The study of nomads from pre-ethnographic times is difficult. Traditionally seen through a lens of contemporary texts usually reflecting some degree of animosity, and always reflecting a naive depiction of the “other,” scholars attempting to explore the nature of nomadic pastoralism in ancient times have faced both a scarcity of sources and basic limitations within those sources. Even when we temper the biases and lacunae in these sources using modern historical and anthropological critique, we are left with tremendous gaps in our direct basic knowledge of ancient pastoral societies.
While not a panacea to these problems of sources, archaeology provides an old-new avenue for exploring ancient pastoral nomadism. Even given the obvious limitations of the archaeological record (preservation, precision in dating, telescoping of events, selective coverage, etc.), the mere fact that we can and do find ancient nomadic camps, often in large numbers, allows us to “do archaeology” on them, generating not only new data sets offering insights to various traditional historical questions, but also, as a matter of course, generating new questions.
In particular, beyond the enriched data set that archaeology can bring to the study of ancient nomadism, the archaeological record studied as an integrated sequence provides a deeper time perspective on the development of the phenomenon. It contrasts with the textual record in two ways. First, the earliest periods in the development of specialized pastoralism, the Neolithic and Chalcolithic, are beyond the chronological range of writing — archaeology is the only means for examining the rise of the phenomenon. Second, even within the span of time for which texts have documented pastoral societies, for obvious reasons, the earlier the period, the fewer the historical texts. Although the archaeological record of pastoral sites in the settled zone has suffered significantly from the destruction wrought by development, beginning in ancient times (and including the practice of agriculture), and in such cases the archaeological record is patchy at best, the record in the desert is relatively even. We can trace the archaeology of desert pastoralism back to its origins and the different periods are roughly comparable in terms of the representativeness of the record. This facilitates comparison between periods and cultures, providing the means for a longue duree approach to the analysis.
Focusing on the Negev as a case study, shifting patterns of relations between nomadic and sedentary groups can be traced over the long term, beginning with the origins of the phenomenon in the Late Neolithic and extending forward through classical and indeed recent times. These are reflected in fluctuating demographic patterns, changes in desert subsistence systems,
STEVEN A. ROSEN
Changing patterns of trade relations, shifting settlement systems (including movement of the border between desert and sown), and evolving material culture systems. Archaeologically I suggest that the long span in the Negev may be divided into four basic complexes, the earliest Timnian complex, the early historical complex (second millennium and early first millennium B. C.), the classical complex (beginning with the nabateans and extending through the Early islamic period), and the recent Bedouin. each complex shows a specific package of social, economic, technological, and ecological adaptations.
The negev as a case study: settlement patterns
Archaeological investigations in the Negev highlands in the early 1980s, initiated as a large salvage project before the redeployment of the israeli army out of the sinai Peninsula, resulted in the systematic documentation of thousands of previously unknown sites spanning the entire breadth of the archaeological record (fig. 4.1). Numerous excavations were also conducted. The surveys were published (and are still being published) in a series of standardized monographs each covering 100 sq. km (for the central Negev, see Avni 1991; Baumgarten 2004; Cohen 1981, 1985; Haiman 1986, 1991, 1993, 1999; Lender 1990; Rosen 1994; and in ecological and cultural contrast, for the northern Negev, see Beit-Arieh 2003a; Gazit 1996; Govrin 1991). Excavations were published in various forums and formats. Other research conducted in adjacent areas like the southern Negev (e. g., Avner 1990, 1998; Rothenberg 1972c), sinai (e. g., Amiran, Beit-Arieh, and Glass 1973; Bar-Yosef 1984; Beit-Arieh 2003b; Dahari 2000; Rothenberg 1972a-b), and southern Jordan (Henry 1992, 1995) complements the picture from the central Negev and indeed is often the key to understanding the phenomena reflected in the record of the central Negev.
