Copper Age ditched enclosures in central Iberia

PEDRO DÍAZ-DEL-RÍO COPPER AGE DITCHED ENCLOSURES IN CENTRAL IBERIA Summary. The interpretation of European Neolithic enclosures must take account of their wide variability in chronology, size, shape, topographical position and material. Such interpretations should rely on the comparative analysis of the processes at work in particular regions. Newly recovered data from small early third millennium cal BC ditched enclosures in central Iberia, with high densities of features and domestic refuse, support the hypothesis of permanent habitation. This paper argues that the variability in late Neolithic–Chalcolithic enclosures throughout Iberia is a result of the cycling of fission and fusion characteristic of segmentary social dynamics. introduction The economic conditions of production play a key role in the understanding of Neolithic and Copper Age political dynamics. Using the technologies available to them, segmentary societies throughout Europe developed a variety of political fission–fusion processes that only occasionally resulted in the constitution of chiefdoms. The expansion of any segmentary political economy relies on the ability to attract, increase and maintain surplus and labour. Thus, aggregational processes are at the centre of any comprehensive understanding of early prehistoric agricultural societies. Copper Age Iberia (3500–2250 cal BC) presents one of the clearest cases of prehistoric economic intensification (Chapman 1990; Gilman 1981). The introduction of a Neolithic production economy c.5500 cal BC resulted in a gradual consolidation of village communities, a generalised phenomenon by the end of the fourth millennium cal BC. During this and the next millennium, tribal groups developed diverse political dynamics characteristic of segmentary societies, particularly a variety of aggregation processes involving the construction of monumental enclosures. Although Iberia was involved in general European processes, such as those which gave rise to megalithism or bell-beakers, most prehistoric cultural processes can be interpreted without the need of any extrapeninsular referents (Chapman 1990). As a result, Spanish and Portuguese scholarship has had limited international impact, and English-language surveys continue to portray the south-east and south-west as the representative areas of prehistoric Iberia. The Copper Age is probably the best example of this limited view. Stone-walled sites are present throughout most of the Portuguese Atlantic façade and southern Spain, although the south-east Millarian and south-west Vilanova de São Pedro ‘cultures’ are the best known (Fig. 1A). Ditches had been found in this and other regions at least since the 80s, but documentation of their OXFORD JOURNAL OF ARCHAEOLOGY 23(2) 107–121 2004 © Blackwell Publishing Ltd. 2004, 9600 Garsington Road, Oxford OX4 2DQ, UK and 350 Main Street Malden, MA 02148, USA. 107 COPPER AGE DITCHED ENCLOSURES IN CENTRAL IBERIA Figure 1 (A) Distribution of fifth to third millennia cal BC enclosures in Europe (modified from Darvill and Thomas 2001, 8). (B) Fourth to third millennia cal BC sites with ditches and ditched enclosures in Iberia. (1) Papa Uvas; (2) Valencina de la Concepción; (3) Peñon Gordo; (4) La Minilla; (5) Perdigoes; (6) La Pijotilla; (7) Los Pozos; (8) Marroquíes Bajos; (9) Ciavieja; (10) Niuet; (11) Arenal de la Costa; (12) Las Matillas; (13) La Esgaravita; (14) Gózquez de Arriba; (15) Cerro de la Mora; (16) Loma de Chiclana; (17) Las Pozas; (18) San Miguel; (19) Matallana; (20) Las Bodegas; (21) Vega de los Morales; (22) Las Canteras (complete references in Díaz-del-Río 2003). OXFORD JOURNAL OF ARCHAEOLOGY 108 © Blackwell Publishing Ltd. 2004 PEDRO DÍAZ-DEL-RÍO widespread distribution and variable scale, and their interpretation as enclosures (Fig. 1B), only became possible after the ‘empirical revolution’ (Harrison and Orozco 2001). Two factors triggered this revolution: the multiplication of functionally oriented research projects (Gilman 2000, 27) and the transformation of field archaeology, as a direct consequence of the political regionalisation of heritage management (Martínez Navarrete 1998). As a result, excavations with contextual data can now be measured in hectares. The increase in the scale of observation calls for a re-evaluation of the traditional view