“The Bridge between Eastern and Western Cultures” Volume 5 Number 2 Winter 2008

In This Issue

About

The Silk Road is a semi-annual publi-cation of the Silkroad Foundation sup-plied in a limited print run to libraries.We cannot accept individual subscrip-tions. The journal can be viewed at:<http://www.silkroadfoundation.org/toc/newsletter.html>. Please feel freeto contact us with any questions orcontributions. Guidelines for contribu-tors may be found in Vol. 2, No. 1(June 2004) on the website.

The Silkroad Foundation14510 Big Basin Way # 269Saratoga, CA 95070

Editor: Daniel C. Waughdwaugh@u.washington.edu

© 2008 Silkroad Foundation© 2008 by authors of individual ar-ticles and holders of copyright, asspecified, to individual images.

From the Editor’s Desktop

Beyond the Sensational: The Reiss-Engelhorn-Museums’ “Origins of theSilk Road”

Every blockbuster exhibitionwants the public to believe that itis “sensational,” since that mayensure large crowds and therevenues to recoup the huge costsinvolved in borrowing, trans-porting and mounting theartifacts. The Reiss-Engelhorn-Museums’ current “Origins of theSilk Road: Sensational New Findsfrom Xinjiang, China” is noexception in this regard.1 But in aworld where Silk Road exhibitionspop up on the landscape likeflowers after a thunderstorm in thedesert, this one goes well beyondthe sensational. With fewexceptions, the material here,displayed first in Berlin beforemoving to Mannheim, is beingshown for the first time outside ofAsia. Readers of this journal willhave about a month yet in whichto visit Mannheim, before theexhibition closes June 1. It isworth the trip.

Thanks to the PublicBroadcasting System, the National

Geographic and some recentbooks (e.g., Mallory and Mair2000), the sensational part of thestory—the largely “west asian”mummies unearthed in Xinjiang—is, in a sense, already well known.In Mannheim, the curators havebeen rather more cautious aboutsensational claims concerningwestern influences in China.Theirs is a nuanced presentationwhich meshes very nicely with thelead article in this issue of ourjournal by the distinguishedGerman archaeologist, HermannParzinger. Dr. Christoph Lind’scuratorial team in Mannheimselected but one of famousmummies (a well-wrapped infantfrom Zaghunluq, ca. 800 BCE,item no. 1012) and instead invitesus to learn about cultures whoseethnic identities remain uncertain.The focus in the first instance isthe ordinary artifacts of the every-day lives and deaths of ihabitantsin the oases and mountains ofXinjiang. We come away with anappreciation for the complexity of

their cultures andinteractions inCentral Eurasiastarting as far backas the late thirdmillennium BCE,some two millenniabefore the periodwhen conventionalnarratives of the silkroads begin. Whilethere is “animal-style” gold work[Fig. 1; exhibit no.166; Fig. 13, p. 6below], bronze, silkFig. 1. Gold neck ring or diadem, Jiaohe-Gou Xi,

Tomb 1, Western Han Period (206 BCE-24 CE).

© Cultural Heritage Bureau of the XinjiangUighur Autonomous Region, China

• The Mannheim Silk Road Exhibit 1

• Silk Roads in Prehistory 7

• Salvage of Nanhai Shipwreck 16

• The Boma Cup 20

• Xiongnu Elite Tomb Complexes 29

• A Xiongnu Satellite Burial 36

• Tahilt Archaeological Survey 42

• Kazakh Food as Culture 48

In Upcoming Issues

· Rahile Dawut on Muslim shrines inXinjiang

· Elliot Sperling on Tibeto-Mongolrelations

· Odbaatar on a Uighur cemetery nearKharbalgas

· Vesna Wallace on Tibeto-MongolBuddhist relations

· Staffan Rosén on Korea and the SilkRoad

· George Saliba on Islamic Scienceand the Silk Roads

and more…

and some noteworthy examples ofHellenistic artistic influences, thereal interest is the material cultureof clay, wood, grains and animalproducts such as bone and wool.As the curators recognize, alimitation of the selection is thatmost artifacts come from burials[Fig. 2], not from settlement sites,few of which have been locatedand studied. It is somewhat ironicthat the chronological desig-nations for so many of the artifactsare Chinese dynastic dates, eventhough so little of the indigenousculture displayed here had muchof anything to do with Chinaexcept by the accident of politicswhich finds the archaeologicalsites within the borders of today’sPeople’s Republic.

The decision was made toorganize the exhibition byexcavation site, rather thanattempt to preserve a strictchronology throughout. Thus we

begin with theimportant “BronzeAge” cemeteries(which also happento be amongst theearliest, dating be-tween 2200 and1100 BCE) at Qä-wrighul and Xiaohe[Fig. 3], not far fromLop Nor. Then we

move on to other sites in theeastern Tarim Basin and Turfanregion (notably Subexi andYanghai). The spectacular“Yingpan Man” [Fig. 11, p. 4

below; no 162] (the burial of aSogdian merchant?) is one ofmany interesting finds from anearly first millennium CE site justto the west of Qäwrighul. There isa particularly rich selection of theremarkable 2000 to 2500-year-oldtextiles excavated at Sampulanear Khotan and at Zaghunluqnear Chärchän, both on the“Southern Silk Road.” The exhibitconcludes with Kiziltur (just southof the Tien Shan Mountains nearKuqa) and sites in the Ili RiverValley [Fig. 4], where the artifactsrange from striking first-millennium BCE bronzes [Fig. 5,facing page; no. 185] to the goldburial mask unearthed with otherTurkic period artifacts at Boma [foran image, see p. 26 below].

The site at Xiaohe (“LittleRiver”; nos. 3-36) may serve toillustrate both the promise of theongoing archaeological work inXinjiang and the challenges toproper scientific study of thematerial. Although discovered, itseems, by Ördek, who had ledSven Hedin to Kroraina (Lou-Lan)and who would much later in 1934guide Hedin’s archaeologist FolkeBergman to Xiaohe, this sizeablenecropolis of some 2500 squaremeters was not systematicallyinvestigated until its re-discoveryin 2000. At that time, a kind of

Fig. 2. A grave excavated at Zaghunluq.

© C

ultura

l H

eritag

e Bure

au o

f th

e Xin

jian

g U

ighur

Auto

nom

ous

Reg

ion,

Chin

a

Fig. 3. Archaeological sites in the east-ern Tarim Basin and in the Turfan re-gion.

© 2

008 R

eiss

-Engel

horn

-Muse

en

Fig. 4. Sites in the Tien Shan Mountain region and espe-cially the Ili River valley.

© 2008 Reiss-Engelhorn-Museen

2

“salvage archaeology” becamenecessary [cf. Figs. 6, 7], to stayahead of the looters, who were atwork as soon as the news of the

find and its GIS coordinatesbecame known (Mair 2006).Unfortunately the radioactivity inthe area, due to nuclear testing,has made it impossible to obtain14C dates.

Although now desert, Xiaohemust have been created in aperiod when the region was wellwatered. The boat-shaped coffinsinclude both human remains and“substitute” figures made oforganic materials. The burial of“substitute” figures or placing ofanthropoid sculptures in or nearthe graves was not uncommon atother sites as well [Fig. 8; exhibitno. 57]. The wooden poles and“paddles” marking the graves atXiaohe, as well as certain otherobjects, suggest a ritual concernover fertility and reproduction. Asis the case though with so manyother archaeological sites, we canonly guess at the meaning of manyobjects — for example [Fig. 9,next page; no. 6], a small long-nosed wooden mask attached tothe chest of a elderly womanwhose attire included a striking feltcap and fur boots. Graves atXiaohe contained delicately wovenbaskets or sieves, throwing dartsbound with feathers, arrows, rawsheep’s wool and woolen textiles.

Of particular interest are a 5th-century BCE five-stringedKonghou harp found at Yanghai(no. 69), the oldest such yet

known, and a perfectly preservedcompound bow, arrows and armguard from Subexi (nos. 77, 78).Among the most importantexhibits are some of those whichmake the least striking visualimpression. For example, thereis a rare 5th century BCEearthenware bellows valve from

Fig. 5. Bronze figure. 71 -Tuanchang, Burial Site, 5th-3rd cen-tury BCE.

© C

ultura

l H

eritag

e Bure

au o

f th

e Xin

jian

g U

ighur

Auto

nom

ous

Reg

ion,

Chin

a

Figs. 6, 7. Chinese excavations inXinjiang.© Cultural Heritage Bureau of the Xinjiang Uighur Autonomous Region, China

© C

ultura

l H

eritag

e Bure

au o

f th

e Xin

jian

g U

ighur

Auto

nom

ous

Reg

ion,

Chin

a

Fig. 8. Mask made of clay and wood,Yanghai, Burial Site II, Tomb 2063,ca. 5th century BCE.

3

Yanghai (no. 61), part of theequipment needed for smeltingore. A set of bowls (no. 79) fromSubexi, dating from the 5th-3rd

centuries BCE, contains food

offerings, includingmeat cubes and millet;a pouch from Sampula(no. 129), dated to theHan Dynasty era,contains millet biscuits.Among the numerousexamples of pottery area rare early 2nd-mil-lennium BCE vesseldecorated with animage of a man and awoman (geometricdesigns are the morecommon), a unique 5th

century BCE twin cup(no. 73), and a rare, forSampula, wheel-turnedearthenware jug(no.125). Spindles withcarved weights may becommon enough (nos.95, 97), but rarely dothey preserve traces ofyarn. Thread spools (no.127) (possibly originallyattached to a loom?) areunusual.

The most vividmemory of the exhi-bition for me is thequality and variety ofthe textiles. Naturally,one’s attention is drawnto Yingpan Man [Figs.10, 11] and the now wellknown fragment of alarge wall hanging withHellenistic motifs fromwhich trousers weresewn for a burial atSampula [Fig. 12, facingpage; no. 113].3 Un-doubtedly the garmentsin those graves weresewn specifically for theburials, as were someother objects (socks,gloves). However, forthe most part we areseeing what appear tohave been every-dayclothes, some soiled orworn from wear. Thereare trousers, skirts,kaftans, shirt fronts,and hats. The great majority arecrafted from wool, some examples(e.g., the caps, nos. 86, 87)

showing an exquisite ribbed meshweave technique. Some of theweaves are plain; others brightly

© C

ultura

l H

eritag

e Bure

au o

f th

e Xin

jian

g U

ighur

Auto

nom

ous

Reg

ion, Chin

a

Fig. 9. Mask, willow wood, withstrings, textile fibers and red paint,Xiaohe, Tomb C, Bronze Age (ca.2200-1100 BCE.

Fig. 10. Details of textiles on“Yingpan Man.”

© C

ultura

l H

eritag

e Bure

au o

f th

e Xin

jian

g U

ighur

Auto

nom

ous

Reg

ion,

Chin

a

© C

ultura

l H

eritag

e Bure

au o

f th

e Xin

jian

g U

ighur

Auto

nom

ous

Reg

ion,

Chin

a

Fig. 11. Burial outfit of a deceased man, Yingpan,Tomb 15, Eastern Han to Jin Period (25-420 CE);wool and silk.

4

colored (e.g., no. 99, a dress withbrown skirt but bright red and bluechecker pattern top and redpiping). Of particular interst areseveral skirts from Sampula with

woven bandsdepicting bac-trian camels,trees or fan-tastic creatures(nos. 102, 104,106, and 117).There is at leastone textilefragment (no.152), probablyimported, withan unusual ta-queté weavingt e c h n i q u e .With one pos-sible exception(no. 96), all thesilks date fromthe Han Dyn-asty period orlater.

The spa-ciousness ofthe displayarea is one ofthe great vir-tues of thee x h i b i t i o n .These oftenquite ordinaryclothes of or-dinary peopleare not rele-gated to ob-scurity in acorner — manyin fact take ona starring roleas the centralfocus of a roomor an alcoveand thus de-mand the view-er’s close at-tention. I wasf o r t u n a t eenough tobegin viewingthe exhibitbefore its for-mal openinghour, but evenonce the public

began to arrive, I never felt that Iwas being hurried along by crowdsas had been the case in thecramped and poorly laid out spacein which the Musée Guimet

mounted its Afghan treasuresexhibition last year in Paris.

Apart from the Silk Roadexhibition in Mannheim, I wouldstrongly recommend that thoseinterested in archaeology visit theexcellent Palatinate Museum inhistoric Heidelberg, only 15minutes away on the train. Thereone can enjoy a rich collection ofpre-historic and Roman periodmaterial. If your calendarexcludes Mannheim before June,then at very least obtain thecatalogue of the Silk Road exhibit(Ursprünge 2007), in which youcan view the objects and read anumber of interesting essays,including one on daily life in theearly Xinjiang communities writtenby Dr. Ulf Jäger, a contributor tothis journal. And in the meantime,enjoy on the pages below asubstantial amount of newinformation about other dis-coveries which are expanding ourknowledge about cultures of thesilk roads, their pre-history, andtheir legacy down to the present.

Daniel C. WaughProfessor EmeritusThe University of Washington,Seattle<dwaugh@u.washington.edu>

References

Mair 2006

Victor H. Mair, “The Rediscoveryand Complete Excavation ofÖrdek’s Necropolis,” The Journalof Indo-European Studies 34/3-4(Fall/Winter, 2006): 273-318.

Mallory and Mair 2000

J. P. Mallory and Victor H. Mair. TheTarim Mummies: Ancient Chinaand the Mystery of the EarliestPeoples from the West. London:Thames and Hudson, 2000.

Ursprünge 2007

Ursprünge der Seidenstraße:Sensationelle Neufunde ausXinjiang, China, Alfried Wieczorekand Christoph Lind, eds.Stuttgart: Konrad Theiss Verlag,2007. ISBN 978-3-8062-2160-2.

Fig. 12. Trouser fragment, made, apparently, from a largewall hanging, Sampula, Tomb I, Han Period (206 BCE-220CE). The provenance of the textile is uncertain.

© C

ultura

l H

eritag

e Bure

au o

f th

e Xin

jian

g U

ighur

Auto

nom

ous

Reg

ion,

Chin

a

5

Notes

1. The curatorial team for theexhibition was headed by Dr.Christoph Lind; apart from theReiss-Engelhorn-Museen inMannheim, the presenters in-cluded the Martin-Gropius-BauBerlin and the Eurasia Departmentof the German ArchaeologicalInstitute. Funding was providedby several institutional andcorporate sponsors. The ex-hibition objects are from theCultural Heritage Bureau of theXinjiang Uighur AutonomousRegion, China.

2. With the exception of the twomaps provided by the exhibit

organizers, my images here arelimited to the few publicity photosmade available by the CulturalHeritage Bureau of the XinjiangUighur Autonomous Region,China. The item numbers arethose in the caption list kindly sentme by Dr. Christoph Lind; Iassume they correspond to thepublished catalogue numbers,which I have not yet had theopportunity to check.

3. The Sampula textile was alsopart of the Metropolitan Museum’sexhibition “China: Dawn of aGolden Age, 200-750 AD”; see<http://www.metmuseum.org/special/china/s3_obj_1.R.asp>.

Fig. 13. Gold foil belt buckle ornament depicting a griffin attacking atiger, Jiaohe-Gou Bei, Tomb 1-mb, Han period (206 BCE- 220 CE).

© C

ultura

l H

eritag

e Bure

au o

f th

e Xin

jian

g U

ighur

Auto

nom

ous

Reg

ion,

Chin

a

6

The ‘Silk Roads’ Concept Reconsidered: AboutTransfers, Transportation and TranscontinentalInteractions in PrehistoryHermann ParzingerStiftung Preußischer KulturbesitzBerlin

Today’s widely used designation,the “Silk Road,” goes back to theGerman scholar Ferdinand vonRichthofen, who travelledextensively in China during thesecond half of the 19th century. Heformulated the term “the silkroads” (Höllmann 2004, 37;Waugh 2007). In doing so, hechose the plural form, being quitecognisant that not just one singletrack was involved, but a widelybranching and ancient trans-continental network oftransportation and communica-tion. Yet, now, at the beginning ofthe 21st century the plural form isnearly forgotten; instead, thesingular form, the Silk Road, iscommonly used to designate whatRichthofen meant. In fact, theplural is more correct and reflectsmore precisely what thistranscontinental travel andcommunication network reallywas. This system of travel andtrade routes is indeed ancient anddeveloped over millennia.Archaeological evidence, which isour focus here, has greatlyextended our knowledge of thecultural exchange across Eurasiafrom prehistoric times anddemonstrates the antiquity of thisnetwork.

The geographical setting ofthe silk roads: barriers androutes

The natural environment of theregions through which the silkroads proceeded is exceedinglyvaried. High, precipitous moun-tains covered with snow andseemingly endless deserts wereobstacles along the silk roads.These natural barriers often

dictated the route to be taken.Therefore a knowledge of thegeographical framework isessential for understanding thedistribution of cultures and theircontacts since earliest times.

The eastern terminus of theclassical, major route of the silkroads was the old imperial cityChang’an (Xi’an) in northern China(Höllmann 2004; Debaine-Francfort and Idriss 2000; Baumer2002). The track ran westwardand crossed the Huangho (YellowRiver) in the province of Gansu,whence it then led further to thenorthwest [Fig. 1]. At thesouthwestern reaches of the Gobidesert the main road forked intoa southern and a northern route.The southern route ran parallel tothe Kunlun mountain range alongthe southern fringe of theTaklamakan in the region nowknown as Xinjiang. The northerntrack first crossed the Gobi desert,then circumvented the northernrim of the Taklamakan, followingthe Tian Shan mountains. A

branch left the northern route atTurfan and continued in anorthwesterly direction towardssoutheastern Kazakhstan (Semi-rechye).