Beyond the specifics of the different periods, several general features of this record are integral to the themes of this paper. The surveys in the central Negev present a consistent picture of major demographic peaks and declines (Rosen 1987b) and seem to contrast with more stable sequences in the southern Negev (e. g., Avner 1998; Avner, Carmi, and segal 1994). The northern Negev shows periods of rise and decline, but these are not necessarily coincident with those of the central Negev and seem to reflect different social and historical trends or events. To a major extent, these demographic fluctuations seem to reflect shifts in the edge of systematic agricultural settlement, coincident in some cases with shifting pastoral territories. Figure 4.2 summarizes roughly the north-south shifts of the agricultural settlement over time, as reflected in the presence of evidence for agriculture and associated sites during a particular time period on a latitudinal scale. The types of evidence vary with different periods, as agricultural practices change. For the earlier periods (Bronze Age and earlier), the presence of high proportions of flint sickle blades in a lithic assemblage is a good proxy for the intensity of agricultural practices (Rosen 1997: 58), and the clear contrasts between desert and Mediterranean zone lithic assemblages in the proportions of sickles must surely be a reflection of intensity of practice.1 For later periods, the presence of agricultural terracing, direct evidence for run-off irrigation systems, is a clear reflection of systematic farming. Although
Microenvironments appropriate for agriculture, and given the environmental limitations of the desert for agriculture, relative proportions of sickles indeed seem to reflect intensity of practice.
The dating of these terrace systems has been the subject of some debate (e. g., Haiman 1990, 1995a), and certainly they have been reused by recent Bedouin, the presence of homesteads in association with the dam systems is a strong indicator of agricultural settlement. In classical and recent times, texts and ethnography also inform on agricultural practices (e. g., Bailey 1980; Bruins 1986; Evenari, shanan, and Tadmor 1982; Kraemer 1958; Marx 1967; Mayerson 1960). Figure 4.2 provides a simplified north-south perspective on agricultural penetration; the real distribution of agricultural21 practices obviously does not follow east-west latitudinal lines, but rather the natural geography, and enclaves of farming may exist in otherwise nonagricultural regions. The true pattern is less even and perhaps even somewhat patchy. Figure 4.3 presents maps suggesting the edge of agricultural practices in two different periods, the Byzantine period and the Early Bronze Age II, and figure 4.4 presents an impressionistic perspective on the actual edges of systematic agricultural exploitation in different periods, with an emphasis on “impressionistic.” Thus, figure 4.2 is intended to summarize the fluctuations and should be taken as an indicator of depth of penetration rather than a detailed geographic statement. As an aside, it is quite important to note that these shifts in the agricultural zone do not correspond in any one-to-one sense with climatic fluctuations in some absolute sense of the word. It is clear that the Early Bronze Age (and for that matter, the Chalcolithic period) were times of considerably greater rainfall than the Byzantine period (e. g., A. Rosen 2007: 150-71 and references; also Bar-Matthews, Ayalon, and Kaufman 1998), yet with considerably less agricultural penetration into the desert.
Complementing this picture of agricultural expansion and contraction, we can also construct similar schematics of fluctuating pastoral occupation in the central Negev. However, in contrast to agricultural settlements, defining pastoral presence requires some discussion, both in terms of what exactly we mean by presence in referring to mobile peoples, and what types of evidence can be used to infer that presence.
For the purposes of this analysis, defining pastoral presence perhaps is more easily accomplished by stating what it is not: it is not the presence of a shepherd opportunistically grazing in an area, nor is it the occasional (non-cyclical) penetration of a group into or through a region. Furthermore, it is clear that desert pastoralists commonly penetrate the agricultural zone, either in the course of seasonal migration or during seasonal dispersal. Thus by pastoral presence I intend the presence of primary residential camps in what Binford (1980) has termed a system of residential mobility. Although Binford was referring to hunting-gathering groups and contrasted residential mobility (the movement of the entire residential social unit) with logistical mobility (movement of individuals or subgroups to and from base camps to secondary resource exploitation sites), the key here is the distinction between residential camps and other sites. Thus, pastoral presence is the common use of a territory by families and clans as expressed in the presence of residential base camps. I refer to this kind of presence from here on as pastoral tribal presence, to distinguish it from sedentary occupations and from opportunistic exploitation. In the classical era, there is clear overlap between the agricultural and pastoral zones (Avni 1996; Haiman 1995a; Rosen 1987a; Rosen and Avni 1993), but the primary pastoral tribal presence, in this case in the sense of the largest aggregate encampments, is in the areas south of the Ramon crater.
