Iron Age habitation patterns on the southern and ...

2 downloads 0 Views 4MB Size Report
'Are barrows indeed invariable companions. 12 The abandoning of a ...... Augustan legionary fortress and Flavian canabae legionis at Nij- megen (PhD thesis ...
Michael Meyer (Hrsg.): Haus – Gehöft – Weiler – Dorf. Berliner Archäologische Forschungen 8 (Berlin 2010)

Iron Age habitation patterns on the southern and northern Dutch Pleistocene coversand soils: The process of settlement nucleation Stijn Arnoldussen (Groningen), Richard Jansen (Leiden)

Introduction Studying Dutch Iron Age settlements provides a chance to understand the crucial trajectories of change that differentiate the transient and dynamic settlement pattern of the later second and first millennium BC and the settled and nucleated settlement pattern that is characteristic for the Roman period (e.g. Waterbolk 1995, 17; Wesselingh 2000, 213–214; Vos 2009, 225). The first is described as the ‘wandering farmstead model’ and thought typical of Bronze Age and Iron Age communities, both in northern and southern Netherlands (e.g. Schinkel 1998, 26; 2005; Harsema 2005, 548). It is generally assumed that only with the first century AD, the first formally recognizable ‘villages’ emerge. They can be identified as small village communities in a social sense: not only by their nucleated nature, with larger numbers of houses but also by the presence of outer boundaries shaped as fences or ditches (e.g. Waterbolk 1995, 21; 23; Harsema 2005, 550; Wesselingh 2000, 214–217). The process of settlement nucleation is traceable by two, interrelated, processes. First, there appears to be an increase in location stability, or alternatively, a decrease in relocation distance, during the final millennium BC. Over time, it appears that houses are more frequently rebuilt at closer distances to their predecessors, eventually suggesting a long-term occupancy of individual house-sites. Second, during the last centuries BC, settlements are increasingly evidenced by the number of closely situated – and judging by type and orientation presumably contemporary – houses, occasionally with ditches surrounding several house(-site)s. Although these processes have been studied systematically for the southern Netherlands’ sandy soils (Gerritsen 2003), no comparable systematic study is available for the northern sandy soils, despite the availability of a vast potential data set. Rather, much attention has been devoted to the typological seriation of house types (e.g. Huijts 1992; Waterbolk 2009) that provides us with a reasonably fine-grained chronological control that allows tracking changes in settlement dynamics in the long-term. Consequently, in this contribution we will try to align the long-term narratives explaining Iron Age habitation patterns for both the southern- and northern coversand landscapes. Is the ‘wandering-farmstead model’ characteristic for the whole Iron Age? Is it true that only from

the first century AD onwards larger settlements or ‘villages’ can be recognized archaeologically?

Periodization; chronology of the Dutch Iron Age First we want to discuss briefly the periodization of the study area. Throughout much of later prehistory the material culture recovered from the northern Netherlands hints at cultural affiliations with northern Germany and southern Scandinavia, whereas south of the main river area influences of central European and Atlantic origins dominate (cf. Arnoldussen 2008, 23–24). Accordingly, periodization systems derived from both Montelius’ and the Hallstatt/La Tène scheme have been used in the Netherlands (Van den Broeke 2005, 480). Based predominantly on analyses of funerary assemblages, an additional array of (material) culture groups have been postulated, albeit with sometimes quite different chronological boundaries (fig. 1). The validity of such, sometime rather arbitrary, bound-up cultural zones is frequently – and in our view; rightly – criticized (e.g. Lanting/Van der Plicht 2003, 151; Van den Broeke 2005, 489 note 12). Absolute dates for main periods of the Iron Age are frequently derived from beyond the present-day Netherlands. The start date of 800 cal BC is based on Pare’s (1991) emergence of Gündlingen swords assigned to the Hallstatt C/D transition and/or the dating of the construction wood of the Hallstatt C burial at Wehringen (Lanting/Van der Plicht 2003, 133; 135). The start of the Middle Iron Age at 500 BC – albeit originally defined by the start of the RuinenWommels (op. cit., 171–172���������������������� ; Waterbolk �������������������� 1962) and Marne (Van den Broeke 1987; Lanting/Van der Plicht 2006, 342) pottery traditions – conveniently nearly coincides with the Hallstatt / La Tène transition (Lanting/Van der Plicht 2003, 135; Van den Broeke 2005, 480; 482; 489). ��������������� Similarly, the Middle- to Late Iron Age transition at 250 cal BC is now almost synchronized with the La Tène B–C transition placed at 270 cal BC (Lanting/Van der Plicht 2006, 251) and is ultimately based on fibula morphology elsewhere (op. cit., 247, ref. to Polenz 1971). For the end of the Iron Age, the presence of Roman troops during Drusus’ Elbe campaign in 12 BC is often used (Van Es 1972, 25; Waterbolk 1995, 17), although recently a start date of 19 BC has Originally it was based on the first occurrence of glass bracelets; �������������������������������������������������������������������� Lanting/Van der Plicht 2006, 271. 



See also discussion in Arnoldussen 2008, 77–85. �����������������������������������������������

380

Stijn Arnoldussen, Richard Jansen

Fig. 1 Periodization schemes for the Dutch Iron Age and parts of adjacent periods (after Van den Broeke 2005, 480 fig. 21.2; Lanting/Van der Plicht 2003, 134; 2006, 250–252; 260).

Fig. 2. Palaeogeographic map of the Netherlands, ca. 500 BC, with the tidal inlets in the western coast (Oer-IJ, Rhine, Meuse and Scheldt; north to south) (© RACM/Deltares, 2008).

been forwarded based on the building date of 19–16 BC for a Roman fort at the Hunerberg in Nijmegen (��������� Driessen 2007, 28; Kemmers ��������������������� 2005, 48–49).

adjacent salt-marshes and caused sedimentation in times of (spring)floods and high-tides, as well as drainage in times of low-tides and precipitation (fig. 2).

Varied landscapes; a palaeogeographic introduction to the Netherlands during the Iron Age During the Iron Age, the present-day Netherlands included regions with a dynamic Holocene build-up of sediments and regions without sedimentation since the Pleistocene Weichselian period, apart from local peatgrowth, sand-drifts and anthropogenic plaggen soils. Iron Age habitation, albeit of a different nature, has been documented in both landscapes.

The coastlines The northern coastline of the Netherlands remained open, as due to marine patterns of sand transport, tidal inlets were rarely closed-off (De Mulder et al. 2003, 229; Vos 2006, 8–9). Marine creeks permeated deep into the ���������������������������������������������������������������������� In this contribution, we will adhere to the calibrated date ranges as suggested for the various periods in fig. 1, although specifically without implying that phenomena discussed for such periods are confined exclusively chronologically or geographically to the corresponding ‘culture-group’. 

Iron Age occupation of these dynamic environments was restricted to natural elevations including raised marsh bars, levees of rivers and creeks, and the raised dwellings known locally as wierden or terpen. Colonization of these landscapes seems to have started around the 6th century BC (cf. Lanting/Van der Plicht 2003, 168), when creek sediments at Middelstum, Jengum and Boomborg were settled without prior artificial raising of the soil (Boersma 2005, 563–567). In comparable settings in the Weser and Ems estuaries, occupation of Late Bronze Age and Early Iron Age date is evidenced (op. cit., 561–562), ���������������������������������������������� suggesting that habitation prior to the 6th century BC cannot be excluded (cf. Lanting/Van der Plicht 2003, 204). At the transition point from the northern to the western coast, the relatively small isolated Pleistocene outcrop of Texel, surrounded by coastal marshes and inland peat areas, allowed continuous habitation from the Middle Bronze Age to the Late Iron Age (Woltering 2000). The western coastline was relatively closed, with four main fluvial systems debouching into the North Sea (fig. 2). In this coastal zone, habitation took place on the dune deposits as well as on particular well-drained parcels of the near-coastal peat areas (Lanting/Van der

Iron Age habitation patterns

Plicht 2003, 168). The sites are renowned for their excellent organic preservation. Research at – for example – Santpoort-Spanjaardsberg and Monster-Het Geestje has shown that sand-drifts frequently buried previous (Middle-)Iron Age occupation phases (Van Heeringen 2005, 591, cf. Lanting/Van der Plicht 2003, 207). Unfortunately, the study of sites in this geogenetic setting is hampered by the often several meters` thick cover of younger dune deposits (op. cit., 591). Particularly in the 8th to 7th (cf. Lanting/Van der Plicht 2003, 167) and 4th–1st centuries BC the near-coastal peat areas appear well-suitable to Iron Age habitation (cf. Van den Broeke 1993), but these areas were almost devoid of habitation in the first century BC.

The inlands In the extreme northeastern part of the Netherlands, directly south of the northern salt-marshes, lies a boulder clay plateau with a variable aeolian sand cover (fig. 2, cf. Waterbolk 1995, 1; 3). This Saale period glacial formation, enclosed by peat-growth, is known as the Hondsrug and provided an attractive settlement area in the Iron Age. Various Iron Age settlements were uncovered by the university of Groningen, among which are the well known sites of Hijken (e.g. Harsema 1991) and Wijster (Van Es 1967; Waterbolk 1995, 22). In the central river area fluvial deposits, such as cut-off swales, levees and crevasse-splay deposits, were occupied during the Iron Age. Like with the coastal dunes, ongoing geogenetic activity complicates archaeological research. Former settlement locations may be eroded or covered by thick layers of sediment, both obscuring them from view. At the point were the main rivers discharged into the coastal zone, Middle and Late Iron Age habitation of their levees was dense (Van Heeringen 1987; 1992, 35; 227–232; 2005, 582). From the central river area, Early Iron Age occupation is locally known from inactive levees with thin crevasse-splay covers and crevasse splays proper. Repeated on-site fluvial sedimentation may have E.g. Assendelft (Therkorn 1987; et al. 1984), Spijkenisse and Rockanje (Van Trierum 1992; 2005), Schagen (Therkorn 2004, 130–134), Velserbroek-Hofgeest (Kok 2008, 106–118), Den Haag (De Hingh/Van Ginkel 2009), Vlaardingen-Vergulde Hand (Eijskoot/De Ridder 2003; 2004). 

