17
Research Khirigsuurs, ritual and mobility in the Bronze Age of Mongolia Francis Allard 1 & Diimaajav Erdenebaatar 2 The khirigsuurs are large and complex ritual sites that are major features in the landscape of Bronze Age Mongolia and represent considerable investment. The authors present recently investigated examples of this important class of monument, describe their attributes and offer preliminary deductions of the kind of society they imply – and whether it was truly nomadic. Keywords: Bronze Age, Mongolia, nomadism, ritual monuments Introduction Spanning thousands of years, the broad emergence of mobile pastoralism across the vast Eurasian steppes encompassed the domestication, harnessing and consumption of the horse, the adoption of wheeled vehicles, animal sacrifice and burial, as well as the practice of inhumation under tumuli or kurgans (Khazanov 1994: 90-7; Anthony 1998; Levine 1999; Anthony & Brown 2000; Kuzmina 2000, 2003). Although marked by regional variation and still disputed on points of ultimate causes and chronology, this process is generally recognized as having been preceded by a settled and agricultural mode of life. Archaeological surveys and excavations have helped reveal the gradual expansion of forms of mobile pastoralism across Eurasia. By the early first millennium BC, much of the steppes appears to have been occupied by nomadic herding societies whose armed horsemen and migratory populations helped spread technological and stylistic innovations over large distances. Explanations for the transition to greater residential mobility have included increasingly arid conditions, population pressure from settled groups, and the demand for the products of a specialised pastoralist economy (Cribb 1991: 12-4; Khazanov 1994: 85-90). In its focus on the origin and transmission of cultural and economic features, the sweeping perspective favoured by studies of Eurasian steppe prehistory has until recently been inimical to the more detailed charting of developments at the regional level. Furthermore, although cross-regional studies often include discussions of developments in southern Siberia’s Minusinsk Basin, Tuva and Gorno-Altai regions, they typically do not extend further east to take into account Mongolia’s vast grasslands, an omission partly justified by the scarcity of accessible published research. Recent work in Mongolia by local and foreign archaeologists has helped bring into sharper focus the broad outline of its prehistory, revealing in the process the existence of regional trajectories within the territory itself. Neolithic sites in eastern and southern Mongolia have yielded ceramics as well as evidence 1 Department of Anthropology, Indiana University of Pennsylvania, USA. (Email: [email protected]) 2 Institute of History, Mongolian Academy of Sciences, Mongolia (Email: [email protected]) Received: 25 July 2003; Accepted: 2 February 2004; Revised: 14 June 2004 antiquity 79 (2005): 547–563 547

Allard F Et Al 2005 (Mongolia - Bronze Age Rituals & Mobility)

Embed Size (px)

Citation preview

Page 1: Allard F Et Al 2005 (Mongolia - Bronze Age Rituals & Mobility)

Res

earc

h

Khirigsuurs, ritual and mobility in theBronze Age of MongoliaFrancis Allard1 & Diimaajav Erdenebaatar2

The khirigsuurs are large and complex ritual sites that are major features in the landscapeof Bronze Age Mongolia and represent considerable investment. The authors present recentlyinvestigated examples of this important class of monument, describe their attributes and offerpreliminary deductions of the kind of society they imply – and whether it was truly nomadic.

Keywords: Bronze Age, Mongolia, nomadism, ritual monuments

IntroductionSpanning thousands of years, the broad emergence of mobile pastoralism across the vastEurasian steppes encompassed the domestication, harnessing and consumption of the horse,the adoption of wheeled vehicles, animal sacrifice and burial, as well as the practice ofinhumation under tumuli or kurgans (Khazanov 1994: 90-7; Anthony 1998; Levine 1999;Anthony & Brown 2000; Kuzmina 2000, 2003). Although marked by regional variation andstill disputed on points of ultimate causes and chronology, this process is generally recognizedas having been preceded by a settled and agricultural mode of life. Archaeological surveysand excavations have helped reveal the gradual expansion of forms of mobile pastoralismacross Eurasia. By the early first millennium BC, much of the steppes appears to have beenoccupied by nomadic herding societies whose armed horsemen and migratory populationshelped spread technological and stylistic innovations over large distances. Explanations forthe transition to greater residential mobility have included increasingly arid conditions,population pressure from settled groups, and the demand for the products of a specialisedpastoralist economy (Cribb 1991: 12-4; Khazanov 1994: 85-90).