of the social dynamics of Copper Age Iberia, where the complex societies of the south-east and south-west were contrasted with a non-intensified and ‘primitive’ hinterland. iberian meseta The largest number of these ditched enclosures is found in the Spanish Meseta. This is the largest geographical unit of Iberia, covering more than one-third of its total expanse. Located at the centre of the peninsula, it is a 600 m high, 181,000 km2 Tertiary plateau partially surrounded by mountains. The Central System chain, running north-east–south-west, divides this geographic unit into the so-called north and south Mesetas. Two of Iberia’s main rivers, the Tagus and the Duero, cut through the limestone surface of the plateau in an east–west direction, exposing marls and clays that have created broad valleys with rich agricultural soils and pasture lands. Interfluvial plateaux generally present dry lands with low agricultural potential, although springs and other concentrations of humidity create occasional permanent areas of pasture. Both Mesetas have a continental climate, with a hot dry summer (24°C mean), a cold rainy winter (4°C mean), and strong contrasts in mean rainfall between the peripheral highlands (over 1000 mm/yr) and the river basins (about 500 mm/yr). In the northern Meseta and Tagus valley, alluvial soils have been used for cultivation and pasture at least since the late sixth millennium cal BC (Kunst and Rojo 1999). Until very recently Neolithic to Bronze Age settlements were primarily defined by clusters of pits. Unstratified sites with underground storage facilities, hearths and other domestic features suggested an economy based on shifting cultivation and the herding of sheep and goats. The last ten years of research have changed this consensus view. Copper Age sites are structured settlements with increasing evidence for dwellings, ditches and occasionally stone-walled enclosure (Delibes et al. 1995; Díaz-del-Río 2001). There is evidence for salt exploitation (Delibes et al. 1998), some specialised lithic technology, and a simple copper metallurgy (Rovira 2002). The third millennium BC saw the first peak of agrarian intensification. The economy, based on domestic production, approached the limit of intensification due to the introduction of simple plough agriculture and the increased exploitation of animals for their secondary products. Under these conditions, reasonable options for further intensification and increasing surplus production relied either on expanding the amount of labour input or on developing labour-intensive systems. All this new evidence suggests that Meseta societies had a certain level of social complexity, comparable to that of their VNSP or Millarian contemporaries. By now some fifty ditched enclosures have been documented in the Meseta. In the northern Meseta their existence was discovered in 1999, as a consequence of aerial photographic surveys (Delibes 2001; Olmo 1999). These sites have not been excavated, but ceramic and lithic remains collected on their surface suggest they are mainly of Copper Age date. In the middle Tagus basin (Madrid region), however, the three known ditched enclosures are all the result of OXFORD JOURNAL OF ARCHAEOLOGY © Blackwell Publishing Ltd. 2004 109 COPPER AGE DITCHED ENCLOSURES IN CENTRAL IBERIA open-area excavations (Díaz-del-Río 2001). These sites were all constructed and backfilled in the first half of the third millennium cal BC. problems The newly-discovered ditched enclosures in the Spanish Meseta raise a series of analytical problems that require a three-pronged approach. We must determine, first, what activities were performed at these sites. Second, we must assess how these sites integrate in contemporary regional site distributions. Surveys indicate high densities of Chalcolithic sites, generally clustering in river basins. Recent open-area excavations suggest that a considerable number of these sites were enclosed. As a consequence, spatial analysis of site distributions is required to determine whether there are hierarchical settlement patterns. Third, we must compare the variability in scale of Meseta enclosures with that of similar, contemporaneous sites elsewhere in Iberia. The sizes of enclosed settlements with similar technological and economic bases range from less than a hectare to more than 90 ha. The biggest sites show an