Except ���������������������������������������������������������������������� for the area north of the Oer-IJ inlet, to where occupants may have relocated; Van Heeringen 2005, 582–583. 

Other ��������������������������������������������������������������������� sites – of equally extensive scale yet published predominantly in Dutch – are Angelslo–Emmerhout (Kooi 2008a–b), Borger (De Wit 2005; Kooi/De Wit 2005), Noordbarge (Harsema 1980; Waterbolk 1995, 15) and Peelo (Kooi/De Langen 1987; Kooi 1991/92). 

E.g. Zijderveld (Theunissen/Hulst 1999, 156–185), Wijk bij DuurstedeDe Horden (Hessing 1989), Culemborg-Den Heuvel (Arnoldussen/Van Zijverden 2004). 

E.g. Kesteren-De Woerd (Sier/Koot 2001), Geldermalsen-Hondsgemet (Van Renswoude et al. 2009), Tiel-Passewaaij (Heeren 2006), NijmegenOosterhout (Van den Broeke 2001), Cuijk (Ball 2007). 

381

prevented habitation, but also served equally to protect or erode older Iron Age occupation traces (cf. Arnoldussen 2008, 415–417). To the north and northeast of the central river area, Saale ice-pushed deposits and gently undulating coversand landscapes are found respectively. Iron Age habitation took place on the sandr-slopes and ice-pushed plateaus (e.g. Verwers 1972, 87; Van Tent 1988; Bijlsma/Schrijer 2003). The coversand ridges in the central-eastern Netherlands have seen relatively more (extensive) research, in which various Iron Age settlements have been uncovered�.10 The southern Netherlands coversand areas are geogentically quite comparable to the central-eastern ones, as they too consist of gently sloping coversand ridges separated by brook-valleys and small local marshes. The archaeological research within this region benefits from a long tradition of academic interest. Since the 1970s, the universities of Leiden and Amsterdam have been involved in extensive fieldwork projects, in which often large areas were uncovered.11 It is particularly within these large-scale projects that questions on the dynamics of Iron Age settlements may be answered best (e.g. Schinkel 1998; Gerritsen 2003; Kerkhof 2008). In the southeastern most part of the Netherlands, some tertiary outcrops can be found, as well as deposits of aeolian loessic sediments. From the latter some Iron Age settlements are known, yet they differ in details – most notably by the different roof-bearing structure of their houses – from those of the southern coversand areas (Lanting/Van der Plicht 2003, 167; Van Hoof 2002; 2007; Tol/Schabbink 2004). Despite the significant variability in geological genesis – Pleistocene versus Holocene, dunes versus coversand ridges, active versus largely inactive landscapes – Iron Age habitation has been uncovered in all of them, albeit with a sometimes different nature. Here we will particularly use the evidence from the southern and northern coversand areas to investigate the role(s) of increased location stability and nucleation, and thus try to determine, whether it is indeed only with the first century AD that This area is comparatively poorly researched with only a few larger excavations: Ede–De Vallei (Bijlsma/Schrijer 2003), BennekomStreekziekenhuis (De Leeuwe 2008), Apeldoorn-Nuon terrein (Williams 2007, 26 ff). 

E.g. Zutphen-Leestense enk (Fontijn 1996), Deventer-Colmschaterenk (Verlinde 1999; 2000; Klomp/Hermsen 2002; Hermsen 2003), Raalte (Groenewoudt et al. 1998; Van der Velde 2007), Dalfsen (Van der Velde et al. 2001); Epse (Prangsma 2002), Zutphen-Looërenk (Bouwmeester et al. 2008, 86–200). 10

������������������������������������������������������������������ See Gerritsen 2003, 26–28 for a complete overview and references. Here we mention the largest excavations: Someren (ca. 20 ha; Kortlang 1999), Weert (ca. 15 ha; Tol 1998), Nederweert (8,3 ha; Hiddink 2005a), Lieshout (8,6 ha; Hiddink 2005b), Oss (ca. 60 ha; Fokkens 1996). The municipality of Breda uncovered large areas during excavations in the 90’s (32 ha; Koot/Berkvens 2004). 11

382

Stijn Arnoldussen, Richard Jansen

Fig. 3. Model of the Bronze Age to Early Iron Age territorial organisation by Roymans/Kortlang (1999; after Roymans/Kortlang 1999, 52 fig. 10).

recognizable ‘villages’, with all the social implications of co-residence – as reflected in mutual support, feelings of togetherness etc. (Arnoldussen 2008, 66–68) – can be identified.

to Bronze Age farmhouses?’ and ‘To what extent do such models hold true outside the chronological and regional confines for/from which they have been compiled?’ are valid, yet unanswered.13

Origins: Bronze Age settlement patterns In order to evaluate the Iron Age settlement dynamics it is necessary to examine the point of departure. For the Bronze Age only a few models of settlement dynamics exists (Arnoldussen 2008, 76–85), and their applicability is limited due to the ����������������������������������������� fact that these are predominantly descriptive models. They provide only a simplified visual representation of archaeologically documented patterns. This means that several important aspects remain unclear, of which causality, representativeness and scale are most salient (op. cit., 84). For example, the model compiled by Roymans and Kortlang (1999; fig. 3) clearly shows the essence of most models of Bronze Age habitation patterns: after a given period of habitation, the house(-site) was relocated to another location which was only infrequently priorly used. In short, farmsteads ‘wandered’ (e.g. Schinkel 1998, 26; 2005; Harsema 2005, 548)��. Set aside the obvious question why these farmsteads wandered, the model leaves much to the imagination.12 Questions like ‘How many farms functioned contemporaneously?’, ‘Are barrows indeed invariable companions The abandoning of a house(-site) is generally explained by assumed limited durability of the construction wood, depletion of fields, or by household dynamics; see detailed discussion and criticism in Arnoldussen 2008, 88–96. 12

The Middle Bronze Age For the Middle Bronze Age B, the data uncovered in the southern Netherlands do indeed seem to reflect a system of wandering farmsteads. At Oss, where the scale of excavation is sufficient to make inferences on absence of phenomena, farmhouses from the Middle Bronze Age B are found several hundreds of meters apart (fig. 4) and all house-sites appear to have hosted only a single housephase (Fokkens 1991). This may indeed be the imprint left by a system of ‘wandering farmsteads’. The limits of house-sites and settlements are however difficult to establish, as is the meaning of several clusters of pits without houses in the direct vicinity. In the northern Netherlands, during the Middle Bronze Age B the single-phased house-sites are also common (e.g. Harsema 1991; Kooi 2008a; Arnoldussen 2008, 206). They are, however, supplemented by a fair number of house-sites that show multiple use-phases. Not only are houses overbuilt (c. 18 %), but the farmhouse of c. 31 % of the house-sites was extended (Arnoldussen 2008, 209; 405). At Angelslo-Emmerhout (fig. 5; Kooi 2008b), this resulted in houses with a maximum length of 75 m. For initial answers; Arnoldussen 2008; 85; 394–395; Bourgeois/ Arnoldussen 2006. 13

Iron Age habitation patterns

383

Fig. 4. Settlement remains from the Middle Bronze Age B (c. 1500–1000 cal BC) at Oss-Ussen. Note that house-sites seem to have hosted only a single house-phase.

Fig. 5. Overview of the Middle Bronze Age B farmhouses at Angelslo-Emmerhout (after Kooi 2008b). Inset shows the possible phasing of farmhouse 52 (after Kooi 2008a, 66 fig. 8 c).

384

Stijn Arnoldussen, Richard Jansen

Fig. 6. Late ���������������������������������������������������������������������������������������������������������������������������� Bronze Age houses from the southern and central part of the Netherlands (1: Tiel-Medel 8 (river area; Van Hoof/Jongste 2007, 77 fig. 5.14), 2: Breda (Berkvens 2004, 103 fig. 6.10), 3: Boxmeer (Van der Velde 1998, 23 fig. 3.8), 4: Sittard (loess area; Tol/ Schabbink 2004, ��������������������������������������������������������������������������������������������������������������������������� 27 fig. 15) and north(east)ern area of the Netherlands and adjacent Germany (���������������������������������������� 5: Elp (Waterbolk 1964, 105 fig. 6), 6: Daverden (D; Precht 2004, 398 fig. 1), 7: Roden (Harsema 1993, 101 fig. 1) and 8: Leesten (Fontijn 1996, 39 fig. 3).