In its focus on the origin and transmission of cultural and economic features, thesweeping perspective favoured by studies of Eurasian steppe prehistory has until recently beeninimical to the more detailed charting of developments at the regional level. Furthermore,although cross-regional studies often include discussions of developments in southernSiberia’s Minusinsk Basin, Tuva and Gorno-Altai regions, they typically do not extendfurther east to take into account Mongolia’s vast grasslands, an omission partly justifiedby the scarcity of accessible published research. Recent work in Mongolia by local andforeign archaeologists has helped bring into sharper focus the broad outline of its prehistory,revealing in the process the existence of regional trajectories within the territory itself.Neolithic sites in eastern and southern Mongolia have yielded ceramics as well as evidence

1 Department of Anthropology, Indiana University of Pennsylvania, USA. (Email: [email protected])2 Institute of History, Mongolian Academy of Sciences, Mongolia (Email: [email protected])

Received: 25 July 2003; Accepted: 2 February 2004; Revised: 14 June 2004

antiquity 79 (2005): 547–563

547

Page 2: Allard F Et Al 2005 (Mongolia - Bronze Age Rituals & Mobility)

Bronze Age Mongolian khirigsuurs

of sedentary occupation, agriculture and, in the case of the Tamsagbulag culture of easternMongolia, animal domestication (Derevyanko & Dorj 1992: 172-81). In contrast, althoughthe remains of domesticated horse, sheep and cattle have been recovered from third andsecond millennia BC campsites and burials in north and north-west Mongolia, there is as yetno clear indication of agricultural settlements predating the establishment of a fully nomadicpastoralist economy in that area. Possibly, as one archaeologist has suggested, subsistence inthat part of Mongolia at this early time consisted of hunting and a mobile form of cattleherding, with agriculture playing little or no role (Volkov 1995: 320).

What the archaeological record of Mongolia does make abundantly clear is the dramaticcultural transformation of much of the territory by the mid-late second millennium BC.From this time until the mid-first millennium BC, a period that is the focus of this articleand that roughly corresponds to the region’s developed Bronze Age, the landscape is markedby a profusion of stone built sites, graves and stelae that are found in large numbers along thevalley bottoms, hill slopes and hill tops. Visible settlement traces have not been detected inassociation with this Bronze Age landscape, neither within the perimeter of the sites nor inassociation with them, while various lines of evidence point to the ritual and funerary natureof the sites, as well as the possibility that they were built and used by mobile populationswho left only ephemeral traces of their settlements behind.

Monuments of the Mongolian Bronze AgeMongolian archaeologists generally recognise three distinct types of Bronze Age structures,whose spatial distributions overlap with one another (Volkov 1995; Erdenebaatar 2004).The so-called ‘slab burials’ are constructed of large vertical stone slabs whose protrudingsections above the ground surface define the burial’s perimeter. The head of the deceased istypically aligned between north-east and south-east. Grave goods include a variety of bronzes(tools, hunting implements, weapons, ornaments and horse riding bits), stone artefacts(tools and ornaments), ceramics (mostly vessel sherds), bone artefacts, selected bones oflivestock (e.g. skulls, scapula and anklebones), cowries and mother-of-pearl (Ishjamts 1994:151-2; Volkov 1995: 321; Erdenebaatar 2002: 151-203, 239-52; 2004). Together, theseassemblages point to the emerging martial nature of society, the importance of animals inritual, and the existence of long-distance contacts that would have brought exotic materials toMongolia.

With some reaching 2.5m high, ‘deer stones’ are stone stelae whose surfaces displaynumerous carved designs. Often organized into horizontal bands, these designs includecircles (interpreted as a sun symbol or earrings), a few full-figured faces facing east, highlystylised deer with bird-like beaks and backward-flowing antlers, bows and various objects(knives, axes, daggers, swords, battle-picks) hanging from what appear to be belts. Somebelieve that deer stones and their designs represent tattooed humans.

This article focuses on a third type of structure, namely the highly distinctive ‘khirigsuurs’,which generally consist of a central stone mound, a square or circular ‘fence’ of surface stones,as well as small stone mounds and circles. Paths and other structures are also sometimespresent. Although khirigsuurs, slab burials and deer stones have been identified at a numberof single sites (e.g. Erdenebaatar 2002: 207, 226), archaeologists have yet to clarify the

548

Page 3: Allard F Et Al 2005 (Mongolia - Bronze Age Rituals & Mobility)

Res

earc

h

Francis Allard & Diimaajav Erdenebaatar

temporal and functional relationship linking the three types of structures. The grave goodsin the slab burials and the objects depicted on the deer stones at least suggest overlappingchronologies between the mid-second and mid-first millennia BC.