STEVEN A. ROSEN
Furthermore, it should be noted in particular that the recent discovery of rock shelters with dung layers attributable to periods for which there are no campsites in the region (Rosen et al. 2005) does not constitute tribal presence, but obviously reflects the use of the central Negev for seasonal grazing. The near total absence of material culture in these rock shelters is typical for such types of exploitation since the range of activities is severely limited.
The second issue is that of archaeological visibility. The central negev is rich in archaeological sites not associated with agricultural systems, interpreted universally as the remains of residential camps reflecting pastoral systems (e. g., cohen 1999; cohen and Dever 1978, 1979, 1980; Haiman 1995a-b; Rosen 1987a; Rosen and Avni 1997).22 On the other hand, debate has arisen around those periods for which there is little evidence of residential camps of any kind, pastoral or sedentary, for example, in the second millennium B. C. In particular, Finkelstein (1995; Finkelstein and Perevoletsky 1990) has argued that these periods (in the Negev) represent times of increased mobility, with correspondingly less hard archaeological evidence. It is beyond the scope of this paper to deal in depth with this issue (see Rosen 1992), but the arguments for archaeological invisibility are flawed for the earlier periods (at least as late as the end of the third millennium B. C.), when lithic industries which leave large quantities of waste provide a clear marker for occupation sites (in much the same way as for the Paleolithic), and for later periods, when ceramics are common. Furthermore, given the likelihood that tents are a later phenomenon among nomadic groups (Rosen and Saidel in press), architectural remains should also provide markers. Finally, it is also worth noting that tents too leave their archaeological signature (for ethnoarchaeological examples, see Avni 1992; Banning and Kohler-Rollefson 1986, 1992; Eldar, Nir, and Nahlieli 1992; Saidel 2001; Simms 1988; and for archaeological examples, see Banning 1986; Rosen 1993; Rosen and Avni 1997: 19, 44, 54-55, 59).
Given these provisos, figure 4.5 summarizes the north-south extent of pastoral tribal presence in the Negev in different periods. As with the schematic of agricultural penetration (fig. 4.2), the long-term record shows shifting territories, with some periods showing penetration fairly far north, and others a contraction into the southern Negev and Sinai. Significantly, while there appears to be partial correlation between the agricultural and pastoral systems, such that during some periods the two systems seem geographically linked with their borders more or less coinciding and even overlapping, there are other long periods when there appears to be significant distance between the two systems, in essence leaving a blank space in the middle, in the heartland of the central Negev. These blank periods coincide, of course, with those periods of demographic decline mentioned earlier. In particular, this phenomenon can be noted during the second millennium B. C. and during the Middle Ages, in the first half of the second millennium A. D. For comparative purposes, figure 4.3, showing the settlement systems of the Early Bronze Age (ca. 3000 B. C.) (ignoring for the moment the Early Bronze Age IV, at the end of the third millennium B. C.) and the classical era can be compared to figures 4.6 and 4.7, presenting impressionistic maps of the tribal pastoral system/agricultural systems of the Middle Ages and of the Middle Bronze Age, circa 1800 B. C.
The spatial and chronological variability reflected in these figures has its source in social dynamics, and I would suggest especially in the nature of relations between the sedentary societies of the Mediterranean (and steppe) zone, and the nomadic societies of the desert.
Early Bronze sites of south Sinai [Beit-Arieh 1986]), but everyone agrees that other sites are indeed associated with local pastoral cultures.