increase of labour investment in settlement structures and collective burials, intensification of exchange and emerging craft specialisation. This broad variability seems to reflect different degrees of political complexity in different parts of Iberia. Ditched enclosures are one of the most generalised settlement features in Neolithic and Copper Age Europe. Recent overviews suggest that, over the continent as a whole, such sites span more than three millennia (sixth to third millennia cal BC). They are found from Sicily to Scotland, Portugal to the Black Sea, reaching the Caucasus and the Dnieper valley (Darvill and Thomas 2001). The few general analyses of this phenomenon have shown the difficulties of assigning it a single functional interpretation. Ditched enclosures have been suggested to be villages, aggregational or central assembly areas, hydraulic features, refuges, markets, cult places or burial sites (Andersen 1997, 302–9). There are no recorded enclosures before the Neolithic, which suggests their construction may be related to increasing intercommunal restrictions over landed property rights. The great variability in chronology, layout, size, topographic position and features calls for an interpretation based on contextual evidence within particular regions. Simply stated, there is no such thing as a pan-European ‘enclosure phenomenon’. Since the 1970s, two interpretations have been dominant. Processualists have understood enclosures as monumental habitation sites, reflecting claims over resources, as well as spatial and social hierarchies. Wessex has been this approach’s paradigmatic case study for the rise of corporate chiefly strategies (Renfrew 1973; Earle 1991). More recently, arguments have emphasised ritual and symbolic aspects of the enclosures, interpreting them as collective social spaces constructed by small semi-mobile groups. The empirical base for this argument is the absence of evidence for habitation, the contrasts between enclosed and unenclosed sites, and the generalised presence of ‘anomalous’, often structured, depositions in the enclosures. Finally, there is yet another possible hypothesis: that enclosures are monumentalised habitation sites that claim rights over landed property, but do not necessarily reflect the spatially organised social hierarchies frequent in chiefdoms. These interpretations are all plausible and should be evaluated in the light of existing sets of contextualised regional data. Whatever the particularities of the regional context, however, the first key element in testing these hypotheses is evaluating the presence or absence of permanent habitation at these sites. To address this question the following lines of evidence are relevant: OXFORD JOURNAL OF ARCHAEOLOGY 110 © Blackwell Publishing Ltd. 2004 PEDRO DÍAZ-DEL-RÍO • A strong correlation between site position and areas of high agricultural potential. • An abundance of domestic features: hearths, ovens, silos, post-holes and/or dwellings. • The complete assemblage to be expected in a domestic context. High density and fragmentation of pottery. • High frequency and density of lithic remains. Presence of most of the operational chain. • High density of grinding stones. • Evidence of food consumption. High density of faunal refuse. • Pollen evidence for increasing modification of immediate landscape. In short, arguments should be built correlating site location, density of features and nature of remains. In the following sections we will evaluate these variables with respect to the three ditched enclosures excavated in the Meseta. Finally, we will address the regional and broader implications of our evidence. Location The three sites in the Madrid region (Gózquez de Arriba, Fuente de la Mora, and Las Matillas) are the only excavated enclosures in the Meseta. They are all situated in the lowlands of the middle Tagus basin, but in completely different topographical settings. Gózquez de Arriba (Fig. 2) is located on the slope of a very small secondary valley, some 500 m from a tributary of the river Jarama, the main regional watercourse. It is situated between a 1 km wide floodplain and a complex of low chalkland hills with permanent pasture spots. Fuente de la Mora (Fig. 3) is located on a small river bluff 400 m north of the Henares river. Its ditch-system encloses the north-east side of the occupied