Investigating the khirigsuursEstablished in 2001, the Khanuy Valley Project on Early Nomadic Pastoralism in Mongolia iscentred in the valley of Khanuy River, located to the north of the Khangay mountain rangein Arkhangai aimag (Figure 1). The project aims to better understand the circumstances thatsurrounded the emergence and development of mobile herding as a way of life in centralMongolia, an objective that it is realising through excavations, surveys, the identification ofcampsites, as well as an ethnographic study (Allard et al. forthcoming). The project is alsoexcavating Golmod-2, a recently discovered Xiongnu (third century BC to second centuryAD) cemetery (Allard 2002). Slab burials, deer stones and khirigsuurs are all found inlarge numbers within the project’s 330km2 research area. This article presents the resultsof fieldwork carried out at many of the khirigsuurs in Khanuy valley and summarisesrecently published information on khirigsuurs in other parts of Mongolia and surroundingregions. It provides an opportunity to describe this little known type of monument tothe wider archaeological public, to investigate its possible association with mobile herding,and to consider the role that ritual may have played in the development of the khirigsuurphenomenon.

The khirigsuur Urt BulagynThe focus of ongoing field studies by the project, the site of Urt Bulagyn (KYR1) serves as abaseline from which to discuss Mongolia’s khirigsuurs. Built entirely of unaltered stones andmeasuring 390 × 390m, it is the second largest khirigsuur in the research area (Figures 2and 3). Its most prominent structure, which has yet to be excavated, is its 5m tall centralmound. A rectangular ‘fence’ consisting of a single line of surface stones and four cornermounds enclose the central plaza, while a poorly defined ‘path’ of surface stones joins thecentral mound to the eastern fence. Beyond the fence, more than 1700 satellite moundsare concentrated on the eastern and southern sides, with the largest mounds located inthe sparse western sector. Measuring no more than 1.2m high from their base, the sevenmounds excavated so far have all contained horse remains, but no artefacts. The twoexcavated mounds illustrated in Figure 4 reveal a feature of the site’s southern sector, namelythe presence of smaller mounds placed to the immediate south-west of larger ones. Near thecentre of the largest of these two satellite mounds, a horse’s skull and a series of five cervicalvertebrae were found, oriented to 120◦. The close proximity of the vertebrae to the skull andthe fact that the vertebrae were found one against the other in correct anatomical sequencesuggests that they were still held together by ligaments when placed there. Partial remains ofa mandible oriented to 117◦ were recovered from the smaller mound. Horse skulls and/orcervical vertebrae were found in all five other mounds excavated at the site.

The outermost ring-shaped area (Figure 3) consists of over 1000 circles that are built ofsurface stones and that range in diameter from 1 to 3m. The wide western sector contains

549

Page 4: Allard F Et Al 2005 (Mongolia - Bronze Age Rituals & Mobility)

Bronze Age Mongolian khirigsuurs

Figure 1. Research area, showing the location of Urt Bulagyn and other selected khirigsuurs.

the most circles as well as many of the largest ones. The circles are often arranged in rows thatare roughly parallel to one another and parallel to the side of the stone fence nearest to them.All five stone circles excavated so far have yielded whitened/bluish small bone fragmentslying between 100 and 200mm below the modern ground surface. A total of over 33 000such fragments was recovered from one of the circles, and fewer than 200 from others.

550

Page 5: Allard F Et Al 2005 (Mongolia - Bronze Age Rituals & Mobility)

Res

earc

h

Francis Allard & Diimaajav Erdenebaatar

Figure 2. The khirigsuur Urt Bulagyn.

The colour and appearance of the fragments indicate a long period of burning, while theirdispersed distribution and the presence of few charcoal remains suggest that the bones werecremated at another location and the fragments later scattered freely within the circle area.A cursory analysis of a sample of fragments suggests that they represent various livestockspecies.

Fourteen slab burials, most of them oriented toward the north-east, are located beyondthe northern fence between the satellite mound and stone circle sectors (Figure 3). Fourburials were excavated. No artefacts, human or animal remains were recovered from two ofthe graves, while a third yielded only a few animal bones. Measuring 5.7 × 5.0m, the largestburial at the site contained partial human remains, along with a few bones of livestockanimals. Located in the same area as the burials, a ‘northern path’ of unknown functionappears as a well-defined scatter of surface stones. A partial excavation of the path revealeda number of large stones lying flat. The slab-burials lie about 15◦ from the west-east axis.

Along with helping to reveal its ritual function, work at Urt Bulagyn is providing insightsinto the construction process itself. For example, it is estimated that roughly half a millionstones were used to build the monument. Obtained from horse teeth found in two differentsatellite mounds, the radiocarbon dates available so far (1040-850 BC and 975-680 BC;these and later dates given at a 2σ range) suggest that the satellite mound sector may have

551

Page 6: Allard F Et Al 2005 (Mongolia - Bronze Age Rituals & Mobility)

Bronze Age Mongolian khirigsuurs

Figure 3. The khirigsuur Urt Bulagyn.

expanded outward over a period of a few hundred years. However, many more dates areneeded to test this hypothesis and clarify whether Urt Bulagyn was built in phases.