Northern and southern routesconverged in Kashgar in westernXinjiang, whence the mainconnection to the west crossed thewestern fringes of the Tian Shanas well as the Pamir range to attainthe Ferghana basin. Despite lowprecipitation, the fruitfulness ofthe Ferghana basin is ensured byrivers that are fed by the meltedsnow and ice from the surroundingmountains. The most importanteast-west route left the Ferghanabasin, passing through theTashkent oasis to Sogdia, whereit traversed the Zeravshan valley,bridged the Amu Darya river, and,after crossing the Karakum desert,progressed across the northernIranian highlands ultimately toreach Mesopotamia and theMediterranean Sea. There wasundoubtedly also a route thatcircumvented the Caspian Sea tothe north and then advancedacross the Eurasian steppe.

The development of thiscomplex and widely branching

Fig. 1. The “Silk Roads” and other trading routes crossing Central Asia.

Copyright © 2008 Hermann Parzinger

7

transportation network – that is,the emergence of the silk roads –cannot be dated with certainty.Historical sources scarcely aid inreconstructing this wearisomeprocess. For example, in ChinaZhang Qian is considered the“progenitor” of the silk roads. Hewas an imperial envoy whojourneyed twice to the west in thelate 2nd century BCE. During histravels he was in Ferghana andreached the upper Amu Darya(Höllmann 2004). The informationwhich Zhang Qian gathered aboutthese distant trans-Pamir regionsdrew close attention at theChinese imperial court. Yet it israther improbable that Zhang Qianwas really the first Chinese whoever reached these areas; morelikely, his report is the first thathas been preserved.

The travel and communicationnetworks across Central Asiawhich were described later as the“silk roads” were neither anachievement of the ancient worldnor even the medieval world. Theirbeginnings lie long before themillennium in which Buddhismemerged. The fact that com-paratively little is known todayabout these early times isprimarily due to an insufficientamount of research, whichhowever has been expanding yearby year. The more knowledgegained about the early cultures ofCentral Asia, the older the

communication network thatlinked them appears, a networkthat in its beginnings as well asmuch later enabled migrations ofpopulation groups as well as thetransfer of goods and knowledge.

Between north and south: therole of Andronovo herdsmenand metallurgists

Looking back into the depths ofprehistory – and that we must do,if we wish to study the emergenceof the silk roads – we dependexclusively upon archaeologicalsources. Long-distance relations inthe area of the later silk roads hadalways been present, perceptibleto varying degrees of clarity inarchaeological contexts. Themigration of certain groups ofpeoples along natural trails, laterutilized for the silk roads, can bedelineated with increasingconfidence in the first half of the2nd millennium BCE. During thisperiod diverse regional groups ofthe Andronovo culture had spreadinto distant parts of Eurasia. Fromtheir origins in present-daywestern Siberia and northernKazakhstan, they expanded theirterritory to the east as far as theYenisei River [Fig. 2].

The Andronovo culture is awidely spread cultural complex,representing the Middle BronzeAge in a great part of the easternEurasian steppe belt (Chernykh

1992; Parzinger 2006). As manyother prehistoric cultures, it ismainly defined by its materialremains, especially a ratherspecific pottery with different S-shaped vessels and incisedgeometric ornaments [Fig. 3,facing page]. Bronze metallurgyand stock breeding are typicalfeatures of this culture, althoughtheir roots date back to the 3rd

millennium BCE. Tin bronzes hadbeen invented even before theMiddle Bronze Age, but theAndonovo culture is connectedwith the first large scaleproduction of jewellery, weaponsand instruments made of tinbronze (Chernykh 1992; Parzinger2006) [Fig. 3]. Sheep, goats,cattle and horses werewidespread, and stock breedingwas the economic basis of theAndronovo population. Not onlythe horse, but also the camel wasof crucial importance for themobility of this culture, and thecamel even enabled theAndronovo population to crosslarge and extremely dry areas.From the steppe and forest-stepperegions between the Ural and theYenisei rivers we know a lot aboutsettlements of this period, someof them rather large. Clearly,sedentary life was quitedeveloped, but the question of thedegree to which agriculture wasknown is still open. Part of theAndronovo population was notconcentrated in villages but livedas mobile herdsmen.

During the first half of the 2nd

millennium BCE mobile groups ofthis Andronovo culture wanderedto the south. They survived the drysteppes and deserts of MiddleAsia, some groups ultimatelyreaching the area of the NamazgaVI culture in southern Turk-menistan and others the territoryof the Sapalli culture in Bactria.Both the Namazga VI and theSapalli cultures are fundamentallydifferent from Andronovo. Insouthern Turkmenistan (NamazgaVI) as well as in southwesternUzbekistan (Sapalli) quite a largenumber of tell settlements havebeen investigated. Both cultures

Fig. 2. Map of the Andronovo culture and its diffusion south- and eastwardsduring the Late Early and Middle Bronze Age (first half of the 2nd millenniumBCE).

Copyright © 2008 Hermann Parzinger

8

are characterized by complexsocieties, living in early urbancentres with public buildings,sanctuaries, workshop areas,living quarters and evenfortifications (Kohl 1984, 1992).Irrigation enabled their popu-lations to develop agriculture withfield crops and even gardenproducts. The centers of thesecultures concentrate in the oasesof the large Central Asian rivervalleys, surrounded by dry steppeor desert. Although there was asyet no writing, Namazga VI andSapalli cultures are part of thenortheastern periphery of theAncient Near East.

At the protourban center ofGonur in Margiana, one of themost important sites of NamazgaVI culture in southeasternTurkmenistan, a temporary campof mobile Andronovo herdsmenexisted in the immediate vicinityof this large, coeval tell settlement(Hiebert 1994). This shows thatindividual Andronovo groups livednear these central communities ofthe south, yet without causing anychange in these more advancedcultures. Areas in the north ofMiddle Asia, by contrast, werebroadly settled by Andronovogroups. At that time Khwarezm onthe lower Amu Darya river, the

Zeravshan valley in Sogdia, theFerghana basin, the Tashkentoasis and Semirechye insoutheastern Kazakhstan wereintegrated into the sphere of theAndronovo cultural community,and thenceforth they followed thedevelopment in the steppe farthernorth rather than that in areas ofoases to the south.

The reasons for the southernexpansion of the Andronovoculture, with some of its groupsmoving even farther south to theborders of Iran and Afghanistan,are unknown. Nonetheless, it isremarkable that the appearanceof the Andronovo culture in MiddleAsia was always associated withmetallurgical activities (Chernykh1992; Parzinger 2006). Numerousartifacts and other evidence pointtowards mining as well as theprocessing of ores. Prehistoricmines and settlements in whichores were extracted andprocessed, ascribable to theAndronovo culture, are known inareas in the modern states ofKazakhstan, Uzbekistan andTajikistan.

Moreover, not only copper oreswere mined, but also tin – thatindispensable component for theproduction of bronze. Whereascopper ores were available farthernorth in the region of the Uralmountains and elsewhere, themuch coveted tin was found onlyin Middle Asia. The exploitation oftin in the Zeravshan valley and ineastern Kazakhstan by theAndronovo culture has beenconfirmed (Parzinger and Boroffka2003). Although objects made oftin bronze are already known priorto this time, they were firstproduced on a large scale bycraftsmen of the Andronovoculture and thence becamewidespread in all of westernSiberia and Middle Asia. Thepeoples of the Andronovo culturewere not only herdsmen, but theywere also highly experiencedminers and adept metalworkers.Possibly it was the abundance ofores in those particular mountainranges in Middle Asia that

Fig. 3. Typical pottery and bronze objects of the Andronovo culturein Siberia and Kazakhstan. After Parzinger 2006.

9

attracted the Andronovo culture tothe south.

Farther south in Xinjiang,during the period of the Afanasevoand Okunev cultures of the 3rd andbeginning of the 2nd millenniumBCE, there are signs of contactwith the steppe cultures of thenorth. The latter extended as fareast as Gansu, as evidenced bycertain bronze objects of theSejma-Turbino type from the Qijiaculture (Debaine-Francfort 1995).Yet it was not until the first half ofthe 2nd millennium BCE thatAndronovo groups fromsoutheastern Kazakhstan slowlypenetrated the Dzungarian basinby means of the Ili River valley,thereby making use of a route thatlater was one of the northernbranches of the silk roads (Mei2000). This passageway throughthe mountains made it easy toreach China from Middle Asiawithout over-coming high pass-es or impene-trable deserts.D z u n g a r i a ncemeteries suchas Sazi haverevealed typicalAndronovo pottery[Fig. 4], whilebronze objects ofthe Andronovotype are known inthe entire region.So it seems thatDzungaria was apart of the largeAndronovo culturalsphere during thefirst half of the 2nd

millennium BCE.

The Andronovo culture playeda central role in the disseminationof the knowledge of bronzemetallurgy in wide parts ofEurasia. Therefore, it seemsreasonable to assume that theculture’s penetration into Xinjiangexerted considerable influenceupon the beginnings of metallurgythere and possibly in neighbouringGansu, through which its effectswere felt even farther in centralareas of China. However, moreresearch must be conducted firstin order better to comprehend andexplain these connections.

Between east and west: theemergence of new cultures

During the following centuries,from the mid 2nd to the start ofthe 1st millennium BCE, tiesbetween the north and the south,especially between northwestern

China and southeasternKazakhstan continued. Necro-polises in Dzungaria yield potterythat displays little recognisablesimilarity to that of cultures incentral and southern Xinjiang andin Gansu; instead it can beassigned to the sphere of the LateBronze Age Karasuk and inparticular Begazy-Dandybaycultures in southern Siberia andKazakhstan (Mei 2000). So thesepopulations of the late 2nd

millennium BCE followed the sameroutes to the south as had theAndronovo groups previously. Yetmovements of these northerncultures farther south cannot bedetected, the archaeologicalevidence thus suggesting thatmigrations from the north to thesouth started becoming lessintensive. This is one of the basicchanges connected with thetransition from the Middle(Andronovo) to the Late BronzeAge (Karasuk, Begazy-Dandybay)in Middle Asia around the middleof the 2nd millennium BCE.

Meanwhile a large part of theregions of the so-called silk roadswas less influenced from thenorth. Cultures of the moresoutherly parts of Central Asiabecame dominant and changedthe cultural picture of the wholearea [Fig. 5]. At that time therewas a characteristic handmade,

Fig. 4. Andronovo pottery from Sazi in Xinjiang. After Mei 2000.

Fig. 5. Late Bronze Age cultures in Central Asia in late 2nd and early 1st millennium BCE.Copyright © 2008 Hermann Parzinger

10

painted pottery, widespreadbetween the Caspian Sea andChina [Fig. 6]. Hence it hasbecome of considerable interest tostudy the patterns of distributionand directions of dissemination,especially since the cultures inquestion occupied areas throughwhich the silk roads later passed.Debates about the evidenceintensified when in recent yearsimproved research in Xinjiangdiscovered cultures connectedwith this handmade, paintedpottery. The question is not simplywhere was the new potteryproduced first and how it spread.What is far more crucial is thequestion as to whether the spreadof this pottery can be connectedwith other cultural developmentsand population groups.

Research has recently linkedcultures with handmade, paintedwares found in western CentralAsia with an “Iron Age Oxusculture,” which includes the groupsJaz in southern Turkmenistan,Tillia in northern Afghanistan andKuchuk in southeastern Uzbek-istan [Fig. 7] (Francfort 2001;Shaidullaev 2002; Parzinger2006). Since previously these

areas had used exlusivelymonochrome wheelmade wares,

the immigration of foreign peopleswas presumed to be the cause ofthis break in ceramicdevelopment. Yet their origincannot be determined withcertainty. Further, becausecultures with similar materialpossessions were present in theTashkent oasis (Burgulyuk) and inthe Ferghana basin (Chust) [Fig.5], with comparable materialreported from Xinjiang as well,one is inclined to presume thatthese groups advanced from eastto west through the Ferghanavalley. In other words, themovement was in the oppositedirection to that of the spread ofmetalworking a few centuriesbefore.

Comparison of the structures ofthese cultures demonstrates thatthe urban-like oasis settlementswith mudbrick architecture andirrigated agriculture are theprimary characteristics of the IronAge Oxus culture, which is based

Fig. 6. Handmade painted pottery from Xinjiang. After Mei 2000.

Fig. 7. Handmade painted pottery from Kuchuk in Uzbekistan.After Shaidullaev 2002.

11

upon an uninterrupted traditionfrom the Bronze Age. By contrast,only elements of this culturalsystem are found in culturesfarther east. This does not refutethe presumed immigration of newgroups to Bactria and furtherwestwards to southern Turk-menistan (Margiana), who mighthave brought handmade, paintedvessels. However, the achieve-ments of the Iron Age Oxusculture in the first instanceemerged from indigenous oldertraditions and not primarily as aresult of influences coming by wayof Xinjiang and Ferghana.

Be that as it may, it isnoteworthy that the change whichthe Iron Age Oxus culture and theother groups with handmade,painted pottery illustrate wascomplete at about the same time–after the middle of the 2nd

millennium BCE – in almost all ofCentral Asia. This result certainlywould not have been possible if acomprehensive and widely

branching network of long-distance travel and communi-cation routes had not already beenpresent. It is surely no coincidencethat this network alreadyforeshadows the underlyingfeatures of the later silk roads.

The first mounted nomads: anew dimension in mobility

The first millennium BCE saw theemergence in the Eurasian steppeof mounted nomadism and a con-c o m i t a n tg r e a t e rm o b i l i t y( P a r z i n g e r2 0 0 4 ;Lebedynsky2006). Inlarge areas ofCentral Asia –e.g. in Turk-m e n i s t a n ,B a c t r i a ,S o g d i a ,K h w a r e z m ,Ferghana and

other regions – the coexistenceand symbiotic relationship ofpeoples living in urban centers andfarming irrigated fields togetherwith nomadic stock-raisers in theenvirons of the oases are well-attested. Mounted nomadsappeared in other parts of CentralAsia as well (Parzinger 2006).There are convincing parallelsbetween bronze objects found inmany areas of Xinjiang [Fig. 8]and those in southern Siberia andeven in the Ordos region – inparticular, specific types of knives,daggers, arrowheads, horse-gear,mirrors and decorative fittings.The last often display the so calledanimal-style ornament (Tierstil),that is so typical of the nomadhorsemen of the older Iron Age(Mei 2000; Parzinger 2006). Suchevidence also appears along abranch of the later silk roads thatleads from the western Takla-makan over the Karakorum rangeto the upper course of the IndusRiver and ultimately to the broadIndus plain in the south.

In recent years increasingnumbers of finds from cemeteriesof nomadic riders of the 1st

millennium BCE in the Karakorumarea show a clear connection tomaterial from Xinjiang as well assouthern Siberia. The famousgolden necklace from Pattan innorthern Pakistan, decorated allover with the typical animal styleimages (Rahman 1990), seems tobe a close parallel to the goldennecklace which we found in 2001in the Scythian elite burial atArzhan in Tuva [Fig. 9], not too

Fig. 8. Objects of “Scythian” character from Xinjiang. After Mei 2000.

Fig. 9. The golden necklace from Arzhan 2 in Tuva,a masterpiece of Siberian animal style.

Copyright © V. Terebenin, State Her-mitage Museum, St. Petersburg.

12

far from the present Russian-Mongolian border (Chugunov,Parzinger and Nagler 2003; 2006).Many other similar examples canbe adduced, all of which confirmthat during the 1st millennium BCEthe mobility of people and thenetwork of cultural interactionsgained a new dimension.

Likewise, during this period andin the following centuries, butabove all in the Christian era, theUpper Indus area was traversedin both directions by countlessmerchants and pilgrims. This isevident from large numbers ofrock inscriptions written inSogdian, Brahmi, Kharosthi,Persian, Hebrew, Tibetan, Chineseand other languages and scripts,as well as petroglyphs and stupasfrom the Buddhist period(Höllmann 2004). Bronze objectstypical of nomad horsemen, whichbelong to the greater sphere ofmaterial culture of the older IronAge attributed to the Scythians,are indicative of the manyinfluences from the Eurasiansteppe. These spread rapidlywithin Central Asia precisely viathe long-distance network fortravel and communication, itselfrooted in Bronze Age and evenolder traditions.

In addition to the afore-mentioned bronzes a furthersource of material from Xinjiangmust be emphasized, whichilluminates the culture of that timein a special manner. In thecemeteries there the deceasedwere buried in tree coffins, incontainers made of reed or simplyin shallow pits. The favourableconditions in the dry desert sandof the Taklamakan resulted in theexcellent preservation of the deadas well as their whole attire andfunerary gifts of organic material(Debaine-Francfort & Idriss 2000).As these desiccated mummiesexhibit unmistakable Europeanfeatures, the theory hasfrequently been asserted thatthese were Proto-Tocharians,ancestors of the later Indo-Germanic Tocharians, whom somebelieve could be localised in

Xinjiang in the 6th–8th century CE.However, the problem of theTocharians is thus far surroundedby too many speculations and toolittle clear evidence. The factremains that in almost the wholeof the Eurasian steppe belt themajority of the older Iron Agemounted nomads was a Europeanpopulation, supplemented bymerely a few Mongolian indi-viduals, who only later, above allin the Christian era, came to be adominant element in thepopulation of Central Asia. Thus,the discovery of Europeanmummies in Xinjiang is by nomeans surprising.

Yet, even more decisive thanthe anthropological characteri-zation of the mummies is the factthat due to the excellentpreservation of the clothing andaccessories, it is possible to makea detailed reconstruction of theirattire. And here there aresurprisingly close similarities tofinds of the Scythian periodrecovered from the frozen kurgansof the Pazyryk culture in the AltaiMountains (Polos’mak 2001). Thecoincidence relates to the type ofclothing as well as to itsornamentation, color andtechnique of production. Thus,

aside from fur, felt, wool and othermaterials, silk played a prominentrole. The nearly completewardrobe of trousers, skirts,blouses, jackets and coatsincludes as well socks and boots,and hood-, hat- or helm-like headapparel (Polos’mak & Barkova2005; Polos’mak et al. 2006). Thelast includes peculiar pillar-likehats found in graves of females.Recent discoveries in theMongolian Altai from a permafrostgrave at Olon-Kurin-Gol [Fig. 10],excavated by a joint German-Russian-Mongolian expedition(Heinken 2007), confirmed theseclose relations between north andsouth and showed that thePazyryk culture extended muchfarther to the southeast than waspreviously known [Fig. 11, nextpage].