area, where visibility is limited. There are permanent and abundant natural springs in the vicinity. Las Matillas (Fig. 4) is situated in the floodplain of the Henares. The three enclosures have different topographical situations. Only Las Matillas would be highly visible. Their only locational similarity is their proximity to watercourses, springs, fertile agricultural soils and areas of permanent pastures. Size, structure and defensive conditions All three enclosures are small in size, ranging from 0.3 to 1.0 ha. At Gózquez, the interior ditch encloses 0.09 ha, the exterior one 0.3 ha, and the total extent of the site is 1.2 ha. The stratigraphic sequence indicates the interior ditch was built first. This was later backfilled, whereupon the exterior ditch was excavated and the interior one re-cut. An episode of heavy erosion following prehistoric occupation removed at least 0.5 m of the archaeological layers in the interior and north-eastern part of the site. The ditch-system was excavated on a slope, thus avoiding the crystallised limestone of the higher hill. The location of the site was selected to minimise construction costs and visibility. At Fuente de la Mora, the topographic setting suggests that only the interior ditch formed an almost complete circle. This ring encloses an area of 0.1 ha within which there is a concentric palisade that in turn encloses 0.06 ha. The total extent of the site is about 1 ha. Erosion of the hill and hillsides has limited the conservation of stratification. As shown in Figure 3, not all ditches were contemporary. The site is situated in a relatively prominent setting, but is easily accessible from the north-east, the side protected by the outer ditches. OXFORD JOURNAL OF ARCHAEOLOGY © Blackwell Publishing Ltd. 2004 111 COPPER AGE DITCHED ENCLOSURES IN CENTRAL IBERIA Figure 2 Gózquez de Arriba (San Martín de la Vega, Madrid). Ditched enclosure and related features. Upper left, visibility from the site. Lower right, dwellings. OXFORD JOURNAL OF ARCHAEOLOGY 112 © Blackwell Publishing Ltd. 2004 PEDRO DÍAZ-DEL-RÍO Figure 3 Fuente de la Mora (Leganés, Madrid). Lower left, location. Lower right, phases. Upper right, dwellings. Centre right, some underground storage pits. OXFORD JOURNAL OF ARCHAEOLOGY © Blackwell Publishing Ltd. 2004 113 COPPER AGE DITCHED ENCLOSURES IN CENTRAL IBERIA Figure 4 Las Matillas (Alcalá de Henares, Madrid). In Las Matillas, partial excavation suggests the existence of a one-ring 0.7 ha causewayed enclosure. Erosion has removed the ditch on the south side of the enclosure, where gravels occur immediately beneath the plough zone. Both its position on the river floodplain and the limited depth and width of the enclosing ditches show a lack of concern for defence. This is confirmed by a roofed space in part of the ditch (Fig. 4). OXFORD JOURNAL OF ARCHAEOLOGY 114 © Blackwell Publishing Ltd. 2004 PEDRO DÍAZ-DEL-RÍO Dwellings and other domestic facilities The lack of evidence for domestic buildings has been used to argue against the hypothesis that enclosures were habitation sites. Domestic structures are infrequent in both enclosed and non-enclosed sites, however. Consequently, the absence of evidence cannot be used as evidence of absence. In any event, despite severe erosion, dwellings, post-holes or foundation trenches are documented at all three sites under consideration here. Five dwellings have been documented at Gózquez (Fig. 2): two are oval, sunkenfeatured buildings, two have foundation trenches (one oval, the other rectangular), and one is circular with a stone foundation. Two of these structures, both seriously affected by erosion, were located in the internal enclosure. Fuente de la Mora has at least four buildings, all with circular foundation trenches. Two were located between the second and third enclosures and probably correspond to the first phase of the ditch-system. The other two are round buildings located in the interior ring during the second phase. At Las Matillas the only roofed structure is located in the ditch. Given that foundation trenches typically are shallow, the survival of only one domestic structure must be due to erosion. In both layout and size (15 to 44 m2) the circular–oval structures found at Gózquez and Fuente de la Mora are identical to other dwellings documented in the Meseta. The second critical aspect of the sites