The khirigsuur phenomenon: chronology, variability and functionResearch in Mongolia and Russia is helping to clarify the chronology and structural featuresof the khirigsuur phenomenon (Askarov et al. 1992: 466-7; Volkov 1995: 324; Erdenebaatar2002: 126-47; 2004). The khirigsuur type of kurgan characterised by a stone mound anda circular or four-sided enclosure is found throughout the western two-thirds of Mongolia,and in regions to its northwest (Tuva and Gorno-Altai) and north (east and west of LakeBaikal). Built of stone slabs, the cists within the central mounds rarely contain grave goods orinhumations, leading to a growing acknowledgement that many khirigsuurs probably had anon-funerary function. Although some khirigsuurs were undoubtedly looted, the apparent

552

Page 7: Allard F Et Al 2005 (Mongolia - Bronze Age Rituals & Mobility)

Res

earc

h

Francis Allard & Diimaajav Erdenebaatar

Figure 4. Two excavated satellite mounds at Urt Bulagyn.

structural integrity of some of the mounds and central cists suggests they were undisturbed.When present, interred individuals are usually positioned with their head pointing to thewest or north-west. Aside from the previously mentioned radiocarbon dates for Urt Bulagyn,the project has dated one small khirigsuur (KYR57) in our research area to 1390-910 BC anda much larger one (KYR 119, Figure 9) to 930-785 BC. When complemented by typologicalanalysis, these dates and the few others available for Mongolia suggest that khirigsuurs werebuilt between the mid-late second millennium BC and the seventh century BC.

553

Page 8: Allard F Et Al 2005 (Mongolia - Bronze Age Rituals & Mobility)

Bronze Age Mongolian khirigsuurs

The results of fieldwork conducted by our team at 27 khirigsuurs in Khanuy valley and byothers in other parts of Mongolia permit us to address the issue of structural and behaviouralvariability more systematically than has been attempted until now. A comparison of overallsite plans reveals an impressive number of noticeable and less evident features present atmost of the research area’s khirigsuurs, six of which are illustrated in Figure 5. Althoughcertain elements are missing at some of the sites (especially the smaller ones), the structuralvocabulary is generally consistent with that identified at Urt Bulagyn: a central mound; afence with corner mounds; a path joining the central mound to the eastern fence; a satellitemound area whose western sector contains few mounds or may be absent altogether; fourlarge satellite mounds placed next to the eastern fence; an association, in the southern sector,between large and smaller satellite mounds arranged along a roughly south-west-north-eastaxis; a concentric outer area of stone circles arranged in rows; and the presence of a longpath and slab burials between the stone circle and satellite mound areas in the northernsector of the sites.

Interestingly, the smallest khirigsuurs also tend to be the most structurally variable, asillustrated by the unusually shaped site KYR10, which lacks a number of elements (Figure 5).Other small khirigsuurs lack a fence altogether. Without additional chronological data, itis difficult to determine whether this is evidence of more relaxed rules of construction atsmaller sites, of the early stages of rule systematisation, or of the lessening of such rules bythe end of the period. Significantly, the few detailed plans published to date of khirigsuurslocated in northern Mongolia permit us to note a similar structural plan to that describedabove, although circular fences may have been more common outside Khanuy valley.

One unexpected finding, illustrated in Figure 6, reveals that the eastern fence is often10-15 per cent longer than the western fence. We suggest two possibilities to account forthis result. First, the fences may have been carefully measured at the time of construction(since such a small difference is difficult to estimate visually), resulting in a shape that wasin some way meaningful. Second, we may be witnessing the ‘keystone effect’, in which thebuilders of the sites faced east and inadvertently made the side furthest to them (i.e. the eastfence) longer than the one closest to them (i.e. the western fence). Here again, the smallestsites (in black on Figure 6) display the greatest variability (e.g. KYR 10 in Figure 5), withvariation in fence length ratios significantly reduced once these small sites are removed fromthe analysis.

As illustrated in Figures 7 and 8 (top), the khirigsuurs in the research area display arelatively consistent orientation along a west/north-west-east/south-east axis, a narrowrange within which other Mongolian khirigsuurs for which a site plan is available also fall. Adetermination of the sites’ intended orientation (i.e. toward west/north-west or east/south-east) is hampered by the presence of prominent elements on both sides of the khirigsuurs(e.g. the path on the east side and the ‘opening’ in the satellite mound sector on the westside), although the orientation of the few recorded inhumations toward the west or north-west may be an important clue. At the very least, such consistency suggests an orientationtoward a feature or event in the sky rather than toward a single marker in the landscape.If this event was the setting or rising of a particular celestial body above the horizon at acertain time of the year, the determination of the body in question will need to await furtherfield observations since mountain ranges on both sides of the valley would have interrupted

554

Page 9: Allard F Et Al 2005 (Mongolia - Bronze Age Rituals & Mobility)

Res

earc

h

Francis Allard & Diimaajav Erdenebaatar

Figure 5. Selected khirigsuurs in the research area.