It is astounding that theseexamples of costume, whichusually are never preserved,manifest many more and closersimilarities between Xinjiang andsouthern Siberia than do othercategories of finds. Hence, we areconfronted with a conspicuousuniformity in clothing that extendsover a large area. Occasionallythese findings correlate with ageneral Indo-Iranian substratum

Fig. 10. The mummy of a mounted warrior, preserved in thepermafrost of the Mongolian Altai at Olon-Kurin-Gol 10.

Copyri

ght

© 2

008 D

euts

ches

Arc

häolo

gis

ches

Inst

itut,

Berl

in

13

in the region. But textiles and furswere always important com-modities of trade in Central Asia.Written sources verify this in laterperiods in a compelling manner,and the situation was probably nodifferent during the pre-Christianera. Fabrics were often traded inthe form of complete garmentstoo. Thus, the question arises asto whether the similarities incostume found between Xinjiangand southern Siberia are to beconnected with members of oneand the same group of languagesor people, or may not ratherreflect an increasingly intensiveexchange of goods and in-formation along the silk roads,where the mobility of peoplereached a new dimension.

Conclusions

The evidence of archaeology aloneis sufficient to demonstrate theexistence of a widely branchingnetwork of travel and com-munication in Eurasia starting asearly as the 2nd millennium BCE.This can be seen in the penetrationof Andronovo groups intoDzungaria and the ensuing spreadof bronze metallurgy intonorthwestern and northern China,a transfer of technical knowledgefrom the northwest to thesoutheast. A few centuries laterhandmade, painted potteryappears relatively suddenly andalmost at the same timethroughout the vast areaextending from southern

Turkmenistan tonorthwestern China.This new warereflects a culturalchange, which couldhardly have beenachieved so rapidlyand uniformly with-out a well-func-tioning and broadlybranching networkof communication. Aweb of intersectingrelations betweennorth and south andin particular with an o r t h e a s t -

southwest orientation can berecognised through the finds leftby mounted nomads of the 1st

millennium BCE. The astonishingsimilarities in attire, observable ingeneral and even in details andfound in distantly separated areasdo not necessarily point to ethnicrelationships, but rather to anintensive exchange of furs, textilesand even garments. And it was atthis time that silk first played animportant role.

The examples presented hereillustrate three main points. Firstly,the roots of this network of long-distance travel and com-munication routes that spanned allof Central Asia reach back into thepre-Buddhist past. The networkdeveloped continuouslythroughout the millennia beforeattaining the form in which it isknown in Antiquity and the MiddleAges as the “silk roads.” Secondly,whereas in later historical periodsgun powder, book printing, themanufacture of porcelain andmany other things weretransported via the silk roadsmainly from the East to the West,the spread of knowledge andproducts during Prehistory seemsto have been much more complex,using the same routes, but indifferent and continuouslychanging directions. And thirdly,it becomes evident that thisnetwork does not concern just oneroad or even a few main tracks,but – as in the sense of the originaldefinition by Ferdinand vonRichthofen – an expansive

network with many branches andcountless trails, all of whichcontributed in the same mannerto the spread of groups of peoplesand the transmission ofknowledge, techniques, wares,religious ideas and forms of artisticexpression. Again and againchanges took place among thecultures involved in this web,whereby the manner, directionand intensity of change weredependent upon the most diversefactors, which cannot always bereconstructed today.

During recent centuries thesignificance of this travel andcommunication network hasdiminished greatly. In the late 19th

century Russian and Britishterritorial interests collided in thearea. For a lengthy part of the 20th

century the millennia-old routeswere cut off through theconfrontation between the easternand western political blocs and tono lesser a degree by the IronCurtain between the former SovietUnion and China. Only in recentyears has a serious opportunity torevivify the silk roads emerged: asa fascinating subject of furtherresearch, and, in its original,millennia-long function, as amedium for the exchange of goodsand ideas and the peacefulencounter between people ofdifferent nationality, skin colourand religion.

Note: This essay is based on akeynote address delivered on 7October 2005 to the Ferdinand vonRichthofen Symposium 2005 atHumboldt-Universität zu Berlin.

About the author

Hermann Parzinger studiedPrehistory in Munich, Saarbrückenand Ljubljana (Slovenia). AsDirector at the GermanArchaeological Institute hefounded its Eurasian Departmentand directed research projects inTurkey, Iran, Kazakhstan,Uzbekistan, Tajikistan, Siberia andMongolia. Since 1996 he hastaught Prehistory at the FreeUniversity in Berlin, and in 1998

Fig. 11. Typical wooden carvings of the Pazyryk cul-ture, found at Olon-Kurin-Gol 10 in the MongolianAltai.

© 2

008 D

euts

ches

Arc

häolo

gis

ches

Inst

itut,

Berl

in

14

he was the first archaeologist toreceive the Leibniz Prize, thehighest scientific award inGermany. In 2003 he was electedPresident of the GermanArchaeological Institute, and inMarch 2008 he became Presidentof the Prussian Cultural HeritageFoundation, which is one of thelargest cultural complexesworldwide. His main scientificinterests are the phenomena ofcultural contacts in Prehistory, thebeginning of sedentary life andagriculture in Europe and Anatolia,the early history of metals andmetallurgy and the emergence ofmounted nomadism in Eurasia. Heis member of several Academiesof Sciences and author of morethan 12 books and 200 articles.

References

Baumer 2002

Christoph Baumer. Die südlicheSeidenstraße. Inseln im Sandmeer.Versunkene Kulturen der WüsteTaklamakan. Zaberns Bildbände derArchäologie. Sonderbände der AntikenWelt. Mainz: Philipp von Zabern, 2002.

Chernykh 1992

Evgenii Nikolaevich Chernykh. AncientMetallurgy in the USSR. The Early MetalAge. Tr. Sarah Wright. New York:Cambridge University Press, 1992.

Chugunov, Parzinger and Nagler 2003

Konstantin V. Chugunov, HermannParzinger and Anatoli Nagler.”Derskythische Fürstengrabhügel vonArzhan 2 in Tuva.” Eurasia Antiqua 9(2003): 113-162.

Chugunov, Parzinger and Nagler 2006

Konstantin V. Chugunov, HermannParzinger and Anatoli Nagler. DerGoldschatz von Arzhan. EinFürstengrab der Skythenzeit in dersüdsibirischen Steppe. München:Schirmer/Mosel, 2006.

Debaine-Francfort 1995

Corinne Debaine-Francfort. DuNéolithique à l’âge du Bronze en Chinedu Nord-Ouest: la culture de Qijia etses connexions. Mémoires de la MissionArchéologique Française en AsieCentrale 6. Paris: Editions Recherchesur les civilisations, 1995.

Debaine-Francfort and Idriss 2000

Corinne Debaine-Francfort andAbduressul Idriss. Kériya, mémoiresd’un fleuve. Archéologie et civilisationdes oasis du Taklamakan. Paris, 2000.

Francfort 2001

Henri-Paul Francfort. “The Cultureswith Painted Ceramics of South CentralAsia and Their Relations with theNortheastern Steppe Zone (Late 2nd –Early 1st Millenium BC).” In: RicardoEichmann and Hermann Parzinger, eds.Migration und Kulturtransfer. DerWandel vorder- und zentralasiatischerKulturverhältnisse im Umbruch vom 2.zum 1. vorchristlichen Jahrtausend.Kolloquien zur Vor- und Frühgeschichte6. Bonn: Habelt, 2001, pp. 231-236.

Hiebert 1994

Fredrik T. Hiebert. Origins of the BronzeAge Oasis Civilization in Central Asia.Cambridge, Mass.: Peabody Museumof Archaeology and Ethnology, HarvardUniversity, 1994.

Heinken 2007

Siebo Heinken.”Der Krieger vom Altai.Was eine Mumie aus der Skythenzeitüber die Geschichte eines antikenSteppenvolks erzählt.” NationalGeographic (Deutschland) 2007, no. 6(Juni): 116-129.

Höllmann 2004

Thomas O. Höllmann. Die Seiden-straße. München: C. H. Beck, 2004.

Kohl 1984

Philip L. Kohl. Central Asia. PalaeolithicBeginnings to the Iron Age. Paris:Editions Recherche sur les Civilisations,1984.

Kohl 1992

Philip Kohl. “Central Asia (WesternTurkestan): Neolithic to Early Iron Age.”In: Robert W. Ehrich, ed. Chronologiesin Old World Archaeology. Chicago;London: University of Chicago Press,1992, pp. 179-195.

Lebedynsky 2006

Iaroslav Lebedynsky. Les Saces. Les“Scythes” d’Asie, VIIIe siècle av. J.-C.– IVe siècle apr. J.-C. Paris: EditionsErrance, 2006.

Mei 2000

Jianjun Mei. Copper and BronzeMetallurgy in Late Prehistoric Xinjiang:Its Cultural Context and Relationshipwith Neighbouring Regions. BAR

International Series 865. Oxford:Archaeopress, 2000.

Parzinger 2004

Hermann Parzinger. Die Skythen.München: C. H. Beck, 2004.

Parzinger 2006

Hermann Parzinger. Die frühen VölkerEurasiens. Vom Neolithikum bis zumMittelalter. München: C. H. Beck, 2006.

Parzinger and Boroffka 2003

Hermann Parzinger and Nikolaus G.O.Boroffka. Das Zinn der Bronzezeit inMittelasien I. Die siedlungs-archäologischen Forschungen imUmland der Zinnlagerstätten.Archäologie in Iran und Turan 5. Mainz:Philipp von Zabern, 2003.

Polos’mak 2001

Nataliia V. Polos’mak. Vsadniki Ukoka(The Horsemen of the Ukok).Novosibirsk: Infolio-press, 2001.

Polos’mak and Barkova 2005

Nataliia V. Polos’mak and L. L. Barkova.Kostium i tekstil’ pazyryktsev Altaia(IV-III vv. do n. e.) (Costume andTextiles of the Altai Pazyryks [4th-3rd

centuries BCE]). Novosibirsk: Infolio-press, 2005.

Polos’mak et al. 2006

Nataliia V. Polos’mak et al., Tekstil’ iz‘zamerzshikh’ mogil Gornogo Altaia IV-III vv. do n. e. (opyt mezh-distsiplinarnogo issledovaniia) (Textilesfrom the “frozen” tombs of theMountain Altai 4th-3rd centuries BCE [anexperiment in interdisciplinaryresearch]). Novosibirsk: Izd-voSibirskogo otdeleniia RAN, 2006.

Rahman 1990

S. Rahman. “Unique Find of GoldOrnaments from Pattan/Kohistan.”Journal of Central Asia 12 (1990): 5-35.

Shaidullaev 2002

Sh. Shaidullaev. “Die früheisen-zeitlichen Siedlungslandschaften inNordbaktrien.” Archäologie in Iran undTuran 34 (2002): 243-340.

Waugh 2007

Daniel C. Waugh. “Richthofen’s ‘SilkRoads’: Toward the Archaeology of aConcept.” The Silk Road 5/1 (2007):1-10. On-line at: <http://w w w. s i l k r o a d f o u n d a t i o n . o r g /n e w s l e t t e r / v o l 5 n u m 1 /srjournal_v5n1.pdf>.

15

The Dream and the Glory: Integral Salvage of theNanhai No. 1 Shipwreck and Its SignificanceXu YongjieZhongshan University,Guangzhou, China

The morning of 22 December 2007saw the integral raising from theseabed of the steel caissoncontaining the Nanhai No. 1shipwreck [Fig. 1]. One week afterthat exciting moment, on 28December, the huge caisson wasbrought into the newly establishedMarine Silk Road Museum onHailing Island, GuangdongProvince [Fig. 2]. The completionof the Integral Salvage Project ofNanhai No. 1 was the dream ofChinese underwater archae-ologists come true!

Nanhai No. 1 (South China SeaNo. 1) was a merchant ship of theSong Dynasty (Northern Song,960-1127; Southern Song, 1127-1279 CE) which sank in the SouthSea of China 800 years ago. It wasdiscovered by accident when theGuangzhou Salvage Bureau and aBritish salvage company co-operated in searching for ashipwreck of the East IndiaCompany in August 1987. Morethan 200 objects, such as Songand Yuan porcelains, a gold belt[Fig. 3] and silver sycee (an ingotused for currency) [Fig. 4, facingpage] came out of the sunkenship. Further analysis led to theconclusion that the ship was amerchant boat of the Song-Yuanperiod sailing on the marine SilkRoad.

A series of underwaterarchaeological investigationsbegan. In November 1989, theUnderwater Archaeological SurveyTeam of the South China Sea’sShipwrecks, including specialistsfrom the National Museum ofChinese History and the JapaneseInstitute of Underwater Archae-ology, conducted the first study ofthe shipwreck. Some shards ofporcelain were collected and the

position of the wreck roughlymapped. In April 2001, acollaborative investigation(involving the UnderwaterArchaeology Research Center ofChina and the GuangdongInstitute of Archaeology andRelics) drew an exact map of thesite. Subsequent study includedthe geophysical exploration,underwater investigation andthe excavation of small-scaletest pits in October 2001, March2002, August, 2002, January2003, and May 2004. Aside fromthe sunken wreck’s hull, thefreeboard and the cabin, the

researchers also found anumber of fine artifacts:more than 4,000 pieces ofunbroken ceramics, somelacquerware, stone imple-ments, ironware, bronze,sycees, a large quantity ofcopper coins and so on.

The underwater archae-ological investigationsshowed that Nanhai No.1, avery large ship with a sharp

Fig. 1. The caisson containing NanhaiNo.1 being lifted out of the sea.

Fig. 2. The caisson being transported tothe Marine Silk Road Museum.

Fig. 3. The porcelain bowl andpot; the gilded belt.

All

photo

gra

phs

© 2

008 G

uan

gdong P

rovi

nci

al I

nst

itute

of

Cultura

l Rel

ics

and A

rchae

olo

gy

16

prow, was made of red cedar. Itwas 30.4 m long, 9.8 m wide and8 m high (excluding the height ofthe mast), with a displacement of

600 tons and acargo capacity of800 tons. It datesmost probably to theSong Dynasty. Thehull of the sunkenship was oriented tothe southwest,indicating the prob-able direction it wassailing.

The bundles of ceramics in theship were mainly from the famouskilns of the Southern SongDynasty: the Longquan Kiln of

Zhejiang, the Dehua Kiln and theCizao Kiln of Fujian, and theJingdezhen Kiln of Jiangxi.Specialists have identified somenoted products like the misty bluewares of Jingdezhen [Figs. 7, 8],the whitewares of Dehua [Figs. 5,6], the green-glazed wares ofCizao and the celadon wares of the

Fig. 4. The sycees.

Fig. 5. A whiteware pot of the Fujian Type.

Fig. 6. A whiteware pot with ahandle.

Fig. 7 (bottom). A qingbai-glaze bowlof the Jingdezhen Type.

Fig. 8 (top). A qingbai-glaze bowlof the Jingdezhen Type.

17

Longquan Kiln [Figs. 9, 10]. Thecopper coins discovered in the ship— for example, Zhi Ping Yuan Bao•¡•½Œ³ , Yuan You Tong Baoxxxxxx, and Huang Song Tong Bao• c ‘ v ’ — mostly belonged to theNorthern Song Dynasty [Fig. 11].Surprisingly, Kai Yuan Tong BaoŠJŒ³’Ê› of the Tang Dynasty (618-907) appeared in the ship. TheJian Yan Tong Bao Œš‰Š’ (1127-1130) and Shao Xing Tong BaoЋ»’Ê (1131-1162) of theSouthern Song Dynasty belongedto the latest coins in the findingsand thus provide a terminus a quofor the date of the ship.Furthermore, several very delicateartifacts came out, including agold belt with a garland pattern

knitted using eight goldenthreads, a large gold bracelet witha dragon pattern, a black lacquerbox with red billowing cloud inlay,a sunflower-shaped bronze mirror,

and a large flared celadon bowl[Fig. 10]. In addition, a few exotic-

looking objects areeye-catching, suchas the gold belt anda cobra’s skeleton,indicating thebreadth of culturaland economic inter-action along themaritime silk road.The present esti-mate, sti l l veryrough, suggeststhat there were60,000 to 80,000objects in the cargo.The preservation ofthe cargo notwith-standing, no humanremains have beendiscovered so far inthe ship.

Taking into accountall the data, in-

cluding the wreck’s position andorientation of the hull, as well asthe type, quantity and details ofthe cargo, it was concluded thatNanhai No. 1 was a large Chinese

wooden ship of theSouthern SongDynasty used foroverseas trade. Itsank for reasonsunknown whensailing out of port tothe open sea.

The hull of theNanhai No.1 is still

very well preserved. Since thecargos on the ship were keptprimarily in the hold near the hull,the porcelain and other relics arelargely intact and thus constitutea national treasure. However, thehydrographic circumstances of theship’s location are very complex.In order to preserve completelythe all the data from the ship, theGuangdong Province CulturalDepartment organized in 2002 aprogram named Integral Salvageand Protection Plan of the NanhaiNo.1. The experts from theGuangdong Province CulturalDepartment, the UnderwaterArchaeology Center of the NationalMuseum and the GuangzhouSalvage Bureau discussed andrevised the program six timesbetween 2002 to 2006 andpresented the final plan to theState Administration of CulturalHeritage. In April 2007, theIntegral Salvage Project of NanhaiNo.1 officially started. In theprocess of salvage work, theGuangzhou Salvage Bureauapplied some pioneeringtechnologies, such as the semi-submersible barge and girderperforation under the seabed, thusguaranteeing a successfuloutcome.