is the high density of domestic underground features. Bell-section pits are common in all the sites. When systematic flotation is applied to their fill, as at Las Matillas, Triticum sp. and Hordeum sp. are recovered. Domestic refuse These sites have high densities of domestic refuse (Table 1). We estimate that around one million ceramic fragments (19 tons) from Gózquez and 700,000 (16.5 tons) from Fuente de la Mora would have been recovered if the sites had been totally excavated. Formal analysis of pottery indicates the presence of a functionally complete domestic set, from large storage jars to cups. The lithic assemblage of Gózquez suggests that the amount of chert may have totalled 39,000 pieces, 17,000 of them flakes. Both Gózquez and Las Matillas have few cores and a small percentage of primary and secondary decortication flakes. The earlier processes of manufacture were probably performed at the flint quarries found throughout the region. Gózquez and Fuente de la Mora also have high densities of complete and fragmented grinding stones. Pollen analysis of sediments from all three sites documents the remarkable human impact on the landscape. Palynological samples from the ditch deposits at Gózquez contain cereal pollen, indicating that agricultural activities were carried out in the immediate vicinity. The evidence suggests that the landscape was an open one, with dominant anthropic pastures (Cichorioideae, Gramineae, Cruciferae, Leguminosae), and with forest species such as oak table 1 Quantitative data from ditched enclosures in the Madrid region Site ha. # ceramic fragments. Kg ceramics # lithic # grinding stones. Kg grinding stones Gózquez de Arriba Fuente de la Mora Las Matillas 0.3 1.0 0.7 112,724 51,453 24,316 2092 1659 407 5894 3800 2924 138 248 ¿ 115 335 ¿ OXFORD JOURNAL OF ARCHAEOLOGY © Blackwell Publishing Ltd. 2004 115 COPPER AGE DITCHED ENCLOSURES IN CENTRAL IBERIA (Quercus ilex-coccifera t.) and juniper reaching no more than 15 per cent of the pollen samples. Gallery woodlands of elm, ash, willow and poplar would also have been present. Palaeoenvironmental data suggest that both Holocene climatic conditions and human action triggered considerable erosion (Martin et al. 2002). Two problems arise when comparing these assemblages with other contemporary sites. First, the three have more refuse than any other early third millennium cal BC sites excavated in the Meseta. Second, because of limited excavation areas in other sites, it is uncertain if those with a high density of refuse were enclosed. In conclusion, all the contextual evidence from these ditched enclosures suggests they were small habitation sites, occupied by few households with an agricultural base. Chronology Radiocarbon dates obtained from Gózquez and Las Matillas indicate that the enclosures were in use and backfilled in the first half of the third millennium cal BC (Table 2). Typological analysis of Fuente de la Mora suggests the same chronological placement. Both absolute chronology and stratigraphical sequence show frequent recuttings and re-formations of the enclosure systems. Gózquez and Las Matillas have residual Neolithic ceramic fragments in Copper Age contexts, and all three have dense Late Copper Age and Early Bronze Age occupation. These EBA contexts are never located directly in the abandoned enclosed area, but are no less than 100 m away. They are represented by clusters of pits and lack any evidence of enclosure. table 2 Calibrated radiocarbon dates from Gózquez de Arriba and Las Matillas (OxCal v.3.5) Site Lab n° BP Cal BC (1 s) Cal BC (2 s) Type Gózquez Gózquez Beta-134866 Beta-134865 4320 ± 130 4100 ± 80 3350-2550 2880-2470 Std. Std. Gózquez Beta-134861 4150 ± 50 2880-2580 AMS Gózquez Beta-134864 4020 ± 60 3350-2650 2870-2800 (15.3%)/2760-2560 (48.6%)/ 2520-2490 (4.3%) 2880-2830 (13.6%)/2820-2800 (5.4%)/ 2790-2660 (40.5%)/2650-2620 (8.7%) 2630-2460 AMS Gózquez Beta-134863 4180 ± 80 Gózquez Beta-134862 4020 ± 50 Gózquez Beta-134858 4100 ± 60 Gózquez Beta-134857 4160 ± 60 2900-2800 (6%)/ 2750-2300 (89.4%) 2920-2560 (93.5%)/ 2530-2490 (1.9%) 2900-2800 (4%)/ 2700-2350 (91.4%) 2880-2550 (86.6%)/ 2540-2490 (8.8%) 2890-2570 AMS Gózquez Beta-134859 4140 ± 50 2880-2570 AMS Matillas Beta-134867 4150 ± 50 2880-2580 AMS 2890-2830 (14.1%)/2820-2660 (50.9%)/ 2650-2620 (1.9%) 2620-2610 (1.3%)/2580-2460 (66.9%) 2860-2810 (16.1%)/2750-2720 (7.2%)/ 2700-2570 (42.6%)/2520-2500 (2.3%) 2880-2830 (13.4%)/2820-2660 (51.2%)/ 2650-2620 (3.6%) 2870-2800 (19.4%)/2780-2770 (3%)/ 2760-2620 (44.9%)/2610-2600 (0.9%) 2880-2830 (13.6%)/2820-2800 (5.4%)/ 2790-2660 (40.5%)/2650-2620 (8.7%) AMS AMS AMS OXFORD JOURNAL OF ARCHAEOLOGY 116 © Blackwell Publishing Ltd. 2004 PEDRO DÍAZ-DEL-RÍO conclusions Several conclusions can be obtained from these enclosures. First, the evidence supports the existence of permanent occupation in all of them. Although Las Matillas is clearly different from the others, the limited available information supports its interpretation as a habitation site. Second, there is no apparent difference in the size or wealth of the sites. All recovered remains are essentially domestic, identical to all contemporary sites in the Meseta. Third, all enclosures are small, and so too are the groups implicated in their construction and occupation. The dynamic expansion and reduction of enclosed areas never exceeds a hectare, although contemporary features can be found up to 300 m away. These relatively constant dimensions may indicate the limits within which segmentary groups fissioned. Fourth, enclosures clearly acted as barriers, monumentalising domestic space and proclaiming restricted access to their resources. Whatever their specific function, they are the first permanent long-term habitation sites documented in the prehistoric Meseta. Finally, although earlier and later occupation is present at the sites, the enclosures were dug and backfilled between 2900 and 2400 cal BC. perspectives Understanding the place of these enclosures in regional contemporary settlement patterns would require a spatial analysis which is impossible with the evidence currently available. If the refuse-densities and sizes of the ditched enclosures are similar, it should be possible to locate them by intensive survey. This would help resolve the ‘palimpsest’ problem generally presented by regional surface collections. In the absence of such spatial analysis and without evaluation of the agricultural productivity of site catchments, it is impossible to demonstrate the presence or absence of settlement hierarchies. There are several arguments against their existence, however. First, it is difficult to imagine that these extremely small sites were politico-economical centres. Second, if these sites were central places, the benefits they obtained are unclear. As Gilman (2001, 78) has stated, chiefs ‘typically feather their nests’, and hierarchical patterns should be somehow reflected in the archaeological record. The scale of ditched enclosures and their characteristics allow an assessment of the limits of political institutions in the Spanish Meseta. The personal ambitions of segmentary leaders in third millennium cal BC Iberia would have been limited by three factors: the environmental and technological conditions for surplus production and accumulation; the capacity to attract, increase and maintain new labour forces; and the social ideological restrictions that supported kinship-based modes of surplus extraction. In central Iberia conditions permitted surplus production and storage, but the scale was never sufficient to finance the expansive interests of local leaders. Production stayed within domestic limits. Under the available technological conditions, increasing surplus production would necessarily imply aggregating social segments, but the size of existing sites does not indicate that would-be chiefs were in fact capable of attracting large labour forces. In other areas of Iberia, the third millennium cal BC presents clear cases of higher population densities, which did not exist either in the preceding Neolithic or the succeeding Bronze Age. As shown in Figure 5, there is a dramatic intra- and inter-regional variability in the size of sites, and consequently, in the labour invested in them. The main trend of most complex settlements seems to be related to the increasing material evidence for internal lineage competition. Many of the larger sites (La Pijotilla, Perdigoes, Valencina de la Concepción, Los OXFORD JOURNAL OF ARCHAEOLOGY © Blackwell Publishing Ltd. 2004 117 COPPER AGE DITCHED ENCLOSURES IN CENTRAL IBERIA Figure 5 Size variability in Iberian Chalcolithic enclosures. (1) Marroquíes Bajos (Zafra et al. 1999, modified); (2) La Pijotilla (Hurtado 