The Late Bronze Age For the Late Bronze Age (c. 1100–800 BC), a continued validity of the ‘wandering farmstead’ is assumed, but again significant differences between the settlement patterns of the southern and northern coversand areas can be outlined. Paramount is a difference in the recognizability of ground plans of houses for this period (fig. 6). In the southern Netherlands’ coversand areas, former rigidly applied rules on the spacing of roof-bearing posts (Arnoldussen/Fontijn 2006, 296) no longer seem to be adhered to during the Late Bronze Age, leaving archaeologists with a variability in house plans that decreases detectability and reduces their use as parallels for other houses within and between regions (Arnoldussen 2008, 223). Shared characteristics are limited to a roof-bearing structure based on two or three rows of main roof-supports, and a generic wide and short rectangular shape (fig. 6, 2–4). It is not only the inner structure of houses that appears to be less (strictly) rule-bound in the Late Bronze Age. Also the layout of settlement sites is different compared to the preceding period. In the southern Netherlands, houses and

house-sites are difficult to recognize and the limits of both house-sites and settlements as a whole remain difficult to identify, with most evidence consisting of isolated (tentative) houseplans.14 In the northern Netherlands, houses of the Elp-type (��������������������������������������� Arnoldussen 2008, 211–212) show better recognizability and greater length compare to those houses known from the southern coversand. More importantly, Elp-type houses continue the previously established northern tradition of prolonging house-site use-life by (repeatedly) extending houses (fig. 7) or by rebuilding on formerly settled areas, such as at the well-known site of Elp (Waterbolk 1987; Bradley 2002). Besides the differences, a new phenomenon occurring both in the southern and northern coversand areas should be mentioned: the emergence of urnfields. Over hundreds of urnfields are known which are frequently interpreted as a static element within the otherwise dynamic settlement system (Gerritsen 2003, 235–254).

For examples see Van der Velde 1998; Tol/Schabbink 2004 and Minsaer 2004. For discussion see Arnoldussen 2008, 227. 14

Iron Age habitation patterns

385

Fig. 7. Overview of the Late Bronze Age farmhouses at Angelslo-Emmerhout (after Kooi 2008a; 2008b). Inset shows the possible phasing of farmhouse 36 (after Kooi 2008a, 62 fig. 4).

Iron Age settlement patterns It is still generally thought that there were few changes in settlement dynamics with the start of the Iron Age. Periodical relocation of farmhouses is still used to describe the settlement dynamics. What does change is that for this period – more than before – the relocation of houses is thought to be linked with the (re-)use of fields within a ‘celtic field’ system (Harsema 2005, 543; Schinkel 2005, 519; Gerritsen 2003, 173). In addition, nucleation is assumed to be manifest by the Late Iron Age (Waterbolk 1995, 15; Gerritsen 2003, 176; Schinkel 2005, 517). House-sites hold more use-phases, suggesting the rebuilding, rather than overbuilding (cf. Arnoldussen 2008, 77) of previously occupied houses. After all, more differentiation between settlements is detectable before the end of the Iron Age. These patterns will be discussed in more detail below.

Early Iron Age: Celtic field relationships To start, the relationships between Celtic field agriculture and Iron Age occupation are prone to be more complex than stated above. Foremost, it remains to be determined when exactly these field systems took on their typical

banked character, as dates prior to the 4th century BC for raised Celtic fields walls are almost lacking (Gerritsen 2003, 167; 176; Spek et al. 2003; Harsema 2005, 543). If the raising (and cultivation) of the banks is a relatively late – i.e. Middle Iron Age or younger – characteristic, Gerritsen’s paradox that ‘the ��������������������������������� use of the banks should perhaps be regarded as an element of a system that replaced the Celtic field system rather than as a characteristic of Celtic field agriculture itself’ (Gerritsen 2003, 174 our emphasis) may very well be true. Second, very few Celtic fields are known from the southern Netherlands (op. cit., 168 table 4.11), in contrast to numerous Iron Age settlements. It remains to be seen whether more intensive post-depositional transformations like reclamation and/or agricultural use are solely responsible (Gerritsen 2003, 170), or whether this represents a real difference in agricultural strategies between the different regions. Third, associations between Iron Age occupation traces and Celtic-field traces are frequently awkward. For example, the Middle to Late Iron Age houses of PeeloKleuvenveld seem to overlap with the Celtic field banks (fig. 8 A), rather than to be ‘conveniently’ situated within them, which one would expect when particular plots were indeed used as house-sites (Harsema 1990, 39–40). Similarly, the Middle to Late Iron Age houses

386

Stijn Arnoldussen, Richard Jansen

Fig. 8. Location of the Iron Age habitation traces in relation to Celtic field walls as observed by aerial photography (medium grey regions) at Peelo (A; after Kooi 1991/92) and Hijken (B; after Waterbolk 1995, 12 fig. 10). Note that fact that Iron Age houses and Celtic field walls often overlap.

Fig. 9. Early Iron Age habitation clusters comprising houses, (rebuilt) outbuildings and fences at Angelslo-Emmerhout (A; after Kooi 2008b, fig. 3b) and Someren-Waterdael (B; after Kortlang 1999, 177 fig. 20, the darker lower right corner represents a peaty depression).

at Hijken are difficult to align with the fences and Celtic field walls as identified through aerial photography (fig. 8 B). Moreover, it can be doubted whether the mobility of house-relocation should actually be connected to the use of Celtic fields, as it remains unsubstantiated that the cycles of agricultural shifts15 should coincide with those of the household (cf. Van Beek 2001, 57). There is, for instance, no evidence to assume that after several decades all agricultural plots in the vicinity of the house that could be economically viable or cost-efficiently exploited were rendered unsuitable for agricultural use (Arnoldussen 2008, 95). 15

E.g. fields, fallow periods and/or crop rotations.

In general, during the Early Iron Age settlements from both coversand landscapes display predominantly singlephased house-sites (but see Schinkel 2005, 524). In various excavations, a cluster of a farmhouse, several outbuildings such as barns and/or granaries and sometimes fences can be identified (fig. 9). Occasionally pits and a well are also found. Although some long-term continuity of house-site usage is suggested by the (repeatedly) rebuilt granaries, the houses display only a single use-phase. This in any case suggests periodical relocation, although the motives behind and periodicity of such relocations remain beyond our grasp.

Iron Age habitation patterns

387

Fig. 10. Middle to Late Iron Age (A) and Early Roman Period occupation at Noordbarge (after Waterbolk 1995, 15 fig. 13). The dashed outlines in A indicate ground plans of tentative Middle to Late Iron Age dating; the irregular and sub-rectangular smaller outlines in B represent semi-subterranean workshops (dutch: hutkommen) and the houses in B are enclosed by a fence placed in a bedding trench.

Middle Iron Age By the Middle Iron Age, prolonging the use-life of previously occupied house-sites by renewing or extending the house appears to become more frequent. It may very well be that the overbuilding of the Middle to Late Iron Age house at Hijken (fig. 8 B) reflects renewal rather than accidental overbuilding, even if the houses are of slightly different subtypes.16 It is tempting to interpret the extension of the Middle- to Late Iron Age farmhouses at Peelo-Es (Kooi 1991/92) along similar lines. Moreover, several of the Iron Age sites in the northern Netherlands, where extending or rebuilding of farmhouses has been documented, evolve into fenced-off Roman period settlements (fig. 10).17 The westernmost Middle or Late Iron Age house-site at Noordbarge (fig. 10 A) hosted a Hijken-type house that was first extended and then replaced by a farmhouse that exceeded the first one in length. 18 The house-site directly to the east-northeast of this one, also initially hosted a small Hijken-type house that was in its second phase rebuilt significantly larger, expanding both the assumed living area and byre. In addition to increased locational stability, the Middle Iron Age also shows an increase in the number of (possible) house plans and house-sites. The high(er) numbers of houses in close proximity at Hijken and Noordbarge can be inferred from figures 8 and 10 (cf. Harsema 2005, 549), but it is expected that also at Borger densities of Middle and Late Iron Age houses 16

Respectively, subtypes Hijken and Zwinderen; Waterbolk 2009, 55; 61 f.

E.g. Noordbarge (fig. 10) and Peelo-Es; Kooi 1991/92, yet not at Hijken. 17

18

See discussion of its dating in Lanting/Van der Plicht 2006, 322.

will prove above average.19 In any case, it is evident that in the Middle Iron Age, in northern and southern Netherlands, settlements comprising multiple phase house-sites and those consisting of single-phased house-sites20 were both common (cf. Kortlang 1999, 172 fig. 17). For the southern Netherlands, the occupation density of Haps (20 houses in a cluster of 80 by 60 m; Verwers 1972, appendix I; Gerritsen 2003, 184 fig. 4.25) is often interpreted as reflecting domestic aggregation (Gerritsen 2003, 184), but Schinkel (2005, 537) has warned that in this particular case the limited space on this pinnacle of a coversand ridge may have caused abnormal densities. To counter such caveats, Oss–Ussen forms an excellent comparison, as space was abundant here during the Early and Middle Iron Age. Here, centuries after the initial Bronze Age habitation (fig. 4), at least three to five house-sites datable to the Early Iron Age could be identified.21 In contrast, the same settlement territory hosted as many as 21 Middle Iron Age house(-site)s (fig. 11, A). It is improbable that poor recognizability caused this, as the Early Iron Age house type – through its long-side entrances, roof-supports outside the walls, rectangular shape and frequent wall-foundation ditches (cf. fig. 9 A) – was well recognisable and common in many regions (Arnoldussen 2008, 292–230). Moving towards the Late Iron Age, a total of 50 houses can be Based on Kooi 1996; Kooi/De Wit 2003; 2005; pers. comm. L. van der Meij, May 2009. 19