The technique of “integralsalvage” involves securing thesunken ship, the cargo andsurrounding silt in a special steelcaisson without disturbing theiroriginal context. The process liftsthe scattered and brittle relics atone time and integrally so thatthey can be moved to theircarefully prepared “crystal palace”— the Marine Silk Road Museum.Once there, the underwaterarchaeological excavation canproceed scientifically in the newenvironment. This approach isdifferent from traditionaloperations such as cofferingsalvage or item-by-item salvage.1

The salvage of Nanhai No.1applied the integrated methods ofsearch, excavation and protectionfor what is probably the first timein the history of underwaterarchaeology. Although relics

Fig. 9. A petaled dish of the Longquan Type.

Fig. 10. The celadon alms bowl of the Longquan Type.

Fig. 11. Copper coins of the SongDynasty.

18

brought up from the ocean maybe mysterious and impressive,their interest often enhanced bymedia coverage, some archae-ologists sti l l look down onunderwater archaeology, becauseits methods have obvious flawscompared to those of landarchaeology. Underwater archae-ology usually undertakes salvageitem by item, and thus lacks anaccurate description and recordsregarding the original context ofthe submarine objects. Whenjudged by the standard of landarchaeology, this method mayseem blind, capricious andfortuitous. The integral salvagemethod applied for the NanhaiNo.1, however, turns theunderwater operation into onesimilar to the excavation work onland, and obtains the same kindof accurate recording of datawhich is possible for undergroundexcavation. The sinking of testcores and methods of underwaterexcavation which were widelyused in the past are equivalent tothe selective sinking of test pits invery limited land excavation. Incontrast, the Nanhai No.1 IntegralSalvage Project has recovered therelics in their entirety, just as inthe comprehensive excavation ofa site on land.

The year 1987 was fruitful forChinese underwater archae-ologists, for it witnessed thefounding of the NationalUnderwater Archaeology Center inthe Chinese Historical Museum aswell as the discovery of Nanhai No.1. The process of excavatingNanhai No.1 — from the initialdiscovery, investigation andsinking of test pits to the integralsalvage — is a record of thegrowth of Chinese underwaterarchaeology during the subse-quent twenty years. The successof the Nanhai No.1 IntegralSalvage Project thus has becomea landmark for Chinese archae-ology.

The raising of the caisson andthe transfer of the ship to the“crystal palace” are yet earlystages in the work — no more than

one-third of the entire project.Careful excavation and pre-servation are still to come and areeven more important for long termsuccess. Many difficult andpressing questions must beanswered: What changes will takeplace after sunken wreck has leftthe original environment? Will thefragile ship framework fall apartafter the box containing the vesselis opened? How do we dehydrateand desalt the wooden hull andsimultaneously minimize oxi-dation? How do we deal with theprobable fissures present in theporcelains which may be causedby saline crystallization on theglazed surface? Also, how do wedeal with the tension between therequirements of careful andextended excavation, on the onehand, and on the other the urgentdemands of the local governmentto develop tourism?

Only when we are able to solvethese problems can thearchaeologists obtain the fullrange of information which theNanhai No.1 contains and providesatisfactory explanations of itsmysteries, both to address theacademic concerns of scholars andthe general curiosity of the public.Perhaps we will then be able toanswer such questions as: Wherewas the Southern Song merchantship built? From whence did theship sail and where was it headed?Was the ship owner a Chinese ora foreign merchant? How manycargos were loaded on the shipand what kind? How were thecabins and cargos arranged? Whydid the ship sink? Was the crewable to escape? Were there anynavigating instruments on board?What was the actual route of themarine Silk Road? What was therole of the Southern Song Dynastyin the marine Silk Road?...

Acknowledgement: Somephotos in this article wereprovided by Professor Cui Yong ofthe Underwater ArchaeologyCenter, Guangdong ProvincialInstitute of Cultural Relics and

Archaeology. I appreciate greatlyhis generous help.

About the author

Professor Xu Yongjie xxxreceived his degrees from JilinUniversity. Between 1998 and2004, he was Deputy Director ofthe Heilongjiang ProvincialInstitute of Cultural Relics andArchaeology. Since 1994 he hastaught Chinese Archaeology in theDepartment of Anthropology ofSun Yat-sen University inGuangzhou. He is also VisitingProfessor at the Research Centerfor Frontier Archaeology in JilinUniversity and the VisitingResearcher at the Research Centerof Ancient Civilization in theChinese Academy of SocialSciences. Among the importantexcavations which he has directedare those at the Dadiwan xxxSite and Donghuishan xxxxGraveyard in Gansu, the XinzhouYouyao xxxx Site in Shanxi, theFenglin xxx City Site inHeilongjiang and the Jiuxianpingxxxx Site in Chongqing. Hisnumerous publications includesome 70 articles and the followingbooks: The Archaeology of MinleDonghuishan (1998); TheArchaeology of Shanxi XinzhouYouyao (2004); Qixinghe — TheSurvey Report of the ShanjiangPlain Ancient Site (The SettlementArchaeology of the Han-Wei Sitesin the Qixinghe Valley) (2004);The Genealogy of the Later PeriodYangshao Culture in China’s LoessPlateau (2007). He can becontacted at <yongjie1957@yahoo.com.cn>.

Note

1. Coffering salvage entailsencircling the sunken wreck andpumping out the sea water, thuschanging underwater archaeologyto land archaeology. In item-by-item salvage, the archaeologistsdove directly to the ship,excavated the relics piece-by-piece, then disassembled theship’s hull, in order to bring itashore and re-assemble it.

19

The Byzantine Element in the Turkic Gold Cupwith the Tiger Handle Excavated at Boma,Xinjiang

Lin YingZhongshan University,Guangzhou, China

In 1997, a hoard with assortedexquisite gold and silver vesselswas found by the local farmers inBoma, Yili (Ili) county, Xinjiang[Fig. 1]. The new finds were soonsent to the Yili County Museum.An Yingxin, the director of the

museum, first reported thetreasure and published apreliminary study in Wenwu (An1999). After comparing the newfinds with other artifactsunearthed in its vicinity, he datedthese precious metal works to thesixth or seventh century CE, thatis, the late Northern Dynasty toearly Tang period. Concerning theowner of the treasure he wrote:

“During the sixthcentury, the TekesiRiver basin was thepolitical center ofthe Western Turks.The court of theqaghan was situatedthere. The discoveryof these high-qualitygold and silver ob-jects will no doubtbe helpful for lo-cating the court ofthe qaghan andanswering otherrelated questions.”

A gold cup with a tiger handleis especially eye-catching amongthese treasures that most likelybelonged to a Turkic chieftain oreven the qaghan himself. It is 16cm high, 10.5 cm in body diameterand weighs 725 g. The rhombus(lozenge) pattern is mold-pressedon the whole body of the cup, andoval red agate is inlaid in eachrhombus. A tiger sculpture isriveted to the upper body of thework as a handle. The tiger has awide, round head, two erect ears,and a long and narrow body. Thestripes that identify the tiger areetched over the whole body of theanimal [Figs. 2, 3].

The Turks were fond of drinkingwine. Two of the ten titles of theirleading officials were derived fromthe words for wine and a drinkingvessel (Cai 1998, pp. 190–91).The gold cup from Bomaresembles the drinking vesselscarved on certain of the Turkicperiod stone images (balbals)found Tuva, southern Siberia (Sun1996, p. 262, Fig. 31). It is verylikely that such cups were used forwine.

The combination of drinking cupand tiger handle reminds us of theclose relationship between thetiger (also the leopard) and

Fig. 1. The location of Boma.

Fig. 2 (left). The BomaCup.

Fig. 3 (below). Thehandle of the Bomacup. All drawings copy-right © 2008 Lin Ying.

Photo

court

esy

of

Prof. W

ang B

inghua.

All

rights

res

erve

d.

20

Dionysus. In Greek myth,Dionysus, the son of Zeus andSemele, once wandered aroundEgypt and India, accompanied byhis teacher Silenus and a groupof satyrs and maenads (Cary etal. 1979, pp. 478–82). After that,he was worshipped by Greeks asthe god of wine. The personalmount of Dionysus was a tiger orleopard. Therefore, the twoanimals constituted a key elementin Greek and Roman depictions ofDionysus. Frequently, they wereportrayed as lying prone on thewine vessel and drinking the wineeagerly. Sculptures of the animalsare usually similar in body shape,but they can be distinguished bythe pattern on the bodies. Thetiger is striped, while the leopardis identified by roundels.

The theme of Dionysus and histiger was also adopted bycraftsman in the Parthian Empire(247 BCE – 226 CE). For example,there is a Parthian drinking vesselwith a tiger handle [Fig. 4].Gradually the tiger image came tobe employed for a specific type ofhandle not even connected withwine or Dionysus, as can be seenin Sasanian and then Islamic

tableware [Figs. 5, 6] (Etting-hausen 1972, pp. 2–8, Fig. 5).

There are some similaritiesbetween the tiger sculpture on theBoma cup and the tiger handle of

the Sasanian vessel:for instance, the thinbody, the pattern ofstripe, and therough treatment ofthe tiger’s feet. Was

the Boma gold cup an import fromSasanian Iran? This seemsunlikely. First, the rhomboidpattern is rare in Sasanian goldand silver works. On the otherhand, it is closer that found onByzantine silverware unearthedfrom an Avar tomb in Albania [Fig.7, next page] (Daim et al. 1996,p. 435, no. 5.502). In addition, thered stone inlay is typicaldecoration for the jewelry andmetal work in the steppe in pre–Islamic times. Most gold objectsin the Boma hoard are inset withred stones. Therefore we believethat the Boma gold cup wasproduced by a local Turkiccraftsman. The tiger handlesimilarly is not likely to be animitation of Sasanian silver work.The tiger was riveted to the upperbody of the vessel where it seemsto be out of proportion with thesize of the cup. In contrast, theSasanian tiger handle is propor-tionate in size to the body of thevessel, and the head and frontclaws of the tiger are located onthe rim of the vessel, as if theanimal is eager to drink the wine.Thus we may conclude that the

Fig. 4. Bronze jug with feline handle, Parthian, 2nd

century BCE, Cincinnati Art Museum. AfterEttinghausen 1972, Fig. 27.

Fig. 5. Sasanian silver jug, partially gilt,with a tiger handle, Iran, 5th centuryCE. The Cleveland Museum of Art. Af-ter Ettinghausen 1972, Fig. 28.

Fig. 6. (left) Bronze ewer with felinehandle, Iran, 8th century CE ; (right)Bronze juglike lamp inlaid with silverand copper, Seljuq period, late 12th

century CE. The Metropolitan Museumof Art. After Ettinghausen 1972, Figs.30, 31.

21

tiger handle of the Boma cup mostlikely had been produced as asculpture before it was attachedto the cup rather than being madespecifically as a handle to matchthe vessel.

When we examinethe development ofsculptures depictingtigers or leopards inByzantium from thefourth century to thesixth century, we findthat the tiger handleof Boma has a closerrelationship to theByzantine works thanto the Sasanian ex-ample. Small sculp-tures of tigers orleopards appeared asearly as the era of theRoman Empire. ByLate Antiquity,Christianity hadreplaced traditionalworship and becomethe state religion. Aspagan customs, thefestival and cere-monies for Dionysuswere finally elimin-ated. It is possiblyagainst this back-ground that thesculptures of tigers

and leopards lost their connectionwith wine and Dionysus began toappear more widely on otherobjects [Figs. 8, 9].

In the meantime, gold andsilver drinking ves-sels with tiger orleopard handlescontinued to be madein the territory of theByzantine Empire.N e v e r t h e l e s s ,c o n t e m p o r a r i e sseemed to like tocollect the handles asprecious objects. Twosmall silver sculp-tures of a tiger and aleopard were dis-covered in theTraprain Hoard inScotland between1914 and 1919 [Fig.10, facing page](Curle 1923, pp. 1–5,79–80). They are ofthe same size, butdiffer in the pattern ofbody decoration.

Those who have studied this hoardbelieve that they were originally apair of handles for a ewer. In 1992,another small silver tiger wasfound in the Hoxne Hoard inEngland [Fig. 11, facing page](Bland and Johns 1993, p. 24).The Hoxne tiger sculpture wasmore exquisitely sculpted thanwere the feline handles found atTraprain. The body of the tiger wasengraved with stripes and inlaidwith niello. In this case, as atTraprain, the sculpture depicted afemale tiger; most likely, pairedwith a sculpture of a male tiger,the two would have been used ashandles for a large silver vessel.The Hoxne Hoard, which wasburied in the beginning of the fifthcentury, included over 200 goldand silver objects and 14,000 goldand silver Roman coins. The goodcondition of these treasuresindicates that they were wellpreserved in their day and laterplaced in the hoard after carefulplanning. This is further testimonythat people liked to collect thesesmall tiger or leopard handles as

Fig. 7. Byzantine silver vessel, Varp, Albania.After Daim et al. 1996, no. 5.502.

Fig. 8. Gold bracelet, Egypt (?), 7th centuryCE. After Ross 1962-1972, Vol. 2, Pl. XXXVIII.

Fig. 9. Silver lamp, Egypt, 7th centuryCE. After Ross 1962-1972, Vol. 2, Pl.XXIX.

22

separate sculptures in daily life.

The sixth century saw the riseof the Western Turk Empire.Sizabul (Ishtemi – Shi–dian–mixxxx in Chinese sources), thefirst qaghan of the Western Turks,sent envoys to Constantinople andestablished diplomatic relationswith the Byzantine Empire. From568 to 576, at least five Byzantineenvoys visited his court. WhenValentinus, the last Byzantineenvoy, set out from Con-stantinople for the territory of theWestern Turks, he was ac-companied by 106 Turks who hadfollowed the previous Romanenvoys back to Constantinopleand now joined the delegation toreturn their homeland (Blockley1986, pp. 171–79). In the sameperiod, the Western Turksextended their military power tothe eastern frontier of theByzantine Empire, aiming topunish the Avars, their oldenemies. When Valentinus arrivedat the court of the qaghan, theWestern Turk soldiers wereundertaking an expedition westand invaded the Crimea (Ibid.). Inthis situation, the Byzantine silverobjects were mostly likely carriedinto the Western Turk Empire asdiplomatic gifts or spoils of war.

There is additional archae-ological evidence for the presentargument. The Seuso Treasure,which became known to the publicin 1990, consists of 14 pieces ofRoman silver tableware. Theinscription on a silver plateindicates that they were onceowned by Seuso, possibly aRoman general of Celtic origin wholived in Pannonia, the Romanprovince south of the frontier onthe Danube. The plate, togetherwith other vessels, was producedaround the early fifth century(Painter 1990; Mango 1990;Mango and Bennett 1994, pp. 195,205–6, 238–39; for the date, seealso Nagy and Tóth 1990). A bigsilver drinking vessel (TheHippolytus Jug) of the SeusoTreasure [Fig.12a, b] has twohandles in the shape of leopards,whose style and technique ofsculpting resemble those of the

Fig. 10. Silver handles of leopards, 5 th century CE,Traprain, Scotland. After Curle 1923, Pl. XXXI.

Fig. 11. (right) Silver sculpture of a tiger, 5th cen-tury CE, Hoxne Hoard, Britain. Ph

oto

copyr

ight

© 2

008 M

ichae

l G

reen

hal

gh/A

rtSer

ve.

All

rights

res

erve

d.

Fig. 12a, b. Silver amphora and its leopard handles, 4th-5th centuries CE(?), Seuso Treasure. After Mango and Bennett 1994, Figs. 5-1, 5-50.

23

feline handles in the Traprain andHoxne Hoards. Prior to their beingre-attached before the auction atSotheby’s, the handles of theSeuso vessel had been separatedfrom its body. According to thestudies by Marlia Mango andKenneth Painter, at least twoRoman silver vessels are similarto the Seuso amphora. One, fromConcesti, Romania, is in thecollection of the State HermitageMuseum in Russia (Matzulewitsch1929, pp. 131–32; Kent andPainter 1977, p. 297). The shapeof the Concesti amphora is thesame as that of the Seuso one.However, the two handles, whichhad already been separated fromthe body of the vessel, aresculptures of centaurs instead ofa tiger or leopard [Fig. 13]. The

other example is some fragmentsof a silver amphora unearthed atTauteni-Bihor, Romania [Dumi-trascu 1973, pp. 116–18]. Whatsurvives of the reliefs from the sideof the vessel suggests that it ismost likely of the same type. Thistype of amphora came from theeastern Mediterranean region,especially from the workshops inConstantinople.

We do not know theowner and source ofSeuso Treasure.Nevertheless, theinscription on thesilver plate illustratesthat Pannonia wasonce a station in thelong journey of thetreasure. Pannonia,the Roman provinceon the Danubefrontier, was firstconquered by Attila,the leader of Huns,and then became a part ofOstrogothic Kingdom. The ownersof the two amphorae found inRomania were most l ikelychieftains of the Huns or Alans,who played important roles in themiddle and lower reaches of theDanube in the fifth century.

Priscus of Panium, the Romanenvoy who went to the court ofAttila in 449 CE, mentioned thatthe Romans gave the Huns manygifts for both men and women,including silver amphorae.Possibly, Roman silver ewers withtiger or leopard handles were sentinto the steppe as diplomatic gifts.In this process, the Roman habitof collecting the tiger handle as anindependent sculpture might havebeen adopted by the nomads. Agold necklace unearthed fromSimleu Silvanici, Romania, pro-vides additional evidence aboutthe interest in objects embodyingthe tiger or leopard sculptures[Fig. 14a, b] (Florescu and Miclea1979, pp. 92-95). The centerpendant of the necklace is a largebead, enclosed by two crossedrings. Two gold leopards, theirfront paws on the knot connectingthe chain and the pendant, faceeach other. Although the leopardsare much smaller than the handlesof Roman silver vessels, in theirshape and creative concept theyare similar. The hoard at SimleuSilvanici dates from the fifthcentury. Its owner was somebarbarian chieftain connected withthe Huns who had probablyimmigrated to the Danube valley.These barbarian tribes ruled the

local Dacians as the militaryvassals of Byzantine Empire.