1997, modified); (3) Vila Nova de São Pedro; (4) Zambujal; (5) Monte da Tumba; (6) Cerro do Castelo do Santa Justa (Gonçalves 1982, modified); (7) Fuente de la Mora; (8) Gózquez de Arriba; (9) Perdigoes (Lago et al. 1998); (10) Los Millares; (11) Fort 1, Los Millares. (Nos. 3, 4, 5, 10 and 11 in Chapman 1990, modified.) OXFORD JOURNAL OF ARCHAEOLOGY 118 © Blackwell Publishing Ltd. 2004 PEDRO DÍAZ-DEL-RÍO Millares) are associated with megalithic cemeteries with collective (though frequently individually treated) deposition of burials. Most of them have an increased presence of ritual or symbolic objects, indications of long-distance exchange of exotic objects and raw materials, and some evidence for semi-specialised crafts – all infrequent at most contemporary sites. The very limited household archaeology from these sites indicates that, contrary to what would be expected in chiefdoms, variability in domestic facilities is generally absent. Ritually-oriented collective facilities are also absent. In segmentary societies, the emergence of high population densities is normally a result of the capacity of certain lineages to attract other groups, through persuasive action (warfare, inclusive ritual and feasting have been typical: Beck 2003), and/or the acquisition of reciprocal obligations. In areas where chiefdoms were a generalised phenomenon, as in late Mississippian North America (Blitz 1999), the possibility of resisting a particular chief’s power would have been generally limited to accepting that of another similar potentate. In late prehistoric Iberia, the intra-regional variation of settlement patterns suggests that, although having an increasingly intensified economy, none of the biggest sites developed long-term hierarchical regional polities. If the internal history of some of the biggest villages is illustrative of a general phenomenon, stable social hierarchies did not develop even at the peak of segmentary aggregation (Zafra et al. 1999). On the contrary, and as probably happened in Los Millares (Gilman 2001, 81), internal lineage competition had the leading role in the evolution and devolution of the existing societies. Driven by lineage factionalism and based on the contradiction between communal structures and household interests, all third millennium BC aggregations ended in fission. It may be, although it is not evident, that certain lineage chiefs were capable of controlling and accumulating enough surplus to finance their ambitions in the short term, but eventually they were unable to maintain their labour force, generally the main limitation for the consolidation of elites (Stein 1994, 41). Comparatively, it is difficult to observe qualitative differences in the economic base of all these regional processes. Where aggregations were bigger, enclosures were more monumental; refuse, prestige items and ritual objects more abundant; and competition between lineages, tougher. The ‘cycling’ of aggregational patterns that can be tracked throughout the third millennium cal BC highlights the failure of expanding corporate strategies to overcome the restrictions created by kin groups and lineage factionalism. Copper Age Iberia shows an evident variability in political dynamics, resulting in multiple aggregation and fission processes characteristic of segmentary societies. This variability makes Iberia a relevant case in the study of prehistoric political economy. Acknowledgements This paper could not have been written without the collaboration of Susana Consuegra Rodriguez and Alfonso Vigil Escalera. I am especially grateful to Tim Earle and Antonio Gilman for their critical review, comments and editing. Departamento de Prehistoria Instituto de Historia, CSIC C/ Serrano 13, 28001-Madrid, Spain E-mail: diazdelrio@ih.csic.es OXFORD JOURNAL OF ARCHAEOLOGY © Blackwell Publishing Ltd. 2004 119 COPPER AGE DITCHED ENCLOSURES IN CENTRAL IBERIA references andersen, N.H. 1997: Sarup Vol. 1. The Sarup enclosures (Moesgaard, Jutland Archaeological Society Publications, 33(1). beck, R.A. Jr. 2003: Consolidation and Hierarchy: Chiefdom variability in the Mississippian Southeast. American Antiquity 68(4), 641–61. blitz, J.H. 1999: Mississippian Chiefdoms and the Fission-Fusion Process. American Antiquity 64(4), 577–92. chapman, R.W. 1990: Emerging complexity: the later prehistory of south-east Spain, Iberia and the