E.g. Herpen-Wilgendaal (south; Ball/Jansen 2002, 100 fig. 7.1) or Eelde-Fluitenberg (north; Schrijer/De Neef 2007). 20

Pits and wells may furthermore indicate house-sites, where a house plan has not been recognised or is situated outside the excavation. 21

388

Stijn Arnoldussen, Richard Jansen

Fig. 11. Left: Early (open rectangular regions) and Middle Iron Age houses (shaded rectangular regions) in relation to funerary sites from various periods (halftone grey) at Oss-Ussen (after Schinkel 2005, 524 fig. 23.8). Right: Late Iron Age (open rectangular regions) and Early Roman period Age houses (shaded rectangular regions) in relation to funerary sites from various periods (halftone grey) at Oss-Ussen (after Schinkel 1998; fig. 102; Schinkel 2005, 528 fig. 23.12).

mapped at Oss-Ussen, leaving however the central area unbuilt. From at least the Early Iron Age onwards this space is in use as a cemetery, resulting in an extensive urnfield by the end of the Iron Age. The distribution of houses clearly indicates a nucleation of house-sites that evolved into (fenced-off) Roman period settlements (fig. 11, B). In the northern part, a particular Late Iron Age house appears to have been rebuilt two times on the same house-site, suggesting an intensification of the pattern of prolonged house-site usage (Gerritsen 2003, 61; fig. 11 B, inset). A similar argument has been made for a Middle to Late Iron Age house-site at Someren, where three consecutive farmhouses are situated amidst a semi-circular array of outbuildings (Kortlang 1999, 179 fig. 21; Gerritsen 2003, 102).

Late Iron Age: nucleation The evidence from Oss-Ussen shows (fig. 11, B) that the aggregation of houses peaks during the Late Iron Age, effectively demonstrating that this is an internal development and not a phenomenon influenced by direct Roman rule or proximity. Besides the sheer number of houses, also the first differentiation of settlements is traceable during the Late Iron Age. Whereas in the Middle Iron Age mainly differences in the numbers of house-sites and differences in their number of house-phases could be used to compare settlement sites, in the Late Iron Age the first settlement demarcations appear as well as sites of perhaps altogether different nature. In the southern Netherlands, systems of linear – albeit not meticulously straight – ditches are frequently en-

countered on Late Iron Age settlement sites. At OssHorzak, two 1.5 m wide parallel ditches running between the Late Iron Age habitation could be traced over 300 m (Jansen in prep; Van As 2009). The ditches lie 4 m apart and contained a dozen of sling pellets and clusters of pottery. Late Iron Age house-sites are found both north and south of this structure that formed a real physical barrier in the landscape. Entrances have not been found thus far. A similar example from the northern Netherlands is the 240 m long palisade with several entrances that is associated with Early and Middle Iron Age ceramics at Gees (Waterbolk 1959; Lanting/Van der Plicht 2006, 280; 342), which cannot unambiguously be interpreted as delimiting a settlement. Nonetheless, both cases are evidently large-scale (cultural) landscape features.

Enclosed settlements At Oss-Almstein (fig. 12) a row of Late Iron Age houses conforms to a settlement boundary that probably had a counterpart to the north of these houses. Within these boundaries the partly excavated settlement resembles the occupation density of Haps, albeit that here space was not limited. Settlement boundaries are not exceptional, as only 1 km to the west the site of Oss-Schalkskamp also yielded Late Iron Age houses and outbuildings within a sub-rectangular enclosure consisting of ditches (Fokkens 1992; Gerritsen 2003, 187 fig. 4.27– 28). These ditches (2 m wide and 1 m deep) had been repeatedly re-cut and contained a large and variable assemblage of debris datable to the period 150–75 cal BC. Pottery, metal slag, bracelets, loom-weights, spindle

Iron Age habitation patterns

389

Fig. 12. Late Iron Age occupation traces at Oss-Almstein, possibly reflecting three occupation phases of two to three contemporaneous house-sites.

whorls and a cache of 210 sling pellets were found within the ditches (Gerritsen 2003, 186). Additional examples of enclosed settlements, for which it is not certain yet that these functioned in the last century BC, may be the sites of Weert-Laarderweg and Voerendaal. The Laarderweg site suggests a double enclosure of c. 6.5 ha of which only 25 % is excavated. The enclosure girds only a single outbuilding (Tol 1998; Gerritsen 2003, 184). Voerendaal yielded a Late Iron Age ditch-phase preceding a Roman period villa construction (Willems/ Kooistra 1988).

Fortified settlements Quite remarkably, and despite various large scale excavation campaigns, only a few comparable Late Iron Age enclosed settlements can be identified in the northern Netherlands (Waterbolk 2009, 30–31; Harsema 2005, 549). In this region, however, a distinctly different type of site emerges; the ‘fortified settlement’ (fig. 13; Harsema 1990, 47; 2005, 550, cf. Waterbolk 2009, 139). At Zeijen-I a square enclosure of c. 0.12 ha with three entrances and a ditch and bank was constructed during the final century BC and contained only a few outbuildings (fig. 13 A; Waterbolk 1977; 2009, 133). In a second, still Late Iron Age phase, the enclosure was enlarged and

large three-aisled buildings lined its sides (fig. 13 B).22 Based on the absence of longitudinal functional divisions in these buildings, Waterbolk (1977; 2009, 142–143) has argued that these were not true byre-houses, but possibly sheds or byres. To our minds the presence of refuse pits in most of them (fig. 13 B; op. cit., 166) could support an interpretation as normal houses. At Zeijen-II, a rectangular area of c. 0.16 ha was surrounded by up to seven palisades – partially placed on banks (fig. 13 D; op. cit., 111–113; 2009, 145). At Vries, a rounded rectangular to ovoid area measuring c. 0.2 ha was enclosed by four to seven phases of palisades during the Late Iron Age (fig. 13 C; op. cit., 129–140; 2009, 146). Because of the elaborate gate works at these sites (Waterbolk 1995, 16; 2009, 147–148) and the fact that at Zeijen-II and Vries only outbuildings were recognised on the insides these sites are interpreted as fortified storage- and/or trade centres (op. cit., 15; 2009, 142; Harsema 2005, 550). To conclude, it is clear that already in the Late Iron Age – particularly the last century before the presence of any Roman troops – a well differentiated settlement pattern can be outlined. Single and/or unenclosed house-sites are supplemented by settlements, in which a communal outer boundary may be indicated and sites that appear to have See Waterbolk 2009, 180 for a claimed Roman period date for this phase. 22

390

Stijn Arnoldussen, Richard Jansen

Fig. 13. Fortified sites from the northern Netherlands (Late Iron Age phases). All towards north and to the same scale. A: Zeijen-I; phase 1, B: Zeijen-I; phase 2, D: Zeijen-II; phase 2, C: Vries; phase 2 (after Waterbolk 2009, 70 fig. 44; 142–146 figs. 105–108).

had more of a fortified and probably not solely domestic outlook (Van Es 1990, 55–56; Waterbolk 2009, 184). According to Roymans (1990, 190–198) the site of Voerendaal referred to above may also have been a fortified site in the Late Iron Age. Two specific observations deserve further mention. The first is that in this period the cultural landscape beyond the house-sites proper is marked by extensive trajectories of ditches or palisades.23 While this is as yet a poorly understood It is suggested that house-sites are marked by extensive trajectories of ditches or palisades even earlier: Waterbolk 1959; Arnoldussen 2008, 252–253. 23

phenomenon, it seems more to relate to purposes of delineating and categorization of the wider agricultural landscape as such, rather than to the confinement of settlement space (cf. Arnoldussen 2008, 251–252; 432). The second observation is that, based on our present knowledge, Late Iron Age settlements comprising multiple houses surrounded by a communal ditch appear to be confined to the southern coversand area. However, with the start of the early Roman period, enclosing ditches are occasionally used in the northern region as well (e.g. Fochteloo; Van Es 1990, 55; Waterbolk 2009, 20; fig. 10 B; infra), suggesting that perhaps the absence of Late Iron Age ditched settlements there is more related to chance than to regionally specific traditions.

Iron Age habitation patterns

391

Fig. 14 Early (A) and Middle Roman period (B) occupation at Oss - Horzak (Jansen in prep.) Houses and ditch systems are indicated. Along the northeast excavation limit, the contemporaneous grave field is situated.