The Xiongnu xxx (the ancestorsof the Huns) had a tradition ofrespecting the tiger and leopardas symbols of the heroic spirit. Thegold and silver works excavatedfrom the Xiongnu tombs in northChina show that the image of thetiger was highly valued. Chinesesources assert that the Turks weredescended from the Xiongnu. For

Fig. 13. Late Roman silver ewer,Concesti, Romania. After Matzulewitsch1929, Tafel 30.

Fig. 14a, b. Gold necklace with leoparddecoration, Simleu Silvanici, Romania.After Florescu and Miclea 1979, Figs.80, 81.

24

instance, the account of the Tujuexxx (Turks) in the Beishi xxx (thestandard history of the NorthernDynasty, composed by Li Yanshouxxxx in the seventh century)states: “The ancestor of the Turkslived in the west of Xihai (the WestSea) as an independent tribe.They were originally a minor tribeof Xiongnu.” The Kül Tegin stelefrom the Orkhon Valley which waserected in 732 CE mentions achieftain named Bars bäg, whomarried the sister of the qaghanand was granted the title Qaghan(Rui 2001, p. 251, n. 26). “Bars”was an old loan word from Iranian,referring to a leopard. In Turkic,its meaning also embraced thetiger and other large felines. Thata member of the Turkic elite wasnamed after a tiger or leopardindicates that the worship of thesefierce animals continued in Turkicsociety. Possibly it was this culturaltradition which explains why theqaghans accepted the tiger orleopard handles from the Romanmetal vessels and then had themattached to their own preciousobjects.

The Turks had developed thetechnique of producing highquality gold and silver work. ATurfan document from the Astanagraveyard recorded that Jinshixxx (master of gold) Mo-pan-tuoxxxxx was sent by the qaghan tothe Gaochang Kingdom. Accordingto the Turfan documents, the foodprovided for foreign guests inGaochang was classified into threecategories. The master of gold wassupplied with the first-class food,indicating his high status in Turkicsociety. Clearly the craft of goldand silver manufacture was highlyvalued by the Turks. In 678 CE(the first year of the Shengli periodin the reign of Wu Zetian xxxx ),Mo-ch’o xxx (Qapaghan), Qaghanof the Turks, complained that thegold and silver wares sent byEmpress Wu Zetian were “allcounterfeit instead of qualifiedproducts” (cf. Yu 1996, pp. 206-7). The high connoisseurship ofTurkic aristocrats for the gold andsilver objects further illustratesthat the Turkic craftsman were

able to imitate the tiger handlesof Roman tableware.

In conclusion, the tiger handleof the Boma gold cup was verypossibly made in imitation of thetiger sculptures of some Romansilverware. When the worship ofDionysus had gradually dis-appeared in Christian Byzantium,the tiger sculpture came to beconnected again with wine in anew cultural context. The TurkicEmpire connected the Byzantine,Iranian, Indian, and Chinesecivilizations. The Turks not onlytransmitted the material andcultural achievements betweenEast and West, but also combinedin their own distinct culture theelements of different civilizations.Even if the influence of this newcreation was not long-lived, it wasthrough the Turks that foreigncultures penetrated into Eurasiamore profoundly. The Boma goldcup reveals the unique achieve-ment of the early Turks in thecultural exchange between Eastand West.

Acknowledgements

I want to express my gratitude toQu Junxiong xxxx and JiangZehuan xxxx in Hope Workshop,School of Chemistry, ZhongshanUniversity, who kindly made thedrawings for my article. My specialthank goes to Professor DanielWaugh for all his help, patienceand painstaking work to preparethis publication.

About the Author

Lin Ying is Associate Professor inthe Department of History,Zhongshan (Sun Yat-sen)University, at Guangzhou, China.Her dissertation examines howByzantine culture flowed into theTang Empire through nomadicpeople and Central Asianmerchants along the Silk Road [LinYing, Tangdai Fulin congshuoxxxxxxxxxx (Fulin, the Ruler ofTreasure Country: Byzantium andChina in the Tang Period), Beijing,2006]. She is now working on theeconomic and cultural contacts

between the Roman Empire andChina. She may be contacted at<linying2288@ yahoo.com>.

References

An 1999

An Yingxin xxx . “Xinjiang yilizhaosuxian gumuzang chutujinyinqi deng zhengui wenwu”( xxxxxxxxx xxxxxxxxxxxxxxxxxx) (The gold and silverunearthed from Zhaosu county,Yili, Xinjiang). Wenwu xxx 9(1999): 4-11.

Bland and Johns 1993

Roger Bland and Catherine Johns.The Hoxne Treasure: AnIllustrated Introduction. London:British Museum, 1993.

Blockley 1985

R. C. Blockley, introd., ed. and tr.The History of Menander theGuardsman. Liverpool: FrancisCairns, 1985.

Cai 1998

Cai Hongsheng xxxx.Tangdaijiuxinghu yu Tujue wenhuaxxxxxxxxxxxxxxx (The Sog-dians and Turkic Culture during theTang Period). Beijing: Zhonghuashuju, 1998.

Cary 1949

Max Cary et al., eds. The OxfordClassical Dictionary. Oxford:Clarendon Press, 1949.

Curlie 1923

Alexander O. Curle. The Treasureof Traprain, A Scottish Hoard ofRoman Silver Plate. Glasgow:Maclehose, Jackson and Co.,1923.

Daim et al. 1996

Falko Daim et al. Hunnen +Awaren: Reitervölker aus demOsten: Burgenländische Landes-ausstellung 1996, SchlossHalbturn, 26. April-31. Oktober1996: Begleitbuch und Katalog.Eisenstadt: Amt der Burgen-ländischen Landesregierung,Abteilung XXII/1 – Kultur undWissenschaft, 1996.

Dumitrascu 1973

Sever Dumitrascu. Tezaurul de la

25

Tauteni-Bihor (The Treasure ofTauteni-Bihor). Oradea, 1973.

Ettinghausen 1972

Richard Ettinghausen. FromByzantium to Sasanian Iran andthe Islamic World: Three Modesof Artistic Influence. The L. A.Mayor Memorial Studies in IslamicArt and Archaeology, v. 3. Leiden:Brill, 1972.

Florescu and Miclea 1979

Radu Florescu and Ion Miclea.Tezaure transilvane: la Kunst-historisches Museum din Viena.Bucharest: “Meridiane,” 1979.

Kent and Painter 1977

John P. C. Kent and Kenneth S.Painter, eds. The Wealth of theRoman World, Gold and Silver, AD300-700. London: BritishMuseum, 1977.

Mango 1990

Marlia Mundell Mango. “DerSeuso-Schatzfund. Ein Ensemblewestlichen und östlichen Kunst-schaffens.” Antike Welt, 21

(1990): 70-88.

Mango and Bennett 1994

Marlia Mundell Mango and AnnaBennett. The Sevso Treasure. Part1. Art historical description andinscriptions. Methods of manu-facture and scientific analyses.Journal of Roman ArchaeologySupplementary Series, no. 12.Ann Arbor, 1994.

Matzulewitsch 1929

Leonid Matzulewitsch. Byzan-tinische Antike: Studien auf Grundder Silbergafässe der Ermitage.Berlin: W. de Gruyter, 1929.

Nagy and Tóth 1990

Mihály Nagy and Endrével Tóth.“The Seuso Cauldron: Is the SeusoTreasure from Hungary?” Minerva:The International Review ofAncient Art and Archaeology 1/10(1990): 22-23.

Painter 1990

Kenneth Painter. “The SeusoTreasure.” Minerva: The Inter-national Review of Ancient Art and

Archaeology (London) 1/4 (1990):4 ff.

Ross 1962-1972

Marvin C. Ross. Catalogue of theByzantine and Early MedievalAntiquities in the Dumbarton OaksCollection. 3 vols. WashingtonD.C.: Dumbarton Oaks Center forByzantine Studies, 1962-1972.

Rui 2001

Rui Chuanming xxxxx. Gu Tujuebeiming yanjiu xxxxxxxxxx (Astudy on the ancient Turkicinscriptions). Shanghai: Shanghaiguji chubanshe, 2001.

Sun 1996

Sun Ji xxx . Zhongguo shenghuoxxxxxx. Shenyang: Liaoningjiaoyu chubanshe, 1996.

Yu 1996

Yu Taishan xxxx , ed. Xiyu wen-hua shi xxxxxxx (A generalhistory of Xiyu). Beijing:Zhongguo youyi chuban gongsi,1996.

The gold mask found at Boma along with the gold cup.

Copyr

ight

© C

ultura

l H

eritag

e Bure

au o

f Xin

jian

g U

ighur

Auto

nom

ous

Reg

ion,

Chin

a

26

Xiongnu Elite Tomb Complexes in the Mongolian Altai

Results of the Mongol-American Hovd Archaeology Project, 2007

Bryan K. Miller, University of Pennsylvania, USAJamsranjav Bayarsaikhan, National Museum of Mongolian History, Mongolia

Tseveendorj Egiimaa, National Museum of Mongolian History, MongoliaChristine Lee, Arizona State University, USA

The Altai mountains have servedas an area of intensive traffic andinteraction within Central Asia,and the critical passes throughthese mountains were certainlyareas of contention over whichvarying tribes and politiesthroughout history and prehistoryfought to gain control. Due to theunique nature of this section ofMongolia’s western frontier, theMongol-American Hovd Archae-ology Project was established toexplore the remains of earlypolities within this area and worktoward an understanding of howthey both controlled and exploitedthe region. While the vast majorityof archaeological sites attributedto the Xiongnu empire (3rd centuryBCE – 2nd century CE) lie within

central Mongolia and the southernBaikal area, remains of cemeteriesin the Altai evidence the expansionof this nomadic conquest state intothe key peripheral region whichgave them access to the oasisstates and trade routes of the SilkRoads. Monumental tombs of theXiongnu elite have beendocumented numerous times atseveral sites in the core area ofXiongnu sites throughout Mongoliaand South Siberia (Rudenko 1969;Konovalov 1976; Navaan 1999;Mission 2003; Miller et al. 2006;Miniaev and Sakharovskaia 2006,2007), but the existence of asimilar royal cemetery at Tahiltin-hotgor signifies a strong imperialpresence within the Altai. Insummer of 2007, the Mongol-

American Hovd ArchaeologyProject began excavations at thissite in order to explore imperialmonuments at the periphery andfurther clarify the dynamiccharacter of these elitecemeteries.

Tahiltin-hotgor cemetery islocated in Manhan sum of Hovdaimag and was first discoveredand documented through trialexcavations of two small graves in1961 (Volkov and Dorjsuren1963). Further excavations wereconducted in 1987-1990 (Navaan1999) on two mounded rampedtombs, and in September of 2006the cemetery was mapped inpreparation for further exca-vations. The cemetery stretches

Fig. 1. Tahiltin-hotgor cemetery (excavated tombs noted). Inset map indicates location in western Mongolia.Unless otherwise noted, all images are copyright © 2008 Hovd Archaeology Project.

27

east to west between two smallmountains and sits at the mouthof one of the major mountainpasses through the Altai to theDjungar Basin of Xinjiang,northwest China [Fig. 1]. Threetomb complexes were in-vestigated in 2007 with the aimof understanding not only themounded tombs, which havereceived most of the attention ofprevious excavations, but also theaccompanying features of the eliteinterments. We define a tombcomplex as a collection of featuresassociated with the interment ofa single deceased that togetherare arranged as one group. In thecase of the Xiongnu elite tombcomplexes, three main featureshave been thus far documented,all of which are found at Tahiltin-hotgor.

The first feature, and the mostprominent, is the square rampedtomb mound. This earthen moundis oriented approximately north-south and is held together bystone walls. A ramp or “funerarypath” (Mission 2003) leading upto the top of the mound extendsout from the southern wall. Thetomb mounds often have stonelines that run north-south andeast-west across the surface, themost recurring instance being anorth-south line running throughthe center of the ramp and themound. Beneath these tombmounds and under numerouslayers of stone, earth, and/orwood, lies a burial chamber witha decorated wooden coffin inside.As mentioned before, thesemonumental tombs have receivedgreat attention thus far, but recentstudies have begun to drawattention to the featuressurrounding the main tomb andthe details of the pit and moundabove the chamber the chamber.Small graves adjacent to the

larger tombs have been excavatedat several of the elite cemeteriesin Mongolia and Buriatia, andrecent studies have begun toaddress them directly as aparticular phenomenon (Miniaevand Sakharovskaia 2002; Miller etal. 2006). These satellite burialsaverage about three per tombcomplex and are usually arrangedin a north-south line to the east,west, or both sides of the maintomb. A few instances ofsignificantly large lines ofaccompanying burials exist at GolMod 2 cemetery in Arhangai (Milleret al. 2006) and confirm the arc-shape of these arrangements ofburials which flank the tombmounds. The satellite burials bearsurface demarcations of stonerings, and the deceased areinterred in wooden coffinsbetween one to two metersbeneath. These resemblestandard Xiongnu burials (Torbat2004). The third feature of tombcomplexes – stone lines – wasdiscovered recently at Gol Mod 2cemetery (Allard et al. 2002) andsubsequently at Tahiltin-hotgor in2006. Parallel east-west lines of

stone pairs are arranged north ofsome of the tomb mounds at bothcemeteries. These lines wereexcavated for the first time in 2007at Tahiltin-hotgor.

Square ramped tombs1

The two square tombs excavatedin 1987-1990 (Navaan 1999) arelocated in the middle of thecemetery, and each were assigneda number during the 2006 survey.A comparison of the drawings andmeasurements in the originalreport with the tomb remainsvisible now at the cemeteryshowed great inaccuracies, butthe general descriptions appearmore or less correct. The firsttomb (THL-83) was completelyexcavated to expose the entiregrave pit and the area beneath theramp. At the bottom of the pit wasa log chamber, inside of which hadbeen placed a wood plank coffinwith golden flower decoration [Fig.2]. The grave furnishings and theremaining grave goods, includinga large pot in the log chamber andpieces of a chariot atop thechamber, appear to indicatemortuary tradition similar to that

Fig. 2. Tomb 83. After Navaan 1999.

28

documented in royal Xiongnutombs of central Mongolia. Thesecond tomb excavated (THL-82)contained an unlooted doublenested coffin with the completeremains of a woman, thus earningit the name “the Princess Tomb”[Fig. 3]. The excavations cored asingle pit into the center of thistomb to access the chamber; sono excavations were done toreveal the entire structure of thisgrave. A bronze spouted pot andtwo bronze sticks2 found in the

tomb are now ondisplay at theNational Museum inMongolia, thoughthe jewelry found byher head and on herwaist do not residein locatable col-lections.

Tomb 64

In order to docu-ment properly thestructure of thesquare tombs at this

particular site andcompare them tosites in centralMongolia, we choseone large tomb, THL-64, and its associatedfeatures [Figs. 4, 5].The stone cross lineswithin the tomb

structure and the “door” closingoff the ramp closely resembledthose in elite tombs elsewhere.Near the ground level and in thecenter of the mound, at the exactplace where the two lines cross,was found a dense deposit of burntearth and charcoal. The stones atthis intersection do not appeardisturbed, and it is likely that thisdeposit relates to the originalinterment ceremony and not theoccurrence of looting. Unlike mostother tombs, the ramp onlydescends to the first step ledge of

the pit, 1.6 m below the surfaceand 4.4 m above the bottom ofthe chamber. Therefore the ramplikely did not serve as a practicalmeans for rolling the coffin into itsfinal resting place, but rather tookon more of a ritually symbolicfunction related to “funerarypaths.” The outer chamber at the

Fig. 3. Tomb 82. After Navaan 1999.

Fig. 4. Tomb 64 complex

Fig. 5. Tomb 64.

29

bottom was surrounded by stonesand built from notched,interlocking logs. Inside the logstructure lay a wood plank coffinnested in the southeast corner ofthe chamber, much like the innercoffin of Tomb 83 which had beennested in the southwest corner. Alattice and flower iron decorationcovered the outside of the coffin,and scant remnants of silk confirmthat the coffin would have beenwrapped entirely before the iron

decoration was mounted [Fig. 6].The wood of the coffin was fittedtogether with bow-tie shapedtenons across the planks andmortise holes with plugs inside theedges of the wood planks [Fig. 7].

A thin gold circle and thin goldcrescent were found lying next toeach other, and from theimpressed marks and small rustedpin holes on both of these, theyappear to have been previouslymounted on a flat surface – verylikely the surface of the coffin lid[Fig. 8] which was tossed asideduring looting. This circle andcrescent pair echo other such pairsfound in numerous Xiongnu graves(Mission 2003; Torbat 2004) and

are equated withXiongnu rituals ofthe sun and moon.“And the chanyu[emperor] at dawnleaves camp andmakes obeisance tothe sun’s beginning,and at dusk makesobeisance to themoon” (Shiji 1959,110: 2892).

Remains of foodcontainment andserving were placedinside the log cham-ber just north of thecoffin which held thedeceased: an iron

ladle, a painted red wooden tray,a pair of chopsticks, cattlevertebrae and a sheep/goat legbone, and a large handmade grainstorage pot [Fig. 9]. Severalturquoise tear-shaped insets, fivegreen stone insets carved in the

shape of scallop shells, and a bonependant which probably adornedthe deceased were found tossedout of the coffin from looting [Fig.10, facing page]. Lastly, to thenorth of the log chamber is asemicircular niche cut into the wall

Fig. 6. Lattice and quadrefoil iron coffin decoration.