555

Page 10: Allard F Et Al 2005 (Mongolia - Bronze Age Rituals & Mobility)

Bronze Age Mongolian khirigsuurs

Figure 6. Khirigsuur fence length statistics.

its path as it rose or set. One possibility being considered is that the sites were orientedtoward the setting of the sun or Venus in the western sky during the spring–summermonths or toward their rise in the eastern sky during the autumn–winter months.

Excavations of satellite mounds at khirigsuurs located in and outside the research areahave revealed further consistency in the orientation of horse heads, most of which pointto the east or south-east (Figure 8, bottom). As in the case of site orientation, more workis needed to determine the celestial feature used to orient the horse skulls. However, theorientation and skeletal data available so far on the age of the horse at the time of deathdoes point to an intriguing possibility, namely that the horses were sacrificed during thelate autumn months as offerings to the rising sun. It is worth noting that in present-dayKhanuy valley, the killing of horses for meat is also done at this time of the year, when thefat content is the greatest.

Also significant is the extent of variation in the treatment of horse remains among, andwithin, khirigsuurs. Some of the excavated satellite mounds in the research area containedonly scattered fragments of horse teeth, while a few yielded nothing at all. At Urt Bulagynitself, one mound contained cervical vertebrae, front teeth, but no other skull elements,

556

Page 11: Allard F Et Al 2005 (Mongolia - Bronze Age Rituals & Mobility)

Res

earc

h

Francis Allard & Diimaajav Erdenebaatar

Figure 7. Full coverage survey area, showing the location and orientation of khirigsuurs.

557

Page 12: Allard F Et Al 2005 (Mongolia - Bronze Age Rituals & Mobility)

Bronze Age Mongolian khirigsuurs

Figure 8. Site and horse head orientations.

while in another, the vertebrae were linedup along the skull’s northern side, instead oftheir more usual position south of it. At onekhirigsuur in northern Mongolia, some of the13 mounds contained various combinations ofhead, neck and hoof remains, while horse partswere accompanied by ceramic sherds in onemound, and fragments of a child’s skeleton inanother (Erdenebaatar 2002: 211-3).

At some of the khirigsuurs, there is evidencefor an extended period of growth and use.As mentioned earlier, the radiocarbon datesobtained from satellite mounds at Urt Bulagynsuggest its possible outward growth over aperiod of a few centuries. The large khirigsuurKYR119 in Khanuy valley presents furthersupporting evidence (Figure 9). The curvedsite plan indicates the presence of three paths,each with a slightly different orientation, while

the stone circle areas are seen to interdigitate in and among islands of previously constructedsatellite mounds. Kurgans are also known to have been reused as burial grounds over anextended period of time, and in some cases expanded for such a purpose (Mallory & Adams1997: 651). In attempting to assign single functions to monuments, archaeologists also riskignoring their significant potential as multifunctional components of the social and politicallandscape. Thus, aside from their apparent role as elite tombs and clan burials, kurgans havealso been interpreted as beacons marking territorial boundaries or routes of communication(Koryakova 2000). There is no reason to deny the possibility that kurgans and khirigsuurswere also used as powerful reminders of ancestral links to specific sectors of the landscape aswell as seasonal gathering places for social and ceremonial occasions.

Khirigsuurs were clearly important as stages for the carrying out of a range of rituals,some of these incorporating animals and/or the movement of celestial bodies. As pointedout above, one of the striking features of the khirigsuur phenomenon is the significant andwidespread regularity witnessed at all levels of ritual practice and space, including whatinitially appear to be minor structural and behavioural elements. To be sure, such consistencyhints at the presence of ritual specialists and the generational transmission of ritual knowledgeover centuries, although differences among and within khirigsuurs (e.g. varying practicesassociated with nearby satellite mounds) should not be ignored and may suggest that ritualpractice was permitted to vary within certain limits at any one time. It should be notedthat among nomadic pastoralists, instances of leadership by religious specialists are notuncommon. In traditional Mongolia, a clan chief was sometimes both political leaderand shaman (Jagchid & Hyer 1979: 171). Although no direct links are here suggested,Mongolian shamanism may offer further clues regarding ritual practice during the BronzeAge. Thus, shamanism is marked by an absence of centralised religious authority as well as bylimited but real variation among individual shamans in how rituals are conducted (Sarangerel

558

Page 13: Allard F Et Al 2005 (Mongolia - Bronze Age Rituals & Mobility)

Res

earc

h

Francis Allard & Diimaajav Erdenebaatar

Figure 9. The khirigsuur KYR119.