Aftermath: Roman period settlements During the Roman period, two trends initiated during the Late Iron Age intensify. The first of these is the tradition of delimiting settlement sites at large (and smaller domestic units within it) with fences or ditches. The Roman periods at Wijster (Van Es 1967) show an enormous diversity in the patterns of subdividing settlement space. Initial parcelling was presumably executed to fence-off house-plots and adjacent fields (Waterbolk 1995, 17; 22 fig. 19a) and was followed by a phase in which individual contemporaneous house-sites appear to be integrated within a shared system of fences (Waterbolk 1995, 23 fig. 19b). During the Late Roman period this is transformed into a system, wherein fences seem to enclose groups of houses, most probably representing village neighbourhoods, and leave open roads (Waterbolk 1995, 17; 24 fig. 19c). From the excavations at Oss, but also Nistelrode-Zwarte Molen (Jansen/Van Enckevort 2008), a comparable image emerges for the Roman Period (e.g. Schinkel 1998; Wesselingh 2000). Parcelling was intensively executed to fence-off house-plots, settlements and adjacent fields. Here we present as an example two phases of the Roman Period occupation at Oss-Horzak (fig. 14; Jansen in prep.; Schamp 2001; Pruysen 2007). In the initial phase,

the ditch system seems to gird the habitation with an irregular to oval ditch system (fig. 14 A). In the later periods this ‘organic’ ditch system was no longer in use (fig. 14 B); instead rectangular plots, sub-rectangular enclosures and double-ditch enclosed areas can be recognised. Particularly this latter technique, straight double ditches around the core of a Roman period settlement, was already documented in the Westerveld excavation 1.9 km to the (south)west (Wesselingh 2000, fig. 72). At the Horzak site it looks like the whole landscape was rearranged, on a large scale. To illustrate the latter, the ditches of the Roman period settlement Oss-IJsselstraat (Wesselingh 1993), 400 m east of Horzak, are orientated in the same direction. The second trend initiated during the Late Iron Age that intensifies during the Roman period is the increased variability or differentiation in the possible configurations of settlements. This is a continuation of a trajectory of change that may have started after the Middle Iron Age. In the longterm, a transition is discernible, wherein delimiting features such as fences, palisades or ditches shift in scope from what initially may rightly be called constructions at a landscape scale (e.g. the Middle Iron Age palisade at Gees and the Late Iron Age double ditches at Oss Horzak) towards a scale that is more focussed on

392

Stijn Arnoldussen, Richard Jansen

the aggregate settlement (e.g. Oss-Almstein; fig. 12) in the Late Iron Age and on ever smaller geographic (and social) units, down to individual households during the Roman period. It is moreover important to stress that during the Roman period, enclosed settlements can involve groupings of houses at various spatial scales, ranging from larger habitation clusters (cf. Van Es 1990, 55– 56) that may represent entire ‘villages’ (e.g. Oss Horzak [fig. 14], Oss – Westerveld [Wesselingh 2000, fig. 72), Noordbarge [fig. 10 B], Wijster [Van Es 1967]) or smaller clusters of a few houses (e.g. Rhee [Waterbolk 2009, 180; Van Giffen 1940], Peelo – Es [Kooi 1991/92]), individual farmsteads (e.g. Peelo-Haverland [Kooi/Delger/Klaasens 1987; Waterbolk 2009, 180]; Oss-De Geer [Jansen/Van Hoof 2003]) or combinations thereof (e.g. Fochteloo [Van Es 1968, 8–9; Van Giffen 1958] or Wijster).

Conclusion In this contribution we set out to investigate whether settlement patterns, defined by the processes of increased location stability and nucleation, is traceable only from the first century AD onwards. Basing on the rich data sets of Iron Age settlements from the southern and northern Dutch coversand areas, we have summarized the changes in settlement patterns from the Middle Bronze Age into the Early Roman period. We have shown that single-phased ‘wandering’ farmsteads as known for the earlier periods (cf. Arnoldussen 2008, 405) continue into the Roman period. However, in the Middle Iron Age the first signs occur that occupants may have invested in a prolonged longevity of particular house-sites. Middle Iron Age houses are regularly found overlapping each other, being extended and possibly even rebuilt on the same farmyard. During the Late Iron Age, the overall density of (sometimes agglomerated) house(-site)s within the wider settlement territory increases significantly (c. 240 % in the case of Oss-Ussen). More importantly however, is the observation that in this period the first enclosed or ‘fortified’ sites can be identified. This suggests that increased domestic aggregation and settlement demarcation were processes that occurred prior to any actual Roman presence and therefore cannot be related causally to the start of Dutch proto-history. Moreover, we have stressed the diversity in nature (fences, palisades and ditches) and scale (settlements at large, neighbourhoods or habitation clusters, individual house-sites) that settlement boundaries may take in both regions during the Roman period.

Summary In this chapter, the rich dataset on Iron Age settlements from the Dutch coversand landscapes is reviewed in order to better understand Iron Age habitation patterns in general and the process of settlement nucleation in par-

ticular. In order to facilitate understanding, brief attention is first paid to specifics of Dutch Iron Age periodization and landscape genesis. A discussion of the different palaeogeographical regions and their Iron Age occupation history in general terms serves as the backdrop for a more in-depth analysis of the settlement patterns in the sections that follow. A brief digression into the settlement dynamics of the preceding Bronze Age (essentially a combination of single-phased house-sites indicative of ‘wandering farmsteads’ and multiple-phased house-sites that show regional patterns such as remarkably frequent extending of farmhouses in the northern coversand areas) moreover allows to study Iron Age settlement patterns from a long-term perspective. With the transition to the Iron Age, the basic model of ‘wandering farmsteads’ is still thought to be applicable, but the relocation is now tied-in with the (re)use of fields within a ‘Celtic field’ system. Moreover, according to traditional views, settlement nucleation and differentiation are visible from the Late Iron Age onwards. We however argue that the supposed interrelation of Celtic field use and Iron Age habitation is far from evident. In the Early Iron Age, single-phased house-sites dominate, but some long-term continuity of house-sites is suggested by the rebuilding of their granaries.��������������� By the Middle Iron Age, prolonging the use-life of previously occupied house-sites by renewing or extending the house becomes more frequent. In addition to increased locational stability (houses are more frequently rebuilt or extended), the Middle Iron Age also shows an increase in the number of houses placed in closer proximity. It is however evident that in the Middle Iron Age, in coversand areas in the northern and southern Netherlands, settlements comprising multiple-phased house-sites and settlements consisting of single-phased house-sites were (contemporaneously) current. In the Late Iron Age houses appear more frequently to be (repeatedly) rebuilt on their house-sites, reflecting intended long-term farmstead usage. In addition, the distribution of houses clearly indicates a nucleation of house-sites that evolves into (fenced-off) Roman period settlements. The evidence from the southern coversand area shows that the aggregation of houses peaks during the Late Iron Age, effectively demonstrating that this is an internal development and not a phenomenon influenced by direct Roman rule or proximity. Besides the increased number of houses, also the first differentiation of settlements (such as enclosed or ‘fortified’ settlements) is traceable during the Late Iron Age and the first settlement demarcations appear. Here again regional distinctions are evident, with settlement( area)s demarcated with ditches being more common in the southern coversand area and ‘fortified’ settlements being more common in the north.

Iron Age habitation patterns

During the Roman period, two trends already initiated during the Late Iron Age intensify. The first of these is the tradition of delimiting settlement sites at large (and smaller domestic units within it) with fences or ditches. The second concerns the increased variability or differentiation in the possible configurations of settlements. In the long-term, a transition is discernible wherein delimiting features such as fences, palisades or ditches shift in focus from what initially may rightly be called constructions at a landscape scale, towards a scale that is more focussed on the aggregate settlement in the Late Iron Age and at ever smaller geographic (and social) units – down to individual households – during the Roman period. It is moreover important to stress that during the Roman period, enclosed settlements can involve groupings of houses at various spatial scales, ranging from larger habitation clusters that may represent entire ‘villages’, or smaller clusters of a few houses, individual farmsteads or combinations thereof. Clearly, from the Middle Iron Age the first signs occur that occupants may have invested in a prolonged longevity of particular house-sites. During the Late Iron Age, the overall densities of (sometimes agglomerated) house(-site)s within the wider settlement territory increases significantly and from this period the first enclosed or ‘fortified’ sites can be identified. This suggests that increased domestic aggregation and settlement demarcation were processes that occurred prior to – and independent of – any Roman presence. We however must stress the the diversity in nature (fences, palisades and ditches) and scale (settlements at large, neighbourhoods or habitation clusters, individual house-sites) that settlement boundaries may take in both regions during the Roman period.

393

Arnoldussen 2008 S. Arnoldussen, A Living Landscape: Bronze Age settlement sites in the Dutch river area (c. 2000–800 BC) (PhD thesis Leiden 2008). Ball 2007 E. A. G. Ball, Het noordoosten van het Land van Cuijk. Een pre- en protohistorisch cultuurlandschap in kaart gebracht. In: R. Jansen/L. P. Louwe Kooijmans, 10 jaar Archol. Van contract tot wetenschap (Leiden 2007) 271–292. Ball/Jansen 2002 E. A. G. Ball/R. Jansen, Van steentijd tot middeleeuwen. Archeologisch onderzoek rond een fossiele beekloop te Herpen-Wilgendaal. Arch. rapp. 11 (Leiden 2002). Berkvens 2004 R. Berkvens, Bewoningsporen uit de periode Late Bronstijd – Midden-IJzertijd (1100–400 v.Chr.). In: C. W. Koot/R. Berkvens (red.), Bredase akkers eeuwenoud. 4000 jaar bewoningsgeschiedenis op de rand van zand en klei. Erfgoed Stud. Breda 1/Rapp. Arch. Monumentenzorg 102, 2004, 95–150. Bijlsma/Schrijer 2003 M. Bijlsma/E. Schrijer, Ede ‘De Vallei’ – DAO. ������������� ADC Rap. 186

Arnoldussen/Van Zijverden 2004 S. Arnoldussen/W. K. van Zijverden, Culemborg-Den Heuvel. Bronze Age remains and Iron Age settlement traces next to a breakthrough channel in the Schoonrewoerd stream ridge. Lunula Arch.