Fig. 7. Coffin tenon construction.

Fig. 8. Gold sun and moon.

Fig. 9. Iron ladleand bone chop-sticks.

30

of the tomb pit where the skulls,cervical vertebrae and legs of onehorse and three horned-sheep/goats were deposited. Two sheepribs were found above the logchamber, one of which has a deepcut mark from the process ofeating the meat on it.

While only the skull, half apelvis and several other bones ofthe deceased were found from thislooted tomb, this represents agreater yield than most royalXiongnu tombs and thus allows usto gain more knowledge about thetomb occupant than usual. Thedeceased was a woman 35-40years old and has evidence on thepelvis bone indicating she gavebirth at least once. The low toothwear for her age probably relatedto access to better processedfood, and the gracile marks ofmuscle attachments show that shedid not do much manual labor.Nevertheless, wear on the lowerspine may indicate continuedhorseback riding, a pattern foundin many of the individuals

excavated here as well as othergraves in Mongolia. Though thistomb was looted, enough remainsthat we may confidently say it isequitable in both structure andburial goods to the larger eliteXiongnu tombs in central Mongoliaand southern Siberia.

Satellite burials

Dorjsuren describes each of thetwo small burials he excavated atTahiltin-hotgor in 1961 as lyingdirectly to the east of squaretombs (Volkov and Dorjsuren1963). From this we may deducethat they were probably satelliteburials. Both graves were markedwith some stones on the surface,maximum 3 m diameter scatter,and about 1.3 m below the surfacelay stone cysts with humanremains. No wooden coffins werefound. In both graves the bodieslay stretched with the headoriented north. In one case thetorso and head were greatlydisturbed and only the legsarticulated, but in the other theleft arm lay stretched to the side,the right arm apparently bent atthe elbow and lying across thechest, and the legs were slightly

bent. The lack ofwooden coffins andbending at the kneessets these burialsapart from thestandard Xiongnuburial of fullystretched supinebodies inside awooden coffin (Torbat2004).

Satellites of Tomb 64

The two satell iteburials excavated tothe east of tomb 64 at

Tahiltin-hotgor [Fig. 11] were verysimilar to those excavated byDorjsuren, and thus quite differentfrom the standard Xiongnu grave.Both individuals were interred instone cysts, not wooden coffins,and both were laid with their legsslightly bent. The secondnoteworthy variance fromstandard Xiongnu graves is thesurface demarcation. Instead ofthe typical thick stone circleovertop the burial pit, these, andall the other satellite gravesexcavated in 2007 were markedby a small stone cluster (which attimes had been scattered) directlyover and approximately the samesize as the burial pit. Unlootedsatellite burials, such as THL-64-2, have allowed us to view theoriginal clustered position of thesurface stone demarcations. Bothof the individuals buried next toTomb 64 were oriented just westof north, as was the main tomb ofcomplex 64. The accompanyingdeceased were teenagers, andTHL-64-2 was possibly a female.If this is true, then it would be thefirst confirmed discovery of afemale in a satellite burial of eliteXiongnu tomb complexes. One of

Fig. 10. Turquoise inlay, scallop-shapedgreen stone, and bone pendant.

Fig. 11. Satellite gravesTHL-64-1 & -2.

31

the main reasons for choosing anentire tomb complex to excavatewas to see the relationshipbetween those in the satellitegraves and the deceased in themain tomb. In this case, theteenagers are oriented to almostexactly the same degree as thecentral barrow, and the traits seenin the human remains show themto also be very closely related toother Xiongnu peoples inMongolia, and thus closelyaffiliated to the woman in the maintomb who also appears closelyaffiliated to other remains withinthe group attributed to theXiongnu (Lee 2007).

Satellites of Tomb 82

Next we chose the twoaccompanying graves on eitherside of tomb 82, previouslyexcavated by Navaan (1999), sothat we could finish the fullinvestigation of this tomb complex[Fig. 12]. Again, both of theaccompanying graves have thesame orientation as the maintomb. Both burials were marked

on the surface withscattered stones,disturbed from theiroriginal position, andthe deceased wereinterred about ameter below thesurface in small pitswith neither woodcoffin nor stone cyst.The deceased inburial THL-82-1 wasfound lying stretched,his body oriented justwest of north, withboth arms at his side,but face down. Thisman had robustmuscle attachmentsfor his lower arms, hands, thighs,and right foot and is the mostmuscular individual at Takhiltin-hotgor, meaning he performedmore labor than the otherindividuals. On either side of thebody, probably in their originalposition, were found bone beltclasps with small holes [Fig. 13].In burial THL-82-2, a small bonepin was found beneath the skull

which had beentossed aside, andthus may have beena hairpin. A sheep/goat astragulus wasalso found to the leftof the waist. Thisman, like the first,also had robustmuscle attachmentson the lower arms,lower legs, and rightfoot. Both of the menburied in thesesatellite burials wereAsian, and thesecond burial hadenough of the skull toshow that he wasclosely affiliated toother Xiongnu. Thewoman buried in the

main tomb (THL-82) excavated byNavaan in 1990 also shows a closebiological affinity to the Xiongnugroup. Again, the individuals in thesatell ite burials are closelyaffiliated to the main individual,and all of them related to theXiongnu biological group (Lee2007).

Satellites of Tomb 25

Lastly, we opened the three burialsflanking the east side of tomb THL-25, even though the main tombwas not excavated, in orderfurther to understand the natureof accompanying burials in thisperipheral royal cemetery [Fig. 14,facing page]. None of these burialsshowed any significant evidenceof disturbance of the surfacedemarcation, and all three hadtight clusters of stones onlyslightly larger than the grave pit.The burials were arranged moreor less in a north-south arc line,though the third grave (THL-25-3), when the exact positions of theburial pits and bodies wereviewed, did not appear to be inline with the previous two graves.Burial THL-25-1 contained thebody of a child about 8 years oldFig. 12. Satellite graves THL-82-1 & -2.

Fig. 13. Bone belt pieces from THL-82-1.

32

oriented slightly east ofnorth and laid supineand stretched, withonly the legs slightlybent. The deceased inburial THL-25-2 was ina stretched supineposition, and orientedalmost exactly north.He was 30-35 yearsold, and had robustmuscular attachments.The condition of histeeth is consistent witha nomadic diet high inmeat, and low inc a r b o h y d r a t e s .Additional detail aboutTHL-25-2 is in thearticle below byJessieca Jones andVeronica Joseph. Whilethe surface stones of burial THL-25-3 appeared not to bedisturbed, many of the deceased’sbones, including the skull cap,were found amongst the stones inthe pit. The rest of the bones werelaid at the bottom of the pit,articulated in a stretched supineposition. The man in this gravewas 40-45 years old. A few ironplatelets were found with smallholes in the corners which mayrelate to their use as armor pieces.The child can be determined asAsian and related to the Xiongnu,while the two adult men, alsoAsian, appear more closely relatedto individuals in Manchuria orSiberia.

The recurrence of robustmuscle attachments on the meninterred in the satellite burialsseems to indicate intense labor.These muscle markers may alsorelate to certain kinds of damageand arthritis that recurs in almostall the individuals. The patterns oftrauma, especially the severetrauma seen in THL-64-1, come

from extreme impact and mostlikely relate to injuries incurredfrom falling off or being thrownfrom a horse. The arthritic patternin these individuals also speaks torepeated horseback riding. Thecombined pattern of arthritis,musculature, and trauma seen inthe individuals here is consistentwith prolonged horseback riding.

Stone lines

One or two stone lines were foundto the north of eight square tombs,the exception being seven parallellines north of Tomb 25. The linesare at least fifteen meters awayfrom the tomb mounds and only afew meters apart from each other.They are often oriented slightly offof an east-west axis so that theyare parallel to the north wall of thetomb mound, which in most casesis oriented west of north. The lineof stone pairs north of tomb 64 isaligned so that the axis of thestone line which cuts north-souththrough the center of the squaremound and southern ramp thenruns directly through the middle

of the northern line,directly between thecenter two pairs. Thetwo stone lines to thenorth of tomb 82 werealso excavated, and theramp of tomb 82 wascleared so as to test therelationship betweenthe stone lines and theaxis of the centraltomb. The centralnorth-south axis oftomb 82, measuredaccording to the centerof the tomb ramp, alsoruns directly throughthe middle of the twoparallel stone lines tothe north. In bothcases, the stone lineswere clearly aligned in

sync with the greater mortuarycomplexes. The results ofexcavation between the stonepairs have begun to shed light ontheir function and use patterns.

The stone pairs in each line lieapproximately two meters apart,though their original positionseems to have been altered insome cases, probably due tostones falling or being knockedover. Between some of the stonepairs, and at times beneath fallenstones, lie deposits of burnt,crushed animal bone. The bonefragments are high fired, resultingin completely black or white boneswith a thin blue/black layer inside.There is also black ash and/orcharcoal deposited with the burntbones, but there is no evidence ofburnt earth between the stones,or burning against the closely setstones. In some cases the burntbones are found with no ash at all.This suggests that the activity ofburning the animal bones was notdone between the stones. Theburning was probably done in adifferent place, after which the

Fig. 14. Satellite graves THL-25-1, -2, -3.

33

offerings were put between thestones. The two easternmoststone pairs of the line in complex64 contain such deposits, but noneof the other four pairs do. In thetomb 82 complex, the first,southernmost line contains burntanimal bones between or beneathevery pair of stones, while onlytwo pairs of stones in the secondline contain such deposits. As thestone pairs seem to “fill up” inorder from nearest stone lineoutward, this may indicate apattern of use over time ratherthan only at the time of intermentof the deceased in the main grave.

This custom of burning animalbones in a place outside of the finaldeposition area and then settingthe remains between stones, findsan exact parallel in Bronze Agekhirigsuur complexes (Allard andErdenebaatar 2005, pp. 3-5) andsheds light on the rituals of theXiongnu which may connect withsuch earlier practices. In theseBronze Age complexes deposits ofhigh-fired crushed animal boneswere placed along with the ashfrom the fires inside small rings ofsingle stones. These stone ringswere aligned in concentric rowsaround an array of stone moundscontaining horse heads, whichwere arranged in rows around astone fence that enclosed a largecentral stone mound. Just like theseparation between centralmound, horse head mounds andstone circles of the khirigsuurs,there is separation in the Xiongnuelite tomb complexes betweenanimals placed above thechamber, to the north of the coffinand inside the chamber, and theburnt animal remains in theexterior stone lines.

Brief discussion

The arrangement and componentsof tomb complexes at Tahiltin-

hotgor, as well as the contents ofthe main tombs, closely resembleelite Xiongnu cemeteries foundelsewhere. The elite graves hereare undoubtedly associated withthe elite imperial culture of theXiongnu polity and thus may beconsidered within the limits of theempire. Analysis of humanremains in these graves seems tocorroborate this association to theXiongnu nomadic polity. Thepresence of Xiongnu peoples inthese elite tomb complexescorrelates easily to the spread ofthe Xiongnu state into this westernfrontier seen also in the materialculture. The possible presence ofpeople from Manchuria, seen inTHL-25-2 and -3, suggests a greatdegree of mobility of differentpeople within the empire and wideconnections across the variedregions under Xiongnu control.Despite the consistencies inmortuary remains for the elite,evidence arises in the satelliteburials of differentiation fromstandard Xiongnu customs.Interments in stone cysts such asTHL-64-1 and THL-64-2 areseldom found in other Xiongnucemeteries, and the flexed legs ofthese two graves and THL-25-1deviate from the prevailing customof fully stretched supine burials inXiongnu graves (Torbat 2004).The most significant difference isthe surface demarcations. Thesmall tight cluster of stones overthe grave pit differs greatly fromthe tradition of stone rings. It isnot clear what these variationsindicate, but it may relate to eitherlocal variations of the Altai orperhaps a social group not yetdelineated within the greatercorpus of Xiongnu remains.

Further excavations of Xiongnugraves outside this monumentalelite cemetery will be conductedin the upcoming 2008 field season.

By investigating burial grounds inthe open valleys and in the highmountain passes we aim furtherto contextualize the burial groundat Tahiltin-hotgor of the imperialXiongnu elite in the peripheralregion of the Altai mountains.

Acknowledgements

Many thanks are to be given forwork by the stupendous Mongolianstudents (and crew) from Hovdand Ulaanbaatar and by thevolunteers (and Daniel Waugh)who came on the SilkroadFoundation 2007 expedition. L.Nyam-Suren was of greatassistance both outside of and inthe field, and Ts. Odbaatar helpedimmensely in the preparations forthe pilot year of the HovdArchaeology Project. Last but notleast, two additional scholars mustbe thanked for their contributionsand dedication to the project:Prokopy B. Konovalov, for bringinghis vast knowledge to the furtherexcavations of Xiongnu remains,and Judy Logan for blessing theproject with her paramount skillsin materials conservation. Supportwas provided by the SilkroadFoundation and a dissertationresearch grant from the HenryLuce Foundation and AmericanCouncil of Learned Societies.

About the Authors

Bryan K. Miller is a Ph.D.Candidate in the East AsianLanguages and Civilizations Dept.of University of Pennsylvania. Hereceived his M.A. in Archaeologyfrom the Cotsen Institute ofArchaeology at UCLA in 2000,after which he began research inTaiwan and Mongolia andmanaged the Xiongnu excavationsportion of the Khanuy ValleyProject. His present researchconcerns historical and archae-ological investigations of early

34

empires in East Asia, namely theQin and Han of China and Xiongnuof Mongolia. E-mail: <millerbk@sas.upenn.edu>.

Jamsranjav Bayarsaikhan iscurator of Bronze and Iron Agecollections at the National Museumof Mongolian History. He receivedhis M.A. in archaeology fromMongolian Educational Universityin 2005, and is presently pursuinga Ph.D. at the National Universityof Mongolia. He is also the chiefMongolian archaeologist for theSmithsonian “Deer Stone” Projectworking in northern Mongolia. Hispresent research concerns BronzeAge ritual monuments and burialsof Mongolia. E-mail:< b a y a r s a i k h a n @nationalmuseum.mn>.

Tseveendorj Egiimaa is amember of the staff of thearchaeology division of theNational Museum of MongolianHistory and a curator of Xiongnumaterials. She received her M.A.in archaeology from MongolianEducational University in 1999.She is presently also anarchaeologist for the Mongol-Russian expedition at the Xiongnuwalled site of Tereljiin.

Christine Lee earned herDoctoral degree in physicalanthropology from Arizona StateUniversity in 2007, researchingthe biological affinities of Neolithicthrough Modern period popu-lations from China and Mongoliausing cranial and dental non-metric evidence. She will be takinga post-doctoral position at theCenter for Frontier Archae-ologyat Ji l in University in China,beginning in autumn 2008.

References

Allard and Erdenebaatar 2005

Francis Allard and DiimaajavErdenebaatar. “Khirigsuurs, ritual

and mobility in the Bronze Age ofMongolia.” Antiquity 79 (2005): 1-18.

André and Yeruul-Erdene 2004

Guilhem André and Ch. Yeruul-Erdene. “Negen Hurel ShiidemniiTyhai (Gol modni hunnu bulshnihereglegdehuuneer).” ArheologiinSudlal (Ulaanbaatar) 22 (2004):110-117.

Di Cosmo 2002

Nicola Di Cosmo. Ancient Chinaand Its Enemies: The Rise ofNomadic Power in East Asia.Cambridge: Cambridge UniversityPress, 2002.

Erdélyi et al. 2000

István Erdélyi et al. ArchaeologicalExpeditions in Mongolia.Budapest: Mundus HungarianUniversity Pr., 2000.

Konovalov 1976

Prokopii B. Konovalov. Khunnu vZabaikal’e. Ulan-Ude: Buriatskoeknizhnoe izd-vo., 1976.

Lee 2007

Christine Lee. “The BiologicalAffinities of Neolithic ThroughModern Period Populations fromChina and Mongolia: The Cranialand Dental Non-metric TraitEvidence.” Unpublished Ph.D.Dissertation, Arizona StateUniversity, 2007.

Miller et al. 2006

Bryan K. Miller, Francis Allard,Diimaajav Erdenebaatar, andChristine Lee. “A Xiongnu TombComplex: Excavations at Gol Mod2 cemetery, Mongolia (2002-05).”Mongolian Journal of Anthro-pology, Archaeology and Eth-nology (Ulaanbaatar) 2/2 (2006):1-21.

Miniaev and Sakharovskaia 2002

Sergei Miniaev and LidiiaSakharovskaia. “Soprovoditel’nye

zakhoroneniia ‘tsarskogo’kompleksa no.7 v mogil’nikeTsaram.” Arkheologicheskie vesti9 (2002): 86-118; in English as:“Sacrifice burials of the royalcomplex no. 7 at the Tsaraamcemetery” <http://xiongnu.atspace.com/Sacrif.htm>.

Miniaev and Sakharovskaia 2006

Sergei S. Miniaev and Lidiia M.Sakharovskaia. “Investigation of aXiongnu Royal Complex in theTsaraam Valley.” The Silk Road 4/1 (2006): 47-51.

Miniaev and Sakharovskaia 2007

Sergei S. Miniaev and Lidiia M.Sakharovskaia. “Investigation of aXiongnu Royal Complex in theTsaraam Valley. Part 2. TheInventory of Barrow No. 7 and theChronology of the Site.” The SilkRoad 5/1 (2007): 44-56.

Mission 2003

Mission archéologique française enMongolie. Mongolie: le premierempire des steppes. Arles: Actessud, 2003.

Navaan 1999

D. Navaan. Hunnugiin uv soyol:arheologiin sudalgaany material.Ulaanbaatar, 1999.

Rudenko 1969

Sergei I. Rudenko. Die Kultur derHsiung-nu und die Hugelgräbervon Noin Ula. Bonn: Habelt, 1969.