2000: 74). In this respect, we may consider the previously mentioned variability in ritualpractice encountered at nearby satellite mounds at single khirigsuurs and wonder whetherit might be evidence of the presence of different, possibly itinerant, ritual practitioners.

559

Page 14: Allard F Et Al 2005 (Mongolia - Bronze Age Rituals & Mobility)

Bronze Age Mongolian khirigsuurs

The more or less consistent structural and behavioural elements associated with thekhirigsuur does not mean that it should be viewed as an isolated phenomenon. In fact,khirigsuurs share a number of such elements with sites of earlier and contemporary culturesin other parts of the Eurasian steppes. Burial under mounds (kurgans) is certainly a well-known and widespread feature of this vast region, as is the important role that livestockanimals played in ritual. Centred in parts of southern Siberia and Mongolia, the thirdmillennium BC culture known as Afanasevo has burials that consist of mounds surroundedby stone fences (Mallory 1989: 223-6). By the second millennium BC, some of the featuresof the eastern steppes’ Andronovo cultures are also shared with the khirigsuur phenomenon,including stone barrows, stone slab cist burials, four-sided and circular enclosures madeof surface stones or vertical slabs and the internment of selected parts of livestock – forexample the familiar head and hoof deposits – that accompanied human inhumations orwere buried in separate pits (Kuzmina 2001). Located in the nearby region of Tuva tothe west of Mongolia, the ninth-eighth century BC kurgan known as Arzhan 1 consistsof a main tomb located at the centre of a large circular structure subdivided into nearly100 compartments built of wooden logs, with many of these compartments containingthe remains of individuals and fully caparisoned horses. A circular stone fence defines theperimeter of the site. Although similar in some ways to Mongolia’s khirigsuurs, Arzhannevertheless stands out as a distinctive monument (Askarov et al. 1992).

DiscussionThe data collected so far by the project in the Khanuy valley point to a number of possibleinterpretations, although poor chronological control and limited settlement data makesome of these highly tentative. The first proposition is that the khirigsuurs were built bymobile populations, possibly nomadic pastoralists. Field surveys in Khanuy valley and innearby regions of Mongolia have yet to identify any visible remains of permanent dwellingsassociated with khirigsuurs, while preliminary testing has revealed only subsurface scattersof ceramic sherds. Such mobile populations may have been herders. Mobile hunter-gatherersocieties typically do not build labour-intensive monuments (Bradley 1993: 1-21), whilepedestrian hunters in such societies are not likely to have selectively carried horse headsback to camp, particularly as the skull is heavy and awkward to hold (Olsen 2003: 95).Thus, it is suggested that the horses were either killed and dismembered at the khirigsuursor that parts of the carcass were hauled to the site by horse-mounted hunters, both scenariosimplying at least some degree of horse herding and domestication. The absence of bits orother parts of the bridle at the khirigsuurs (and possibly contemporary slab burials as well)should not be taken as negative evidence of riding, since mounted herding can be donewhile riding bareback (Shishlina 2003) and harnesses can be made of perishable materials(Levine 1999: 14).

It is sometimes claimed that fully nomadic pastoralism emerged on the steppes of Eurasiano earlier than the beginning of the first millennium BC, with earlier pastoralist populationssaid to have been settled or to have practised various forms of semi-nomadic pastoralism(Askarov et al. 1992; Ishjamts 1994; Khazanov 1994: 90-7). Khazanov maintains thatalthough all of the pre-conditions for pastoral nomadism were in place by the mid-second

560

Page 15: Allard F Et Al 2005 (Mongolia - Bronze Age Rituals & Mobility)

Res

earc

h

Francis Allard & Diimaajav Erdenebaatar

millennium BC, it is not until the beginning of the next millennium that the transformationwas complete, a process that he believes was linked to the onset of very arid conditions.Some archaeologists point to features of earlier steppe cultures that suggest some degreeof residential mobility. For example, the third millennium BC Yamnaya Culture (locatedbetween the Black Sea and Ural River) is marked in some areas by an absence of dwellingsand agricultural implements, the presence of kurgan burials out in the open steppes, theuse of wheeled vehicles, and the fact that the domesticated species (cattle, sheep and horse)were well suited to the steppe environment (Mallory & Adams 1997: 652-3; Anthony 1998;Shishlina 2001). However, the absence of substantial settlements should not be construed asirrefutable evidence of a fully nomadic migratory pattern, either in Yamnaya Culture or inthe Mongolian Bronze Age. In fact, the large number and impressive scale of the khirigsuursin Khanuy valley also suggest some degree of residential stability associated with the buildingof the sites and/or their use over an extended period of time.