Periods of proximity between the systems reflect close relations and indeed, some degree of economic and perhaps cultural integration. Periods of spatial separation reflect different levels of integration, indeed different kinds of relations, at least in the regions surveyed. These spatial patterns and fluctuations combine well with analyses of material culture and texts, suggesting that the long span of pastoral adaptations, from origins to modern times in the Negev, may be divided into four general complexes. Explanation of the patterns of contraction and expansion requires explication of each individual complex in its historical and cultural context.
The timnian complex
The Timnian complex (Eddy and Wendorf 1998; Henry 1992, 1995: 353-74; Kozloff 1972/73, 1981; Rothenberg and Glass 1992; also see Ronen 1970 for lithic descriptions without direct attribution; Rosen in press for recent update; and Zarins 1990, 1992 for counterpart in Arabia), the earliest pastoral complex in the desert, developed in the sixth millennium B. C. as herding replaced hunting, and as social and cultural institutions evolved concomitant to this new subsistence base. In particular the following features are characteristic of the complex, and contrast with the preceding Early Neolithic (Pre-Pottery Neolithic B) cultures: (1) Enclosure and attached-room architecture (contrasting with the clustered rooms of the Early Neolithic); (2) Increasing site sizes such that the larger sites consist of aggregates of several enclosures and attached rooms; (3) The prevalence of desert kites (gazelle hunting drive traps); (4) The presence of desert shrines with cosmological symbolism, sometimes achieving megalithic proportions; (5) The presence of constructed burial fields (tumuli, nawamis) often in association with the shrines; and (6) A material culture assemblage dominated by the chipped stone industry, including small and transverse stone arrowheads, microlithic drills, tabular scrapers, ad hoc blade tools, and a wide range of ad hoc flake tools. Beads made of seashells and other materials are also common. Ceramics appear late in the sequence and are dominated by globular hole-mouth cooking pots.
In general, the package seems to reflect the evolution of a tribal society, contrasting with earlier smaller-scale social formulations. Thus, desert kites reflect cooperative hunting, probably requiring a larger community than earlier hunting strategies, as well as investment in hunting furniture on a scale not seen previously.23 The shrines, especially in their megalithic aspect (Rosen et al. 2007) indicate labor organization of a degree previously not seen in the desert. The presence of burial structures organized in extensive fields and standing upright and visible from some distance most likely reflects a higher order of territoriality, probably a response to the needs of herding. Indeed, the presence of large shrines (with cosmological alignments) and large tombs strongly suggests a social organization capable of drafting labor for construction and the development of an ideological system for legitimizing these new power relations. Geographic variability in site size and material culture suggests seasonal aggregation and dispersion (e. g., Haiman 1992; Kozloff 1981; also see Henry 1992 for an example from Jordan).
A Pottery Neolithic/Chalcolithic date (= Timnian) and A. Goren (cited in Rosen 1997: 39) claims to have recovered transverse arrowheads, an attribute of the Timnian, in his excavation of one kite.
STEVEN A. ROSEN
The Timnian as an archaeological complex spans more than three millennia, from circa 5500 B. C. to circa 2000 B. C. It can be divided into four phases (Early, Middle, Late, and Terminal) distinguishable on the basis of material culture variability, especially arrowhead typology in the early stages and ceramics in the latter. Other chronological markers include tabular scrapers and associated attributes, other lithic types, burial styles, the introduction of metallurgy, and the introduction of ceramics, and for the later stages, ceramic variability.