(Bunschoten 2003). Boersma 2005 J. Boersma, Colonists on the clay. The occupation of the northern coastal region. In: ���������������� L. P. Louwe Kooijmans/P. �������������������������������� W. van den Broeke, H. Fokkens/A. L. van Gijn (red.), The prehistory of the Netherlands (II) (Amsterdam 2005) 561–576. Bourgeois/Arnoldussen 2006 Q. P. J. Bourgeois/S. Arnoldussen, Expressing monumentality: some observations on the dating of Dutch Bronze Age barrows and houses. ���������������������������� Lunula Arch. protohist. 14, 2006, 13–25. Bouwmeester/Fermin/Groothedde 2008 J. Bouwmeester/H. A. C. Fermin/M. Groothedde (red.), Geschapen landschap. Tienduizend jaar bewoning en ontwikkeling van het cultuurlandschap op de Looërenk in Zutphen. ����������������� BAAC Rap. 00.068 (Deventer 2008). Bradley 2002 R. Bradley, Entering the present. Legacies of the immediate past. In: R. Bradley (Hrsg.), The past in prehistoric societies (London 2002) 49–81. De Hingh/Van Ginkel 2009 A. E. de Hingh/E. van Ginkel, De archeologie van Den Haag (Utrecht 2009). De Leeuwe 2008 R. de Leeuwe, Prehistorie tussen de loopgraven. Nederzettingssporen en vondstcomplexen in Bennekom-Streekziekenhuis uit de midden-bronstijd tot de midden-ijzertijd, ca. 1500 tot 500 v.Chr. Arch. rapp. 81 (Leiden 2008). De Mulder et al. 2003 E. F. J. de Mulder/M. C. Geluk/I. L. Ritsema/W. E. Westerhoff/T. E. Wong, De ondergrond van Nederland (Groningen 2003). De Wit 2005 M. J. M. de Wit, Op zoek naar de noordelijke grens van de

protohist. 12, 2004, 59–71. Arnoldussen/Fontijn 2006 S. Arnoldussen/D. R. Fontijn, Towards �������������������������������� Familiar Landscapes? On the Nature and Origin of Middle Bronze Age Landscapes in the Netherlands. Proc. Prehist. Soc. 72, 2006, 289–317.

Bronstijd/IJzertijdnederzetting bij Borger. Een archeologisch inventariserend veldonderzoek (IVO) door middel van proefsleuven op bestemmingsterrein `Natuurtransferium’ langs de Koesteeg/Rolderstraat te Borger. ARC Rap. 136 (Groningen 2005).

Address of the Author Stijn Arnoldussen, Groningen ������������������������������������ Institute of Archaeology, Poststraat 6, 9712 ER, Groningen, the Netherlands, [email protected] ������������������������������� /Richard Jansen, Faculty �������������� of Archaeology, Leiden University, Reuvensplaats 3, PO-box 9515, 2300 RA, Leiden, the Netherlands, r.jansen@arch. leidenuniv.nl.�

References

394

Stijn Arnoldussen, Richard Jansen

Driessen 2007 M. Driessen, Bouwen om te blijven: de topografie, bewoningscontinuïteit en monumentaliteit van Romeins Nijmegen, Amersfoort (PhD Thesis). Rapp. Arch. Monumentenzorg 151 (Amersfoort 2007). Eijskoot/De Ridder 2003 Y. Eijskoot/T. de Ridder, De Vergulde Hand 7.074. Basisverslag. VLAK-verslag 22 (Vlaardingen 2003). Eijskoot/De Ridder 2004 Y. Eijskoot/T. de Ridder, De Vergulde Hand 7.073. Basisverslag. VLAK-verslag 20 (Vlaardingen 2004). Fokkens 1991 H. Fokkens, Nederzettingssporen uit de bronstijd en de vroege ijzertijd in Oss-Ussen, wijk Mikkeldonk. In: H. Fokkens/N. Roymans (red.), Nederzettingen uit de bronstijd en de vroege ijzertijd in de lage landen. Nederlandse Arch. Rapp. 13 (Amersfoort 1991) 93–109. Fokkens 1992 H. Fokkens, Oss-Ussen, Schalkskamp. Verslag van het onderzoek in 1991 en 1992. Brabants ������������������ Heem 44, 1992, 157–167. Fokkens 1996 H. Fokkens, The Maaskant project: continuity and change of a regional research project. ��������������������� Arch. Dialogues 3.2, 1996, 197–216. Fontijn 1996 D. Fontijn, De bewoning krijgt vorm. Nedezettingssporen uit de Bronstijd en de IJzertijd op de Leestense enk. In: M. Groothedde (red.), Leesten en Eme. Archeologisch en historisch onderzoek naar verdwenen buurtschappen bij Zutphen (Kampen 1996) 36–47. Gerritsen 2003 F. Gerritsen, Local identities. Landscape and community in the late prehistoric Meuse-Demer-Scheldt region, Amsterdam (PhD thesis). Amsterdam ����������������������������������������� Arch. Stud. 9 (Amsterdam 2003). Groenewoudt et al. 1998 B. Groenewoudt/Th. Spek/H. M. van der Velde/I. van Amen/J. H. C. Deeben/D. G. van Smeerdijk, Raalte-Jonge Raan: de geschiedenis van een Sallandse bouwlandkamp. Rapp. Arch. Monumentenzorg 58 (Amersfoort 1998). Harsema 1980 O. H. Harsema, Het Drents plateau; de latere prehistorische bewoning. In: M. Chamalaun/H. T. Waterbolk (red.), Voltooid verleden tijd? Een hedendaagse kijk op de prehistorie (Amsterdam 1980) 83–102. Harsema 1990 O. H. Harsema, Drenthe’s plaats en rol in de ijzertijd. Nieuwe Drentse Volksalmanak 107, 1990, 39–49 (169–179). Harsema 1991 O. H. Harsema, De bronstijd-bewoning op het Hijkerveld bij Hijken. In: H. Fokkens/N. Roymans (red.), Nederzettingen uit de bronstijd en de vroege ijzertijd in de Lage Landen. Nederlandse Arch. Rapp. 13 (Amersfoort 1991) 21–29. Harsema 1993 O. H. Harsema, Een prijswinnend lot in de ‘Vijfde Verloting’: Een Bronstijderf ontdekt in Roden (Dr.). Paleo-aktueel 4, 1993, 44–48. Harsema 2005 O. H. Harsema, Farms amongst Celtic fields. ������������������� Settlements on the northern sands. In: ���������������� L. P. Louwe Kooijmans/P. ������������������������������� W. van den Broeke/ H. Fokkens/A. L. van Gijn (red.), The prehistory of the Netherlands (II) (Amsterdam 2005) 543–555.

Heeren 2006 S. Heeren (Hrsg.), Opgravingen bij Tiel-Passewaaij 1. De nederzetting aan de Passewaaijse Hogeweg. Zuidnederlandse Arch. Rapp. 29 (Amsterdam 2006). Hermsen 2003 I. Hermsen, Wonen en graven op prehistorische gronden. Archeologisch onderzoek van nederzettingsresten uit de bronstijd en ijzertijd op de percelen Holterweg 59 en 61 te Colmschate (Gemeente Deventer). Rap. ��������������������������������������� Arch. Deventer 11 (Deventer 2003). Hessing 1989 W. A. M. Hessing, Wijk bij Duurstede ‘De Horden’: Besiedlung und Bestattungen aus der frühen Eisenzeit. ������������������ Ber ROB 39 (Amersfoort 1989) 297–344. Hiddink 2005a H. Hiddink, Opgravingen op het Rosveld bij Nederweert 1. Landschap en bewoning in de IJzertijd, Romeinse tijd en Middeleeuwen. Zuidnederlandse Arch. Rapp. 22 (Amsterdam 2005). Hiddink 2005b H. Hiddink, Archeologisch onderzoek aan de Beekseweg te Lieshout. Zuidnederlandse Arch. Rapp. 18 (Amsterdam 2005). Huijts 1992 C. S.T. J. Huijts, De voor-historische boerderijbouw in Drenthe; reconstructiemodellen van 1300 vóór tot 1300 na Chr. Stichting Historisch Boerderij Onderzoek (Arnhem 1992). Jansen/Van Hoof 2003 R. Jansen/L. G. L. van Hoof, Archeologisch onderzoek Oss-De Geer. Bewoningsporen uit de bronstijd en de Romeinse tijd. Arch. rapp. 19 (Leiden 2003). Jansen/Van Enckevort 2008 R. Jansen/H. van Enckevort, Bewoningssporen uit de Romeinse tijd. In: R. Jansen (Hrsg.), Bewoningsdynamiek op de Maashorst. De bewoningsgeschiedenis van Nistelrode van laat-neolithicum tot volle middeleeuwen. ������������������������������������ Arch. rapp. 48 (Leiden 2008) 95–168. Jansen in prep. R. Jansen, The third decade. Excavations at Oss-Horzak (PhD thesis Leiden). Kemmers 2005 F. Kemmers, Coins for a legion. An analysis of the coin finds of the Augustan legionary fortress and Flavian canabae legionis at Nijmegen (PhD thesis Nijmegen 2005). Kerkhof 2008 M. Kerkhof, Why Wander? A critical re-assessment of late prehistoric habitation patterns and the transition from wandering farmsteads to stable villages (unpublished Ma thesis Leiden 2008). Klomp/Hermsen 2002 M. Klomp/I. Hermsen, Archeologisch onderzoek naar de prehistorische bewoning aan de Holterweg 57 te Colmschate, Deventer. Rap. Arch. Deventer 8 (Deventer 2002). Kok 2008 M. S. M. Kok, The Homecoming of Religious Practice (PhD Thesis Amsterdam/Rotterdam 2008). Kooi/De Langen 1987 P. B. Kooi/G. J. de Langen, Bewoning in de vroege ijzertijd op het Kleuvenveld te Peelo, gem. Assen. Nieuwe Drentse Volksalmanak 104, 1987, 151–165. Kooi/Delger/Klaassens 1987 P. B. Kooi/G. Delger/K. Klaassens, A chieftain’s residence at Peelo? A preliminary report on the 1987 excavations. Palaeohist. 29, 1987, 133–144.