It has been noted that the centralisation of power in nomadic pastoralist societies is aninherently difficult process, in part due to low population densities and the natural tendencyof mobile pastoralists to disperse (Irons 1979; Khazanov 1994: 152-64). Nevertheless,instances of significant demographic recruitment did on occasion occur in Mongolia. Theseincluded not only the khirigsuur phenomenon of the Bronze Age, but also the manylater walled towns and political capitals (e.g. Khar Balgas during the Uighur period andKharkhorum during the Mongol period), elite cemeteries (e.g. Xiongnu burial groundssuch as Golmod-2 in Khanuy valley), as well as large-scale military operations that reachedbeyond the borders of present-day Mongolia (e.g. the Xiongnu military campaigns intoChina and the Mongol advances through the Eurasian continent during the thirteenth andfourteenth centuries AD).

For the Xiongnu, Uighur and Mongol periods, archaeological and textual evidence pointsto highly centralised power structures that oversaw the construction of sites, cemeteries, aswell as the operation of trade and military activities. In contrast, demographic recruitmentand substantial labour investment during Mongolia’s Bronze Age appears to have beenmostly focused on the ritual sphere, whether in the construction of the khirigsuurs or in thetransport and erection of the deer stones. One further distinctive feature of the Bronze Ageappears to be an absence of long-lived primary political and demographic centres, or of theoperation of large-scale military and trade operations. Instead, the widespread distributionof the numerous khirigsuurs, deer stones and slab burials that dot Mongolia’s steppes at thistime points to a highly fragmented and possibly shifting geopolitical landscape.

ConclusionAlthough better chronological control is required to further address this issue, it is possiblethat the earliest khirigsuurs were constructed and used prior to the appearance of mountedwarriors at the beginning of the first millennium BC. Even if that is not the case, we may stillsuggest that ritual in Mongolia’s Bronze Age served not only an important integrative role,but also provided real – but limited – opportunities for self-aggrandizement. Many anthro-pologists have remarked on the fact that the control of ritual practice and its setting offersan important pathway to power (Bloch 1992). While a monumental backdrop and arcane

561

Page 16: Allard F Et Al 2005 (Mongolia - Bronze Age Rituals & Mobility)

Bronze Age Mongolian khirigsuurs

ceremonies can effectively impress upon followers the powers of the religious leadership, theritual subtext itself serves to reiterate the status quo’s guiding principles. With this in mind,we may inquire as to the opportunities that a highly regulated ritual programme such as thatperformed at the khirigsuurs would have provided to ambitious individuals. The distributionof khirigsuurs and other Bronze Age sites throughout Mongolia suggests a decentralisedgeopolitical landscape in which no single individual controlled the entire region.

Future work on the social structure of early Mongolia and the role of nomadism willbenefit from the finding, dating and mapping of settlements, that forms an importantobjective of the next stage of the project. For a good portion of Mongolia’s prehistory andhistory, one is faced with a highly diffuse archaeological landscape characterised by smallstone-built tombs and ritual sites, as well as temporary campsites that are difficult to locate(and which have until recently generated little interest on the part of archaeologists). Ourproject is initiating a systematic settlement pattern study that will help clarify populationlevels and the distribution of occupation in the valley.

ReferencesAllard, F. 2002. A Xiongnu cemetery found in

Mongolia. Antiquity 76: 637-8.

Allard, F., D. Erdenebaatar & J.L. Houle.Forthcoming. Recent archaeological research in theKhanuy River valley, Central Mongolia, in G.R.Tsetskhladze (ed.). Beyond the steppe and the sown:integrating local and global visions. Proceedings ofthe University of Chicago Eurasian ArchaeologyConference. Leiden: Brill.

Anthony, D.W. 1998. The opening of the Eurasiansteppe at 2000 BCE, in V.H. Mair (ed.). TheBronze Age and Early Iron Age peoples of easternCentral Asia, vol 1: Archaeology, migration andnomadism, linguistics: 94-113. Washington:Institute for the Study of Man Inc.

Anthony, D.W. & D.R. Brown. 2000. Eneolithichorse exploitation in the Eurasian steppes: diet,ritual and riding. Antiquity 74: 75-86.

Askarov, A., V. Volkov & N. Ser-Odjav. 1992.Pastoral and nomadic tribes at the beginning of thefirst millennium B.C., in A.H. Dani & V.M.Masson (ed.). History of civilization of Central Asia,vol 1: The dawn of civilization: earliest times to700 BC: 459-72. Paris: Unesco Publishing.

Bloch, M. 1992. Prey into hunter: the politics of religiousexperience. Cambridge: Cambridge University Press.