For the purposes of this paper, the important issues are the changes in external relations between Timnian pastoral groups and the sedentary farming societies of the mediterranean zone. The increasing ties between the two culture regions can be traced from relative autonomy in the early Phase of the Timnian sequence to increasing ties (based especially on metallurgy) in the middle Phase, and on to intensive ties leading to economic asymmetries in the Late and Terminal Phases. specifically, the early Phase shows only minor connections between the desert and the mediterranean zone, reflected primarily in shell bead exchange (Bar-Yosef mayer 1997) and perhaps the diffusion of arrowhead paradigms (the direction of which is not always clear) (e. g., Gopher 1989, 1994). economically, Timnian groups in this period are basically autonomous. In the middle Phase copper moved regularly, if not in truly large quantities, from feinan, and perhaps Timna, into the Beersheva Basin. Two points are crucial for understanding why the Timnian should still be seen as a relatively autonomous economy in this period. first, with the exception of the copper, objects originating in the desert are very rare in sites in the agricultural chalcolithic Beersheva-ghassul culture of the northern Negev. second, in parallel, there is virtually no movement of goods south, and material culture in the desert is virtually all local in origin. essentially, the early and middle Phases of the Timnian reflect a subsistence pastoral economy, one which I have referred to elsewhere as herding-gathering (Rosen 2002).
In distinct contrast, by the Late Timnian (early Bronze Age II in northern terms), both the diversity and the quantity of goods moving between north and south increased dramatically. Thus, the material culture assemblage from arad, the early Bronze age desert gateway town in the northern negev, includes the following desert derived goods: copper, milling stones, shell beads, other beads, tabular scrapers, and pottery. Pottery and milling stones are a new addition to the exchange system, and all other goods significantly increased quantitatively. good evidence for production for export of milling stones and beads at the level of the cottage industry has been documented in the central negev (abadi and rosen 2008; rosen 2003). Trade in copper is also evident (for central negev, segal and rosen 2005; for other regions, e. g., Beit-arieh 1974; rothenberg 1972c; rothenberg and glass 1992). import of ceramics into the desert has also been demonstrated, and the presence of trade stations in south sinai is well established (amiran, Beit-arieh, and glass 1973; Beit-arieh 2003b; Porat 1989). This change in the basic tenor of relations is also reflected in a major increase in the number of sites in this period (fig. 4.8). The combination of increasing intensity of exchange, increase in number of desert sites (undoubtedly reflecting a demographic expansion), and the apparent collapse of this trade system with the abandonment of arad suggests that relations by the Late Timnian can be characterized as economically asymmetric, such that herding-gathering was no longer a viable subsistence system and trade had now become crucial to the desert polities. This economic asymmetry is one of the defining features of recent pastoral nomadic systems (Khazanov 1984). When the trade system collapsed, the large population was no longer viable and the system was abandoned.
The desert early Bronze age iv (= intermediate Bronze age = middle Bronze age i) sites, dated to the late third millennium B. C., have been attributed to the Terminal Timnian on the basis of ceramic and lithic technological continuities with the preceding phase (e. g., Rosen et al. 2006; vardi 2005). Notably, the Early Bronze Age iv corresponds to the pan-Near Eastern collapse of urban society, but reflects a settlement peak in the arid region.
Despite these material culture continuities, the central negev shows a culture stratigraphic break between the phases, associated with the early Bronze age iii and the abandonment of arad.24 Both the settlement patterns and basic domestic architecture in the Terminal Timnian differ from the preceding Late Timnian and some of the sites achieve sizes not previously seen in the region (e. g., ein ziq has some 200 domestic structures) (see especially Cohen 1999: 83-298 for summary). it is beyond the scope of this paper to review these materials in detail, but site distributions (Cohen 1992, 1999: 267-82; Haiman 1996), ceramic petrography (Goren 1996), and the presence of copper ingots originating in feinan (with a possibility of origin at Timna) at several sites (Segal and Roman 1999) suggest close connections with areas east of the Rift Valley. That is, the Early Bronze Age iV “recolonization” may derive from Jordan, and continuities with the Timnian are through the east. The point to be stressed here is that despite the size of the sites there is no evidence for significant agriculture, nor would the desert environment of the period have permitted it. instead, this pastoral society seems to have had its raison d’etre in the desert in the exchange systems with the north and perhaps with Egypt (e. g., Yekutieli 1998), with a primary focus on copper.
World History