Shiji 1959

Sima Qian. Shiji. Beijing, 1959.

Torbat 2004

Ts. Torbat. Hunnugiin jiriin irgediinbulsh. Ulaanbaatar, 2004.

Volkov and Dorjsuren 1963

V.V. Volkov and Ts. Dorjsuren.“Hovd aimagiin Manhan suminnutagt ertiin sudlalin maltagahaiguul hiisen tuhai.” ArheologiinSudlal 2 (1963): 51-68.

35

Yeruul-Erdene 2004

Ch. Yeruul-Erdene. “Gol ModniHunnu bulshni sudalgaani zarim urdun.” Arheologiin Sudlal (Ulaan-baatar) 22 (2004): 76-109.

Notes

1. A detailed description of asquare tomb and its componentsis given in Miniaev andSakharovskaia 2006, 2007.

2. A comparison of these bronzesticks with others found in Xiongnuroyal tombs exhibits intriguingsimilarities, though their exactfunction is still under debate(André and Yeruul-Erdene 2004).

Carved stone erected after our exca-vations marking the completion of thereconstructed Tomb 64.

Excavation of a XiongnuSatellite Burial

Jessieca JonesLudwig-Maximilians-Universität, MunichVeronica JosephBoston University, USA

During the 2007 field season, theHovd Archaeology Projectexcavated several burials at theTahiltin-hotgor cemetery in Hovdaimag, Mongolia, at the foot of theAltai Mountains. A full report of theTahiltin-hotgor excavation isincluded in this issue of The SilkRoad (Miller et al. 2008).1

An earlier survey of the site inSeptember 2006 indicted thatthere are 132 graves which makeup the many complexes in thiscemetery (Miller 2006). Thenumber of satellite burials rangesfrom one to five per burial complexat Tahiltin-hotgor. The purpose ofthis report is to describe theexcavation of one of these satelliteburials and the artifacts foundwithin it. Burial THL-25-2 is one of thethree satellite burialsassociated with theunexcavated tombcomplex THL-25[Fig. 1]. The threesatellite burials, ofwhich THL-25-2 wasthe middle one, wereoriented in a gen-erally north-southline to the east of thelarge, square-ramped tombmound. We havechosen to report onburial THL-25-2 be-cause the skeletalremains were com-

pletely undisturbed and wellpreserved and thus the artifactscould be viewed in their originalcontext.

Grave excavation

After the vegetation, consisting ofscrubby desert bushes, sur-rounding the cluster of stones wasremoved, the loose sandysediment surrounding the stoneswas cleared to define the extentof the feature. When the surfacewas completely visible, arectangular grid was createdaround the stone cluster usingstring and iron pegs. This grid wastied to a reference datum point atthe large square tomb (THL-25)in order to allow us to record thelocation of the burial features in

Fig. 1. Drawing of THL-25 grave complex. All imagesare copyright © 2008 Hovd Archaeology Project.

36

three-dimensional space withregard to the square tomb and theother two satellite burials in thecomplex. A sub-datum wascreated for THL-25-2 (known assub-datum II) as a reference pointfor taking all measurementsrelating to this burial. Accuracy inrecording the features is veryimportant during excavation sothat the context of the artifacts isunderstood and drawings of theburial can be recreated oncefieldwork is completed. We tookmeasurements and photographsat every level of the excavation.

Within our excavation unit, thesediment around the stones wasremoved down to their bases,which represented the groundsurface at the time the feature wasconstructed. The artist of the teamthen drew the excavation unitusing a meter-grid. In order todraw the cross-section of theburial shaft, a level north-southreference line running roughlythrough the middle of the stone

mound was made with a string,and coordinates were taken withreference to sub-datum II fromvarious parts of the stones. Oncethis level was recorded andphotographed, we removed thestones and began excavatingbelow the surface through thehard-packed earth. We were verycareful at this stage so as not todamage any of the grave contents.The buckets of soil that weremoved were dumped outsidethe excavation unit so as not toclutter the working area, but nottoo far away because it was usedto refi l l the hole when theexcavation was concluded.

At a depth of approximately halfa meter (33-56 cm), we un-covered a second layer of stonesoriented in the same north-southdirection. The coordinates of thislayer were recorded and thefeatures photographed. Thesmaller stones were removed andwe continued excavating withtrowels toward the bases of thetwo large stones [Fig. 2]. As soonas we reached the bases of thestones, we created a pedestalroughly the size of the entire graveshaft so that we had enough roomwithin the excavation unit tocontinue working down toward theskeleton. The soil matrix within thegrave shaft was looser than thesurrounding, undisturbed ground.We were able to identify the extentof the grave by the change in soilconsistency and the presence ofvery hard, white veins of clay inthe undisturbed areas.

Human remains

There was no indication that theindividual had been buried in awooden coffin or stone cyst.Similarly, no wooden coffins orstone cysts were found in theother two satell ite burialsassociated with the THL-25 tomb

complex. As we excavatedbeneath the stones, we could seebits of bones sticking out, whichindicated that the skeletal remainswere directly underneath this layerof stones. After we had carefullytrowelled and brushed away thesoil, we came across three smallloose human bones. From then on,every trowel movement and brushstroke was painstakingly executedin order to make sure humanremains and artifacts wereexposed in their original position.The sediment around the threebones was carefully brushed awayand their positions documented insitu and photographed. Before anyof the bones or artifacts wasremoved, each item was assigneda number and then a clear plasticzip-lock bag was labeled with theassigned number. In this case, thefirst three bones were given thenumbers THL-25-2.1, THL-25-2.2and THL-25-2.3 (“THL-25” is thenumber assigned to the maintomb; “-2” is the number of thesatellite burial, and “.1” is thenumber of the item coming out ofthe burial). A log book was keptas the items were taken out,numbered, and bagged. Therecord in the log book had to bemeticulous because the infor-mation had to be entered into acomputer database back in ourfield lab. It was important toassign one person to record-keeping so that confusion ordiscrepancies could be clarifiedlater. After the three small boneswere removed and recorded, wecontinued working down towardthe rest of the skeleton andartifacts. These bones appearedto have been disturbed by rodentactivity, as indicated by smallrodent skeletal remains found atthis depth. As we finally exposedthe rest of the skeleton andartifacts, it became clear that this

Fig. 3. The THL-25-2 grave excavatedto the level of the skeleton, with thetwo large sub-surface stones still inplace.

37

grave had not been looted. Wenoted that the skeleton was wellpreserved and nearly intact.Preliminary analysis indicated thatthe remains belonged to a male.

In order to be able to makedrawings of the exposed skeletonas well as the cross-sectionof the grave, coordinateswere taken from variousparts of the skeleton alongthe same north-southstring line we createdearlier. Day glow plasticcocktail sticks were laid inthe burial pit at strategicpoints of the skeleton alongthis north-south line,keeping the northern pointconsistent for all latermeasurements; theircoordinates were recorded;and photos were takendirectly above the ex-cavated area. Later in thelaboratory, these photoswere used in conjunctionwith the Il lustrator©

program to make accuratescale drawings of theskeleton and cross-sectionsof the burial pit. Theconsistent northern point ofall such photograph-baseddrawings was used tooverlay the consecutiveimages in Illustrator© sothat their stratigraphicrelationship could beproperly understood.

After the exposedremains were documentedand photographed in situ asabove, artifacts that wereabove the skeleton were takenout, given a number and baggedso that the full extent of the burialposition and condition of theskeleton could be noted. Theindividual in this tomb was buried

in a supine position with legsextended [Fig. 3]. His body wasoriented to the north and his leftarm was bent across his chest.The skull was oriented facing eastand had been slightly deformed bythe weight of a large stone placedon top of it. The skeleton was

buried about a meter below theancient surface. All of theseobservations were again recordedand additional photographs madebefore the remains were removed,superficially cleaned with brushes,

labeled and bagged, and taken tothe ger (yurt) that housed thelaboratory equipment. In thelaboratory, the bones wentthrough another cleaning processbefore being securely packed fortransportation to Ulaanbaatar. Noperimortem trauma to the

skeleton was observed,which most likely rules outa violent death. Furtheranalysis by Christine Leerevealed that the interredindividual was an adultmale between 30-35 yearsof age and 159-167 cm tall.(Miller et al. forthcoming).

Artifacts

This burial contained moreartifacts than any of theother satellite graves weexcavated. After all theskeletal remains had beenremoved from the grave,the same process of takingcoordinates of the artifactswas repeated and anotherphoto with the day-glowplastic cocktail sticks wastaken. Correct photo-documentation of all theartifacts in situ is essentialfor post-excavation re-search.

Due to the climaticconditions in this area noorganic material such astextiles, wood, or leatherwere preserved at thisshallow depth. Howeverthere was some evidence oforganic residues around theskeleton. We recoveredseveral artifacts made ofbone and metal. The bone

artifacts were relatively well-preserved but all the iron artifactswere badly corroded. The boneartifacts included a plateperforated at both ends [Fig. 4,

facing page], approximately 16 cm

Fig. 3. In-situ skeletal remains.

38

long , found near the left side ofthe individual, within the bend ofthe left arm. The bone plate isslightly curved with a small holeon one end and a “T”-shaped pieceof metal inserted through theother hole at the opposite end.This plate may have been part ofa bow, of which the rest was notpreserved, or the handle of aninstrument. Similar bone anddeer-horn artifacts have beenfound at the site of Burhan-tolgoiin Egiin gol Valley, Bulgan aimag(Torbat et al. 2003). Near the leftside of the pelvis, we also found asmall, slightly flat object made ofbone that was rounded on bothends and perforated through themiddle [Fig. 5]. On one side, an“X” was incised into the surface,and the opposite side wasuntouched. This may have been atoggle used for fastening clothing.

Around the same area, a smallbone shaft that we had partlyexposed earlier was found to be

embedded in achunk of highlycorroded metal, justbelow the left pelvisnear the femur. Ourconservator, JudyLogan, devised amethod to dislodgethe bone artifactfrom the corrodedmetal with minimaldamage to theartifacts. Afterstudying the situ-ation, it was decidedthat a block-liftmethod would be the best optionto achieve this. The artifact clusterwas cleaned, then two layers ofcheesecloth were carefully placedon top of it. The cheesecloth waswetted so that it would conformto the surface of the object.Paraffin wax pellets were softenedin a hot-water bath and the semi-melted wax was slathered ontothe cheesecloth on all sides, likeicing a cake [Fig. 6]. The

cheesecloth acted as a barrierbetween the artifact and the rigidwax support and prevented themelted wax from sticking to theartifact. After the wax had set,an archaeological spatula wassliced under the dirt supportingthe wax-encased artifact toseparate it from the ground. It wasthen flipped onto a piece ofcardboard so that the artifact waslying on its wax-case. It wasconveyed back to the laboratoryfor further treatment.

On each side of the pelvis wealso found a pair of sheep or goatastragali (ankle bones). They weresmoothed in some areas, butotherwise undecorated. Near theright side of the pelvis, sitting ontop of a chunk of corroded metalwas a larger astragalus, whichcould be from a bovine or horse.Animal ankle bones were found insome of the other burialsexcavated at Tahiltin-hotgor aswell as in many other Xiongnusites throughout Mongolia andsouthern Siberia (see Davydova1995, 1996; Dorjsuren 1961;Miller et al. 2006). In modernMongolia, ankle bones are stillused as gaming pieces. The burialcontained many other fragmentsof corroded iron, but most of themwere too damaged to be identified

Fig. 4. (below) Bone plate,before cleaning, with inset(in different scale) showingend view with iron insert.

Fig. 5. (right) Bone toggle.

Fig. 6. Applying the wax over the cheeseclothin preparation for the block-lift.

39

in the field. However, several ofthe corroded iron artifacts wereintact enough so that they couldpossibly be identified in the futureusing x-ray and other techniques.

Located on the inside of eachankle of the deceased was a pieceof iron, which may have been partof a footwear buckle. Several otherfragments found around andunder the pelvis may have beenpart of a metal belt or fastenersfor clothing. A corroded metal ringwas found near the right hand.This could be a ring used byarchers to protect the fingerswhen drawing the bowstring.

Back in the laboratory, dirt fromthe underside of the block-liftedartifact cluster was carefullycleaned in order to expose thebone-shaft, while leaving thefragile corroded metal in thecheesecloth-wax case. Fragmentsof wood were detected at one endof the corroded metal, whichsuggests that it could have beenbe a knife or other metalinstrument with a wooden handle.Only further cleaning and othertests can reveal its form andconfirm what the object is.Meanwhile, the bone artifactrevealed itself as a pipe-shapedobject – a long, faceted piece ofbone with a small scoop carvedinto one end [Fig. 7, facing page].The shaft of the artifact was nothollow which rules out thepossibil ity of its being animplement used for smoking.There were no visible burn marksor residue inside the bowl of theobject. A similar artifact was foundat the Burhan-tolgoi cemetery inEgiin gol valley, north Mongolia(Torbat et al. 2003), and twosimilar objects have also beenfound in Sudzinsk in Il’movaia pad’(Rudenko 1969).

Discussion

The purpose of these satelliteburials is still being debated. THL-25 is a typical Xiongnu elite burialwith three satellite burials. Theartifacts buried in this satellitegrave are consistent with Xiongnusatellite burials of this size in othercemeteries, even though theburial structure of the satellitegraves at Tahilt is different (Milleret al. 2006; Miller et al.forthcoming).

The presence of traditionalclothing accessories such as ametal belt around the waist,footwear with buckles, andastragali is common in Xiongnuburials. The most unusual objectin this grave is the bone pipe-shaped object. It is interesting tonote that such objects have beenfound in the core Xiongnu areasas well as in the westernmost partof the Xiongnu Empire. This maypoint to some degree of culturalhomogeneity among the differenttribes which constituted thenomadic empire. The types ofmaterials used to produce theseartifacts (bone, metal, etc.) arealso similar to Xiongnu materialculture from other burials.

When we reviewed the humanremains and artifacts together, wewished to hypothesize this was awarrior not significantly past theprime of his life. However, theevidence about possible weaponrywas at best slender. He wasprobably dressed in organicmaterials (most probably wool orfelt) as indicated by organicresidues all around his remains.His garment was fastened with abone toggle and in Xiongnufashion he wore a belt made oflinked metal plaques. He probablywore felt or leather bootsdecorated with metal buckles and

carried his accessories, such as aknife, the pipe-shaped object, andastragali in a pouch of some sorton his left side just below hispelvis. The bone-plate might havebeen part of his bow, and heappears to have had an iron-bladed knife. It is not certainwhether a corroded metalfragment found near the left femuris an arrowhead. Connectionsbetween the material and humanremains found in the grave and thepast lifestyle and practices of theancient Xiongnu can be drawnthrough the context that intactburials and archaeologicalexcavation provide. In this and theother satellite burials at Tahilt,unlike at Gol Mod 2, we did notfind numerous iron arrowheads.

It is interesting to note that nocoffin or stone cyst was foundamong the three satellite burialsin the THL-25 complex. Thereasons could be due to variationsin local tradition, given that theTahiltin-hotgor cemetery is in thewesternmost part of Mongolia, orenvironmental constraints such asthe shortage of wood, or thestatus of the deceased. The exactdating of this satellite burial, THL-25-2, awaits further clarificationbut we can say confidently that itshares similar burial traditions ofother Xiongnu graves.

Acknowledgements

The authors would like to thankBryan Miller and Judy Logan fortheir comments and suggestionson earlier drafts of this article, theSilkroad Foundation, the NationalMuseum of Mongolian History andthe University of Pennsylvania.

About the authors

Veronica A. Joseph received herB.S. in Anthropology fromMichigan State University and is

40

currently a Ph.D. student in theArchaeology Department atBoston University. Her researchinterests include the Qin and Hanempires in China and theirinteractions along the Chinesefrontier with nomadic groups ofMongolia and Northeast Asia. Shemay be contacted at: <vjoseph@bu.edu>.

Jessieca Leo-Jones graduatedfrom the University of Canterbury,Christchurch, New Zealand with aB. Sc. in Chemistry and Economicsand obtained her M.A. in Sinologyfrom the Ludwig MaximilianUniversity (LMU), Munich,Germany. Presently, she isworking on her Ph.D. at the Facultyfor Asian Studies at LMU. Herinterests include Chinese Art andArchaeology, sex and sexuality inthe History of Chinese Medicine,and Xiongnu culture in relationshipto Central Asia. She may becontacted at: <jade.ys@o s t a s i e n . f a k 1 2 . u n i -muenchen.de>.

References

Davydova 1995

A. V. Davydova. Ivolginskiiarkheologicheskii kompleks. T. 1.Ivolginskoe gorodishche (TheIvolga Archaeological Complex.Vol. 1: The Ivolga Fortress).Arkheologicheskie pamiatniki

siunnu, Vyp. 1. St. Petersburg:Peterburgskoe vostokovedenie,1995.

Davydova 1996

A. V. Davydova. Ivolginskiiarkheologicheskii kompleks. T. 2.Ivolginskii mogil’nik (The IvolgaArchaeological Complex. Vol. 2.The Ivolga Cemetery). Arkhe-ologicheskie pamiatniki siunnu,Vyp. 2. St. Petersburg: AziatIKA,1996.

Miller 2006

Bryan Miller. “Khovd Archaeology:Archival Research and FieldSurvey Expedition,” accessed at:<http://www.mongoliacenter.org/f e l l o w s h i p / r e p o r t s / B r ya n .Miller.pdf>.

Miller et al. 2008

Bryan K. Miller, JamsranjavBayarsaikhan, TseveendorjEgiimaa, and Christine Lee.“Xiongnu Elite Tomb Complexes inthe Mongolian Altai: Results of theMongol-American Hovd Archae-ology Project, 2007”. The Silk Road5/2 (2008) (see above).