west Mediterranean (Cambridge). darvill, T. and THOMAS, J. 2001: Neolithic enclosures in Atlantic northwest Europe: some recent trends. In Darvill, T. and Thomas, J. (eds.), Neolithic Enclosures in Atlantic Northwest Europe (Oxford), 1–23. delibes, G. 2001: Del Bronce al Hierro en el valle medio del Duero: una valoración del límite Cogotas I-Soto de Medinilla a partir de las manifestaciones de culto. Zephyrus 53–54 (2000–2001), 293–309. delibes, G., herrán, J.I., santiago, J. de and val, J. del 1995: Evidence for social complexity in the Copper Age of the Northern Meseta. In Lillios, K.T. (ed.), The origins of complex societies in late prehistoric Iberia (Ann Arbor, International Monographs in Prehistory, 8), 44–63. delibes, G., viñé, A. and salvador, M. 1998: Santioste, una factoría salinera de los inicios de la Edad del Bronce en Otero de Sariegos (Zamora). In Delibes, G. (coord.), Minerales y metales en la Prehistoria reciente. Algunos testimonios de su explotación y laboreo en la Península Ibérica (Valladolid, Studia Archaeologica, 88), 155–97. díaz-del-río, P. 2001: La formación del paisaje agrario. Madrid en el III y II milenios BC (Madrid, Comunidad de Madrid, Arqueología, Paleontología y Etnografía, 9). díaz-del-río, P. 2003: Recintos de fosos del III milenio AC en la Meseta Peninsular. Trabajos de Prehistoria 60(2), 61–78. earle, T.K. 1991: Property rights and the evolution of chiefdoms. In Earle, T.K. (ed.), Chiefdoms: Power, Economy, and Ideology (Cambridge), 71–99. gilman, A. 1981: The development of social stratification in Bronze Age Europe. Current Anthropology 22, 1–24. gilman, A. 2000: El desarrollo reciente de la Arqueología Peninsular visto desde los Estados Unidos. Actas do 3° Congresso de Arqueologia Peninsular, 1 (Porto), 27–34. gilman, A. 2001: Assessing Political Development in Copper and Bronze Age Southeast Spain. In Haas, J. (ed.), From Leaders to Rulers (New York), 59–81. gonçalves, V. 1982: Cerro do Castelo de Santa Justa. Un povoada calcolítico fortificado no Alto Algarve oriental. Arqueologia 6, 42–8. harrison, R.J. and orozco, T. 2001: Beyond Characterisation. Polished Stone Exchange in the Western Mediterranean 5500–2000 BC. Oxford Journal of Archaeology 20(2), 107–27. hurtado, V. 1997: The dynamics of the occupation of the middle basin of the river Guadiana between the fourth and second millennia BC. In Díaz Andreu, M. and Keay, S. (eds.), The Archaeology of Iberia. The Dynamics of Change (London), 98–127. kunst, M. and rojo, M. 1999: El valle de Ambrona: un ejemplo de la primera colonización Neolítica de las tierras del Interior Peninsular. Saguntum Extra 2, 259–70. lago, M., duarte, C., valera, A., albergaria, J., alameida, F. and carvalho, A.F. 1998: Povoado dos Perdigoes (Reguengos de Monsaraz): dados preliminares dos trabalhos realizados em 1997. Revista Portuguesa de Arqueologia 1(1), 45–152. OXFORD JOURNAL OF ARCHAEOLOGY 120 © Blackwell Publishing Ltd. 2004 PEDRO DÍAZ-DEL-RÍO martín, M., macías, R., lópez, P., López, J.A., consoegra, S. and díaz-del-río, P. 2002: El paisaje durante el Subboreal en el entorno del yacimiento de Gózquez (San Martín de la Vega, Madrid). XIII Simposio de la Arociacion de Palinólogos enlengua Espan̄ola, 321–34. martínez navarrete, Ma.I. 1998: The Development of Spanish Archaeology in the 20th Century. Archaeologia Polona 335–336 (1997–1998), 319–42. olmo, J. del 1999: Arqueología aérea en Castilla y León. Revista de Arqueología 215, 44–9. renfrew, C. 1973: Monuments, mobilisation and social organisation in Neolithic Wessex. In Renfrew, C. (ed.), The explanation of cultural change (London), 539–58. rovira, S. 2002: Metallurgy and Society in Prehistoric Spain. In Ottaway, B.S. and Wager, E.C. (eds.), Metals and Society. Papers from the session held at the European Association of Archaeologists Sixth Annual Meeting in Lisbon 2000 (Oxford, BAR Int. Ser. 1061), 5–20. stein, G. 1994: Economy, Ritual and Power in ‘Ubaid Mesopotamia’. In Stein, G. and Rothman, M.S. (eds.), Chiefdoms and Early States in the Near East: The Organizational Dynamics of Complexity (Monographs in World Archaeology, 18), 35–46. zafra, N., hornos, F. and castro, M. 1999: Una macro-aldea en el origen del modo de vida campesino: Marroquíes Bajos (Jaén) c. 2500–2000 cal ANE. Trabajos de Prehistoria 56(1), 77–102. OXFORD JOURNAL OF ARCHAEOLOGY © Blackwell Publishing Ltd. 2004 121