Bradley, R. 1993. Altering the earth. Edinburgh:Society of Antiquaries of Scotland.

Cribb, R. 1991. Nomads in archaeology. Cambridge:Cambridge University Press.

Derevyanko, A.P. & D. Dorj. 1992. Neolithic tribesin northern parts of Central Asia, in A.H. Dani &V.M. Masson (ed.). History of civilization of CentralAsia, vol 1: The dawn of civilization: earliest times to700 BC: 169-89. Paris: Unesco Publishing.

Erdenebaatar, D. 2002. The four sided grave andkhirigsuur cultures of Mongolia. UlaanBaatar:Mongolian Academy of Sciences, Institute ofHistory (in Mongolian).

–2004. Burial materials related to the history of theBronze Age in the territory of Mongolia, in K.M.Linduff (ed.). Metallurgy in ancient eastern Eurasiafrom the Urals to the Yellow River. Lampeter: EdwinMellen Press.

Irons, W. 1979. Political stratification among pastoralnomads, in L’equipe ecologie et anthropologie dessocietes pastorales (ed.). Pastoral production andsociety: 361-74. Cambridge: Cambridge UniversityPress.

Ishjamts, N. 1994. Nomads in eastern Central Asia, inJ. Harmatta, B.N. Puri & G.F. Etemadi (ed.).History of civilizations of Central Asia, vol 2: Thedevelopment of sedentary and nomadic civilizations:700 BC to AD 250: 151-69. Paris: UnescoPublishing.

Jagchid, S. & P. Hyer. 1979. Mongolia’s culture andsociety. Boulder: Westview Press.

Khazanov, A.M. 1994. Nomads and the outside world,2nd edn. Madison: University of Wisconsin Press.

Koryakova, L. 2000. Some notes about the materialculture of Eurasian nomads, in J. Davis-Kimball,E.M. Murphy, L. Koryakova & L.T. Yablonsky(ed.). Kurgans, ritual sites, and settlements: EurasianBronze and Iron age (BAR International Series 890):13-8. Oxford: Archaeopress.

Kuzmina, E. 2000. The Eurasian steppes: the transitionfrom early urbanism to Nomadism, in J.Davis-Kimball, E.M. Murphy, L. Koryakova & L.T.Yablonsky (ed.). Kurgans, ritual sites, and settlements:Eurasian Bronze and Iron age (BAR InternationalSeries 890): 118-25. Oxford: Archaeopress.

562

Page 17: Allard F Et Al 2005 (Mongolia - Bronze Age Rituals & Mobility)

Res

earc

h

Francis Allard & Diimaajav Erdenebaatar

–2001. Andronovo, in P.N. Peregrine & M. Ember(ed.). Encyclopedia of prehistory, vol 4: Europe:1-21. New York: Kluwer Academic/PlenumPublishers.

–2003. Origins of pastoralism on the Eurasian steppes,in M. Levine, C. Renfrew & K. Boyle (ed.).Prehistoric steppe adaptation and the horse: 203-32.Cambridge: McDonald Institute for ArchaeologicalResearch.

Levine, M. 1999. The origins of horse husbandry onthe Eurasian steppe, in M. Levine, Y. Rassamakin,A. Kislenko & N. Tatarintseva (ed.). Late prehistoricexploitation of the Eurasian steppe: 5-58. Cambridge:McDonald Institute for Archaeological Research.

Mallory, J.P. 1989. In search of the Indo-Europeans.Language, archaeology and myth. London: Thames& Hudson.

Mallory, J.P. & D.Q. Adams (ed.). 1997.Encyclopedia of Indo-European culture. London,Chicago: Fitzroy Dearborn.

Olsen, S.L. 2003. The exploitation of the horse atBotai, Kazakhstan, in M. Levine, C. Renfrew & K.Boyle (ed.). Prehistoric steppe adaptation and thehorse: 83-103. Cambridge: McDonald Institute forArchaeological Research.

Sarangerel. 2000. Riding windhorses. Rochester:Destiny Books.

Shishlina, N. 2001. Eurasian steppe nomad, in P.N.Peregrine & M. Ember (ed.). Encyclopedia ofprehistory, vol 4: Europe: 124-38. New York: KluwerAcademic/Plenum Publishers.

–2003. Yamnaya culture pastoral exploitation: a localsequence, in M. Levine, C. Renfrew & K. Boyle(ed.). Prehistoric steppe adaptation and the horse:353-65. Cambridge: McDonald Institute forArchaeological Research.

Volkov, V.V. 1995. Early nomads of Mongolia, inJ. Davis-Kimball, V.A. Bashilov & L.T. Yablonsky(ed.). Nomads of the Eurasian steppes in the EarlyIron Age: 319-33. Berkeley, CA: Zinat Press.

563