Iron Age habitation patterns

Kooi 1991/92 P. B. Kooi, Project Peelo: het onderzoek in de jaren 1977, 1978 en 1979 op de es. Palaeohist. 33/34, 1991/92, 165–285. Kooi 1996 P. B. Kooi, Nederzettingssporen uit de Bronstijd en IJzertijd op de Daalkampen te Borger (Dr.). Palaeo-aktueel 7, 1996, 49–51. Kooi/De Wit 2003 P. B. Kooi/M. J. M. de Wit, Een definitief archeologisch onderzoek langs de rijksweg N34 te Borger, gemeente Borger-Odoorn (Dr.). Arc-Public. 71 (Groningen 2003). Kooi/De Wit 2005 P. B. Kooi/M. J. M. de Wit, Borger revisited, de nederzetting op de Daalkampen (Dr.). ���������������������� Palaeo-aktueel 14/15, 2005, 129–133. Kooi 2008a P. B. Kooi, Bronze Age settlements in Drenthe. In: S. Arnoldussen/H. Fokkens (Hrsg.), Bronze Age settlement sites in the Low Countries (Oxford 2008) 59–68. Kooi 2008b P. B. Kooi, Nederzetting uit de bronstijd en de ijzertijd in Angelslo-Emmerhout (Gem. Emmen). Palaeohist. 49/50, 2007/2008, 327–373. Koot/Berkvens 2004 C. W. Koot/R. Berkvens (Hrsg.), Bredase akkers eeuwenoud. 4000 jaar bewoningsgeschiedenis op de rand van zand en klei. Erfgoed �������� Stud. Breda 1/Rapp. Arch. Monumentenzorg 102, 2004, 95–150. Kortlang 1999 F. Kortlang, The Iron Age urnfield and settlement from Someren’Waterdael’. In: F. Theuws/N. Roymans (Hrsg.), Land and Ancestors. Cultural dynamics in the Urnfield period and the Middle Ages in the Southern Netherlands. Amsterdam �������������������������������� Arch. Stud. 4 (Amsterdam 1999) 133–198. Lanting/Van der Plicht 2003 (2001/2002) J. N. Lanting/J. van der Plicht, De 14C-chronologie van de Nederlandse Pre- en Protohistorie IV: Bronstijd en Vroege IJzertijd. Palaeohist. 43/44, 2003 (2001/2002) 117–261. Lanting/Van der Plicht 2006 (2005/2006) J.N. Lanting/J. van der Plicht, De 14C chronologie van de Nederlandse Pre- en Protohistorie V: Midden- en Late IJzertijd. Palaeohist. 47/48, 2006 (2005/2006), 241–427. Minsaer 2004 K. Minsaer, Bewoningsporen uit de late Bronstijd en de IJzertijd te Ekeren. Voorlopige resultaten van het archeologisch onderzoek in 2002 naar aanleiding van de uitbreiding van goederenspoor 27A. Lunula Arch. protohist. 12, 2004, 109–116. Pare 1991 C. F. E. Pare, Swords, wagon-graves, and the beginning of the Early Iron Age in Central Europe. ������������������������������������ Kleine Schr. Vorgesch. Sem. Marburg 37 (Marburg 1991). Polenz 1971 H. Polenz, Mittel- und spätlatènezeitlicher Brandgräber aus Dietzenbach, Ldkr. ������������������������������������������������� Offenbach am Main. Stud. u. Forsch. NF 4 (Langen 1971). Prangsma 2002 N. Prangsma, De eerste bewoners van Waterdijk II. Archeologisch onderzoek in een nieuwbouwlocatie, Epse, gemeente Gorssel. ADC Rappport 142 (Bunschoten 2002). Precht 2004 J. Precht, Leben am goldenen Bach. Eine Siedlung der jüngeren Bronzezeit bei Daverden in Ldks. Verden. In: M. Fansa/F. Both/H.

395

Haßmann (Hrsg.), Archäologie Land Niedersachsen. ������������� 400000 Jahre Geschichte (Bad Langensalza 2004) 398–402. Pruysen 2007 M. Pruysen, Lang Geleden. Een bewoningsreconstructie van de inheems-Romeinse nederzetting te Oss-Horzak (unpublished MA thesis Leiden 2007). Roymans 1990 N. Roymans, Tribal societies in Northern Gaul – An anthropological perspective. Cingula 12 (Amsterdam 1990). Roymans/Kortlang 1999 N. Roymans/F. Kortlang, Urnfield symbolism, ancestors and the land in the Lower Rhine Region. In: F. Theuws/N. Roymans (Hrsg.), Land and Ancestors. Cultural dynamics in the Urnfield period and the Middle Ages in the Southern Netherlands. Amsterdam Arch. Stud. 4 (Amsterdam 1999) 33–62. Schamp 2001 C. R. C. Schamp, Opgraving Oss-Horzak 2000 (unpubl. MA thesis Leiden 2001). Schinkel 1998 K. Schinkel, Unsettled settlement, occupation remains from the Bronze Age and the Iron Age at Oss-Ussen. The 1976–1986 excavations. Analecta Prehist. Leidensia 30, 1998, 5–306. Schinkel 2005 K. Schinkel, Hamlets on the move. Settlements in the southern and central parts of the Netherlands. In: ���������������� L. P. Louwe Kooijmans/P. ���������������� W. van den Broeke, H. Fokkens/A. L. van Gijn (Hrsg.), The prehistory of the Netherlands (II) (Amsterdam 2005) 519–542. Schrijer/De Neef 2007 E. Schrijer/W. de Neef, Fluitenberg, begraafplaats Zevenberg Gemeente Hoogeveen (Dr.). Deel 2: Opgraving en archeologische begeleiding. Steekproefrap. 2007–02/02 (Zuidhorn 2007). Sier/Koot 2001 M. M. Sier/C. W. Koot, Archeologie in de Betuweroute. Kesteren-De Woerd: bewoningssporen uit de IJzertijd en de Romeinse tijd. Rapp. ������������������������������������������ Arch. Monumentenzorg 82 (Amersfoort 2001). Spek et al. 2003 Th. Spek/W. Groenman-van Waateringe/M. Kooistra/L. Bakker, Formation and land-use history of Celtic fields in north-west Europe – An interdisciplinary case study at Zeijen, The Netherlands. European Journal Arch. 6.2, 2003, 141–173. Therkorn et al. 1984 L. L. Therkorn et al., An early Iron Age farmstead: site Q of the Assendelver Polders Project. Proc. Prehist. Soc. 50, 1984, 351– 373. Therkorn 1987 L. L. Therkorn, The structures, mechanics and some aspects of inhabitant behaviour. ��������������������������������������� In: R. W. Brandt/W. Groenman-van Waateringe/S. E. van der Leeuw (Hrsg.), Assendelver Polder papers 1. Cingula 10 (Amsterdam 1987) 177–224. Therkorn 2004 L. L. Therkorn, Landscaping the powers of darkness/light: 600 BC–350 AD. Settlement concerns of Noord-Holland in wider perspective (PhD Thesis Amsterdam 2004). Theunissen/Hulst 1999 E. M. Theunissen/R. S. Hulst, De opgraving te Zijderveld. In: E. M. Theunissen (Hrsg.), Midden-bronstijdsamenlevingen in het zuiden van de Lage Landen. Een evaluatie van het begrip ‘Hilversumcultuur’ (PhD Thesis Leiden 1999) 156–180.

396

Stijn Arnoldussen, Richard Jansen

Tol 1998 A. Tol, De bewoningsgeschiedenis van Kampershoek. In: N. Roymans/A. Tol/H. Hiddink (Hrsg.), Opgravingen in Kampershoek en de Molenakker te Weert. Zuidnederlandse Arch. Rapp. 5 (Amsterdam 1998) 7–35. Tol/Schabbink 2004 A. Tol/M. Schabbink, Opgravingen op vindplaatsen uit de Bronstijd, IJzertijd, Romeinse tijd en Volle Middeleeuwen op het Hoogveld te Sittard. Zuidnederlandse Arch. Rapp. 14 (Amsterdam 2004). Van As 2009 S. van As, Een fysieke barrière. Een uniek greppelsysteem uit de IJzertijd (unpublished BA thesis Leiden 2009). Van Beek 2001 R. van Beek, Elp in perspectief. Een evaluatie en incorporatie van de verschillende aspecten van een bronstijd-cultuur (unpublished MA thesis Leiden 2001). Van Es 1967 W. A. van Es, Wijster: a native village beyond the imperial frontier, 150–425 A.D. (PhD thesis Groningen 1967). Van Es 1968 W. A . van Es, Friezen en Romeinen. In: J. J. Valma (Hrsg.), Geschiedenis van Friesland (Groningen 1968) 1–45. Van Es 1972 W. A. van Es, De Romeinen in Nederland (Haarlem 1972). Van Es 1990 W. A. van Es, Drenthe’s plaats in de Romeinse tijd (en vroege middeleeuwen). Nieuwe Drentse Volksalmanak 107, 51–62 (181– 192). Van Giffen 1940 A. E. van Giffen, Nederzettingen, grafheuvels, leemkuilen en rijengrafveld tusschen Rhee en Zeijen, Gem. Vries. ������������������������ Nieuwe Drentsche Volksalmanak 58, 1940, 192–200. Van Giffen 1958 A. E. van Giffen, Prähistorische Hausformen auf den Sandböden in den Niederlanden. Germania ������������� 36, 1958, 35–71. Van Heeringen 1987 R. M. van Heeringen, Een horde uit de late ijzertijd in de Stevenshofjespolder in Leiden, Bodemonderzoek in Leiden. Arch. Jaarverslag 1986, 1987, 67–73. Van Heeringen 1992 R.M. van Heeringen, The Iron Age in the Western Netherlands (PhD Thesis Amersfoort 1992). Van Heeringen 2005 R.M. van Heeringen 2005, On unsteady ground. Settlements ������������������� in the western Netherlands. In: ���������������� L. P. Louwe Kooijmans/P. ������������������������ W. van den Broeke/H. Fokkens/A. L. van Gijn (Hrsg.), The prehistory of the Netherlands (II) (Amsterdam 2005) 581–595. Van Hoof 2002 L. G. L. van Hoof, ‘En zij begroeven zich een huis’. Structuur en levensloop van een ijzertijderf in de Zuid-Limburgse lösszone. In: H. Fokkens/R. Jansen (Hrsg.), 2000 Jaar bewoningsdynamiek. Thema’s in het metaaltijdenonderzoek (Leiden 2002) 73–93. Van Hoof 2007 L. G. L. van Hoof, The Iron Age habitation. In: P. van de Velde (Hrsg.), Excavations at Geleen-Janskamperveld 1990/1991. Analecta Praehist. Leidensia 39 (Leiden 2007) 245–278.

Van Hoof/Jongste 2007 L. G. L. van Hoof/P. F. B. Jongste, Een nederzettingsterrein uit de midden- en late bronstijd te Tiel – Medel Bredesteeg. Arch. Rap. 64 (Leiden 2007). Van Renswoude/Van Kerckhove et al. 2009 J. van Renswoude/J. van Kerckhove et al, Opgravingen in Geldermalsen-Hondsgemet. Een inheemse nederzetting uit de Late IJzertijd en Romeinse tijd. Zuidnederlandse Arch. Rapp. 35 (Amsterdam 2009). Van Tent 1988 W. J. van Tent, Elst – ‘t Woud. In: W. J. van Tent (Hrsg.), Archeologische kroniek van de provincie Utrecht over de jaren 1980–1984 (Utrecht 1988) 9–13. Van Trierum 1992 M. C. van Trierum, Nederzettingen uit de IJzertijd en de Romeinse Tijd op Voorne-Putten, IJsselmonde en een deel van de Hoekse Waard. In: A. B. Döbken (Hrsg.), BOORbalans 2: Bijdragen aan de bewoningsgeschiedenis van het Maasmondgebied (Rotterdam 1992). Van Trierum 2005 M. C. van Trierum, Peat farmers. Settlements on the peat to the south of the Meuse estuary. ���������������� In: L. P. Louwe ��������������������� Kooijmans, P. W. van den Broeke, H. Fokkens/A. L. van Gijn (Hrsg.), The prehistory of the Netherlands (II) (Amsterdam 2005) 597–601. Van den Broeke 1987 P. W. van den Broeke, Oss-Ussen: het handgemaakte aardewerk. In: W. A. B. van der Sanden/P. W. van den Broeke (Hrsg.), Getekend zand. Tien jaar archeologisch onderzoek in Oss. Bijdragen tot de studie van het Brabantse Heem 31 (Waalre 1987), 101–119. Van den Broeke 1993 P. W. van den Broeke, A crowded peat area: observations in Vlaardingen-West and the Iron Age habitation of southern Midden-Delfland. ��������������������������������� Analecta Praehist. Leidensia 26, 1993, 59–82. Van den Broeke 2001 P. W. van den Broeke, Een gordel van macht en pracht. Het Midden-Nederlandse rivierengebied in het 1ste millenium v.Chr. In: I. Bourgeois/P. Crombé/G. de Mulder/M. Rogge (Hrsg.), Een duik in het verleden. Schelde, Maas en Rijn in de pre- en protohistorie. Publicaties van het Provinciaal Archeologisch Museum van Zuidoost-Vlaanderen – site Velzeke. Gewone Reeks 4 (Zottegem 2001) 131–156. Van den Broeke 2005 P. W. van den Broeke, Late Bronze Age and Iron Age: introduction. In: L. P. Louwe Kooijmans/P. W. van den Broeke/H. Fokkens/A. L. van Gijn (Hrsg.), The prehistory of the Netherlands (II) (Amsterdam 2005) 477–489. Van der Velde 1998 H. M. van der Velde, Archeologisch onderzoek in de Maasbroeksche Blokken te Boxmeer. Rapp. Arch. Monumentenzorg 64 (Amersfoort 1998). Van der Velde 2007 H. M. van der Velde, De steentijd en de metaaltijden (13.000 v.Chr.–100 n.Chr.). In: H. M. van der Velde (Hrsg.), Germanen, Franken en Daksen in Salland. ADC Monogr. 1 (Amersfoort 2001) 535–564. Van der Velde/Van Benthem/Bloo 2001 H. M. van der Velde/A. van Benthem/S. B. C. Bloo, Een huisplaats uit de late Bronstijd te Dalfsen. Archeologisch onderzoek aan het

Iron Age habitation patterns

bedrijventerrein Welsum te Dalfsen. ADC Rap. 95 (Bunschoten 2001). Verlinde 1999 A. D. Verlinde, Isolated houses in Overijssel during the transition from prehistory to protohistory. In: H. Sarfatij/W. J. H. Verwers/P. J. Woltering (Hrsg.), In Discussion with the Past. Archaeological ��������������� studies presented to W. A. van Es (Zwolle 1999) 77–86. Verlinde 2000 A. D. Verlinde, Inventarisatie en onderzoek van archeologische locaties op en rond de Colmschater Enk te Deventer. Rapp. Arch. Monumentenzorg 75 (Amersfoort 2000). Verwers 1972 G. J. Verwers, Das Kamps Veld in Haps in Neolithikum, Bronzezeit und Eisenzeit. Analecta Praehist. Leidensia 5 (PhD thesis Leiden 1972). Vos 2006 P. C. Vos, Toelichting bij de nieuwe paleogeografische kaarten van Nederland. NOaA hoofdstuk 25 (versie 1.0) (www.noaa.nl). Vos 2009 W. K. Vos, Bataafs platteland. Het Romeinse nederzettingslandschap in het Nederlandse Kromme-Rijngebied. Nederlandse Arch. Rapp. 35 (Amersfoort 2009). Waterbolk 1959 H. T. Waterbolk, Nieuwe gegevens over de herkomst van de oudste bewoners van de kleistreken. Akademiedagen ������������������ 11, 1959, 16–37. Waterbolk 1962 H. T. Waterbolk, Hauptzüge der eisenzeitlichen Besiedlung der nördlichen Niederlande. Offa ��������� 19, 1962, 9–46. Waterbolk 1964 H. T. Waterbolk, The Bronze Age settlement of Elp. Helinium 4, 1964, 97–131.

397

Waterbolk 1977 H. T. Waterbolk, Walled enclosures of the Iron Age in the North of the Netherlands. Palaeohist. 19, 1977, 97–172. Waterbolk 1987 H. T. Waterbolk, Terug naar Elp. In: F. C. J. Ketelaar (Hrsg.), De historie herzien; vijfde bundel ‘Historische Avonden’ (Hilversum 1987) 183–215. Waterbolk 1995 H. T. Waterbolk, Patterns of the peasant landscape. Proc. Prehist. Soc. 61, 1995, 1–36. Waterbolk 2009 H. T. Waterbolk, Getimmerd verleden. Sporen van voor- en vroeghistorische houtbouw op de zand- en kleigronden tussen Eems en IJssel. ������������������������������������������ Groninger Arch. Stud. 10 (Groningen 2009). Wesselingh 1993 D. A. Wesselingh, Oss-IJsselstraat: Iron Age graves and a native Roman settlement. Analecta Praehist. Leidensia 26 (Leiden 1993) 111–138. Wesselingh 2000 D. A. Wesselingh, Native neighbours: local settlement system and social structure in the Roman period at Oss (the Netherlands). Ana���� lecta Praehist. Leidensia 32 (PhD Thesis Leiden 2000). Willems/Kooistra 1988 W. J. H. Willems/L. I. Kooistra, De Romeinse villa te Voerendaal, opgraving 1987. Arch. Limburg 37, 1988, 137–147. Williams 2007 G. L. Williams, Een Bronstijd nederzetting op het voormalige Nuon terrein te Apeldoorn. ADC Rap. 756 (Amersfoort 2007). Woltering 2000 Woltering 2000, Occupation history of Texel, IV Middle Bronze Age – Late Iron Age (1350–100 BC). Ber. ROB 44, 2000, 9–396.

Haus - Gehoft - Weiler - Dorf Siedlungen der Vorrumischen Eisenzeit im ntirdlichen Mitteleuropa

Michael Meyer (Hrsg.)

Haus - Gehoft - Weiler - Dorf Siedlungen der Vorromischen Eisenzeit irn n~rdlichenMitteleuropa Internationale Tagung an det Freien Universitat Berli~i vom 20.-22. M a n 2009

Verlag Marie Leidorf GmbH RahdedWestf. 2010