Miller et al. 2006

Bryan K. Miller, Francis Allard,Dimaajav Erdenebaatar andChristine Lee. “A Xiongnu TombComplex: Excavations at Gol Mod2 Cemetery (2002-05).” Mon-golian Journal of Anthropology,

Archaeology and Ethnology 2/2(2006): 1-21.

Miller et al. forthcoming

Bryan K. Miller, JamranjavBayarsaihan, Tseveendorj Egimaa,Prokopy B. Konovalov, Judy Loganand Christine Lee. “Elite XiongnuBurials at the Periphery: tombComplexes at Tahiltin-hotgor,Mongolian Altai” (forthcoming).

Navaan 1999

D. Navaan. Khunnugiin uv soyol:arkheologiin sudalgaany material.Ulaanbaatar, 1999.

Rudenko 1969

Sergei I. Rudenko. Die Kultur derHsiung-nu und die Hügelgräbervon Noin Ula, tr. Helmut Pollems.Bonn: Habelt, 1969.

Torbat et al. 2003

Ts. Torbat, Ch. Amartuvshin, andU. Erdenbad. Egiin Golyn savnutag dah arheologiin dursgaluud.Ulaanbaarar: Mongolian StatePedagogical University, 2003.

Note

1. Additional information aboutTahiltin-hotgor cemetery can befound on-line at the SilkroadFoundation website: <www.s i l k r o a d f o u n d a t i o n . o r g /e x c a v a t i o n /takhiltcemetary.html>.

Fig. 7. Bone “pipe-bowl.”

41

The Tahilt Region:A Preliminary Archaeological Survey of the Tahilt

Surroundings to Contextualize the Tahilt CemeteriesJames T. Williams

Beijing; University of Pittsburgh, USA

The survey component of the 2007Silkroad Foundation Tahiltexpedition provided informationwhich not only complements datacollected by the Xiongnu periodexcavations conducted by theMongol-American Hovd Archae-ology Project but may also standalone as a data related to broaderregional considerations. A largescale survey can provide both achronological and spatial contextfor particular sites, and the typeof survey described here isintended to be the first stage in amethodology that would includemore survey and more targetedexcavation, a methodology thathas been successful in other areasof Central Asia (Honeychurch,Wright and Amartuvshin 2007;Frachetti 2004). The survey of2007 was designed to answersome preliminary questions andassess the suitability of the regionfor higher intensity survey in thefuture. Some of the questions areas follows:• What types and kinds of sitesexist in the Tahilt region?

• What types of environmental orlandscape features are associatedwith archaeological features?

• What are the pre- and post-Xiongnu sites and how do theycompare to the Xiongnucomponent in the archaeologicallandscape?

• What should future surveys ofthis area look like and what arethe potential challenges?

Methodology

The survey was conducted over a40 square kilometer areagenerally to the southwest of theTahilt Xiongnu cemetery. Itconsisted of a systematic

pedestrian survey with transectsspaced at about 500 meters,terrain permitting. This regionallowed the survey to encompassa number of environmental andlandscape features such as rivers,high rocks, open spaces, andvalleys. The rationale for surveyby making large transects acrossthe region involved a number offactors, the first being how bestto understand the Tahilt regionquickly while at the same timehaving the resolution to detectsmall sites or single artifacts. Themethodology could be describedas high resolution but low density.By low density I mean that thesurveyed area represents what Iestimate to be a ten percentcoverage or sampleof the 40 sq kmregion (Honey-church, Wright andAmartuvshin 2007).A small area wastargeted for 100%survey coverage soas to prove that thetarget area was avery low densityarea and not lowdensity due to thesurvey methodology[Fig 1, C]. Futuresurveys would bedone at 100%coverage, as I willdiscuss below. GPScoordinates, sitetype and surround-ing environmentaldata have beentaken to establishthe context of a site.Site typology issimilar to that ofother regional sur-veys conducted innorthern and south

central Mongolia, in order tofacilitate comparison with theirresults (Wright 2006). Sites werealso categorized by the number offeatures. Features are singlearchaeological elements whichmake up an archaeological site.Recording the number of surfacefeatures, those elements that arevisible from the survey, can be auseful ranking tool indiscriminating between larger andsmaller sites.

What are the types and kindsof sites that exist in the Tahiltregion?

Both artifact–based andmonument–based sites indicatethat the chronology of the Tahilt

Fig 1. Overview of the survey region with smallerareas indicated. All images copyright © 2008 HovdArchaeology Project.

42

region appears to be continuousfrom the Paleolithic, roughly100,000 B.P. based on stone tooltypology, to the Turk period (ca.7th century CE). Evidence forspecific later historic periods suchas Khitan or Mongol at this timehas not been detected. For thispaper I designate as monumentalsites those with architecturalfeatures. I include burials and/ortombs but do not l imit thedesignation to these types: i.e.,all burials are monuments, but notall monuments are burials. Sincenone of them have beenexcavated, there is no evidence ofthe use of the monuments. Thissection will address the initialquestion of the survey concerningthe types and kinds of sites in theTahilt region and give a very briefdescription of them.

The Paleolithic and Mesolithiccomponent in the Tahilt region ischaracterized by large stonechopper and axe tool types [Fig2] generally in the areasurrounding the large spring andriver fed basin [Fig 1, A], nownearly dry. The well known Tsenkercave Paleolithic site is upstreamof this basin. So the location ofearly stone tools in this areaseems to relate to the river valley.

The Neolithiccomponent is char-acterized by smallerblade tools and morecomplex tool kits(Larichev, Pforr andChard 1962) and inthe case of Tahiltwould include scra-pers and retouchedbladelets [Fig 3].These were generallyfound in close prox-imity to Bronze Agemonuments andpottery [Fig 4], a factwhich suggests thatsites were used andreused over longperiods of time forhabitation and stonetool production.

Bronze Age sitesare both artifact sitesand monumental.The only potteryscatters found duringthe survey are fromthis period or aremodern or historicsherds which werenot collected. Sincenomadic sites lackn o n - m o r t u a r y

architectural andother extensivematerial remains,the pottery gen-erally would indicatea habitation sitewhich it is useful todistinguish from amonumental site.There are bothhabitation andmonumental sitesfor the Bronze Age,but this pattern doesnot appear for laterperiods. The monu-mental Bronze Agesite types found inthe Tahilt region arek h i r i g s u u rmonuments, slab

burials and other featuresassociated with those two, suchas satellite features that arecharacteristic of khirigsuurs. Akhirigsuur is a circular pile ofstones which may or may notinclude an interment. Thekhirigsuurs at Tahilt have not beenexcavated; so there is no measureof those with interments and thosewithout. However, the types andkinds of features recall khirigsuursfound in Central Mongolia, and thepercentage with interments isexpected to be generally verysmall, if it exists at all (Wright2006). The other characteristics ofkhirigsuurs are most often circularor square surrounds and small

Fig 2. Paleolithic chipped hand tools.

Fig 3. Mesolithic andNeolithic flakes bladesand bladelets.

Figs. 3, 4. Flakes and blades withassociated pottery.

43

satell ite features and/orrectangular pavements of stone[Figs. 5, 6].

Another site type generallycharacterized as Bronze Age isrock art. This would include anydepiction made by humanspainted or pecked into the stone(Jacobson-Tepfer 2006). At Tahiltthere are seven rock art sitesranging from a single panel of asingle animal to multiple panelswith a variety of depictions ofsheep or deer, which is within

normal expectations of rock art forthis area [Fig 7].

The evidence for the Xiongnuand the Iron Age comes mostobviously from the Tahilt cemetery(Miller et al 2008) [Fig 1 B], but inaddition there are three smallercemetery sites in the area,containing between six andtwenty-seven features/burials.Furthermore there are a numberof single Xiongnu burials outsideof the four main cemeteries and anumber of stone ring features that

could be char-acterized asXiongnu but mayr e p r e s e n ta n o t h e rch r ono l og i c a lperiod. The fu-ture excavationof these smallercemeteries andsingle burials willprovide morei n f o r m a t i o nabout their re-lationship to themajor cemetery.Outside of themajor cemeterycomplex no othersquare-rampedtombs have beendetected. The

burials in the other cemeteries areall circular in form [Fig 8 Map; Fig9, both, facing page].

The smallest component to thechronology of the Tahilt regioncomes from the Turk period withtwo sites including standing stonesand stone lines extending fromthem, as is typical of balbal linesfound in other areas of Mongolia.The construction of the standingmonument includes base supportstones generally square-shaped[Fig 10, facing page] and thus not

Fig 5. (above) Example of a Khirigsuur with a square surround.

Fig 6. (below) Selection of Khirigsuur types and features from theTahilt region.

Fig 7. Typical panel of rock art depict-ing sheep, goats and/or deer.

44

to be confused with the deer stoneor Bronze Age standing stonefeatures.

The post-Turk period inMongolia would encompass

medieval and historicperiods such asUighur, Khitan, andMongol. No evidenceof these sites orcharacteristic potteryfrom these periods(Wright forthcoming)was detected duringthe survey. It ispossible that thesample size is limitingthe finds. However,the location of Tahiltis far from the core ofany of the historicempires. In addition,periods such as theMongol might notappear, given thatthey are relativelyshort in comparisonto the major periodsrepresented at Tahilt.What the preliminarysurvey shows is a

continuous use of the landscapefor 100,000 years up to the Turkperiod; however, the use andunderstanding of this landscapewithin the survey area differs withthe culture.

What types of environmentalor landscape features areassociated with archaeologicalfeatures?

The connections between specificlandscape features are, at thispoint anecdotal. Howeverrepresenting them graphically ina GIS is the first step tounderstanding what environ-mental features are important tolocations of archaeologicalelements within the landscape. Inthe future the types of landscapefeatures will be systematicallycategorized to clarify therelationships between thearchaeological record and itsenvironment. In addition this willhelp better to target surveys forthe future.

The first of the majorenvironmental features of theTahilt region are the Hoyt Tsenkerand Dond Tsenker rivers. Thesetwo rivers come together at themodern town of Manhan. Also atthis river junction is a large sitewith over 40 khirigsuur monu-ments ranging in size from ten tothirty-five meters. So far this is thelargest site in the area with someof the largest monuments. Theother main feature is the highrocks that run between the tworivers. While the high rocks do notaffect the overall site locations, thesites are fairly uniform over theentire survey area. The larger

Fig 8. Locations of three or four secondaryXiongnu cemeteries.

Fig 9. Xiongnu burial ring from a secondary cemetery.

Fig 10. Standing stonesassumed to be from theTurk period.

45

sites, both Bronze Age andXiongnu, appear at the base of thehigh rocks and in the dry and rivervalleys [Fig 11]. The large basinmentioned earlier is to the westof the Tahilt cemetery; some of themajor Khirigsuur sites overlookthis basin. To the south of theTahilt cemetery is a large openvalley. This area containedevidence for modern use but wasalmost unused in the past. Asection of this area was selectedfor 100% total coverage survey,and only four sites were detectedwithin that area. Understandingthe relationship between majorlandscape features and thesurrounding archaeologystrengthens our understanding ofeach component of the chron-ology.

What are the pre- and post-Xiongnu sites and how do theycompare to the Xiongnucomponent in the archae-ological landscape?

The pre-Xiongnu component tothe Tahilt region is the most visible

of all the periods, with 40% of allthe sites dating to the Bronze Age,not including the rock art andassociated Neolithic sites. Therelationship between these twoperiods can begin to beunderstood even at the pre-liminary stages of survey. The keydifference between the BronzeAge sites and the the Xiongnusites is that there are no artifactor visible Xiongnu habitation sites,whereas in the Bronze Age wehave both, sometimes in closeproximity to the monumentalsites. We can understand this tomean that during the Bronze Agethe landscape was understood asboth a ritual and habitation area,but the Xiongnu were delineatinga separation between a ritual andhabitation space, thus creating aless homogeneous picture thanthe Bronze Age perspective onspace. This may be unique to theperiphery or be resulting from thelimitations of the survey method-ology. The evidence for the ritualuse of the landscape in the BronzeAge is very visible and would have

been visible to theXiongnu. The evi-dence that thelandscape was prev-iously used forhabitation is andwas less apparent.It is possible thatthis is part of thecriteria the Xiongnuwould have used inchoosing a locationfor the Tahilt ceme-tery. However com-parison of othermajor cemeterysites and the BronzeAge componentwould need to beexplored before wecan be certain. AtTahilt we can see theBronze Age mark onthe land, and laterthe Xiongnu use asection of thislandscape not cur-rently occupied withmonuments for theirown monuments

and ritual. These distinctions alsobring out the behavioraldifferences between the twogroups during these time periods.The khirigsuurs representmonument and ritual withoutburial. The Tahilt cemetery andXiongnu burials represent a ritualand mortuary site and behaviors(Miller et al. 2008). This can alsobe seen for later periods, wherewe have evidence for two Turkstanding stones but no pottery. Itis understood that two sites do notrepresent a significant sample, butfuture survey will easily eitherconfirm or deny this pattern.

Future surveys

The Tahilt region is interestingarchaeologically and ideal forfuture regional survey. However,there are some challenges ofwhich we became aware duringthe preliminary stages. Fullcoverage survey (that is, covering100% of the landscape) would beideal for the Tahilt region forovercoming some of thesechallenges. One of the challengesof survey in the Tahilt region wasthe absence of artifact sites, whichmakes it difficult to determine withany certainty the habitation andsettlement patterns for theXiongnu and Turks. A full coveragesurvey would detect more sitesand increase the sample size, or,if the habitation area is in fact notin the Tahilt region, full coveragesurvey would detect this. Thesame could be said forunderstanding the Bronze Agecomplexity of the region. At thispoint the survey indicates wheresites seem to be located but is notrepresentative of the emergingspatial and cultural complexitystudied during this period (Allardand Erdenebaatar 2005; Wright2006, 2007). Expanding thesurvey to include more en-vironmental zones would alsoincrease the likelihood of detectingnomadic habitation sites for theBronze Age, Xiongnu and laterperiods.

The other aspect that wouldneed to be considered for future

Fig 11. Location of larger sites or those with ahigher feature count.

46

surveys would be to locate sourcematerials for lithics and pottery.Since the site locations seem tobe related to certain landscapefeatures, future surveys wouldalso want to consider the otherelements that might affect thisdecision-making process. Inaddition, source material studieshave been useful in other parts ofMongolia to prove whether potteryis locally produced or brought fromother regions (Honeychurch,Wright and Amartuvshin 2007). Inthe case of Tahilt, since thehabitation record for the Xiongnuand Turks is limited and/or non-existent, pottery sourcing couldhelp in understanding thelocations of habitation. Simul-taneously surveying the palaeo-environment would clarify theenvironmental landscape andprovide other layers of data toconsider.

The Tahilt region has thepotential to answer a number ofinteresting research questions.One single survey or excavationis not enough to characterize anentire region, but, with continuedinterest in the area, a clearerpicture will emerge regarding thelandscape in which the Xiongnuand other cultures of Central Asialived. The study and survey of thisregion will no doubt continue.

Acknowledgements

I would like to thank Bryan Millerfor giving me the opportunity toconduct this survey as a part ofthe project’s larger work with theMongolian National Museum onthe periphery of the Xiongnu, theSilkroad Foundation for itssupport, Josh Wright for his adviceand comments, and the volun-teers of the 2007 season for theirhard work.

About the Author

James T. Williams is currently aLecturer on culture and history atthe University of Chinese Medicinein Beijing. He received his degreefrom the University of Mass-

achusetts, Boston, in Archaeologyand Art History, and will enter thePh.D. program in Archaeology atthe University of Pittsburgh inautumn 2008. He has worked onsix different archaeologicalprojects in Mongolia, China,Jordan and the U.S. He isinterested in the Neolithic andBronze Age emergence of socialand spatial complexity in Chinaand Mongolia. He may becontacted at <jamesw1220@gmail.com>.

References

Allard and Erdenebaatar 2005

Francis Allard and DiimaajavErdenebaatar. “Khirigsuurs, Ritualand Mobility in the Bronze Age ofMongolia.” Antiquity 79 (2005): 1-18.

Fitzhugh 2006

William Fitzhugh, ed. American-Mongolian Deer Stone Project:Field Report 2006. WashingtonDC: Arctic Studies Center, 2006.Available on-line at: <http://www.mnh.si.edu/arctic/html/pdf/2006%20Mongolia%20Field%20Report%20small.pdf>

Frachetti 2004

Michael D. Frachetti. “Archae-ological Explorations of BronzeAge Pastoral Societies in theMountains of Eastern Eurasia.” TheSilk Road 2/1 (2004): 3-8.

Hall et al.1999

Mark E. Hall et al. “ChemicalAnalyses of Prehistoric MongolianPottery.” Arctic Anthropology 36(1999): 133-150.

Honeychurch, Wright andAmartuvshin 2007

William Honeychurch, JoshuaWright and Chunang Amartuvshin.“A Nested Approach to Survey inthe Egiin Gol Valley, Mongolia.”Journal of Field Archaeology 32(2007): 339-352.

Jacobson-Tepfer 2002

Esther Jacobson-Tepfer. “Petro-glyphs and the Qualification ofBronze Age Mortuary Archae-ology.” Archaeology, Ethnology,

and Anthropology of Eurasia 3(11) (2002): 32–47.

Jacobson-Tepfer 2006

Esther Jacobson-Tepfer. “The RockArt of Mongolia.” The Silk Road 4/1 (2006): 5-13.

Larichev, Pforr and Chard 1962

V. E. Larichev, Barbara L. Pforr andChester S. Chard. “The MicrolithicCharacter of Neolithic Cultures inCentral Asia, Trans-Baikal, andManchuria.” American Antiquity27/3 (1962): 315-322.

Miller et al. 2008

Bryan K. Miller, JamsranjavBayarsaikhan, TseveendorjEgiimaa, and Christine Lee.“Xiongnu Elite Tomb Complexes inthe Mongolian Altai: Results of theMongol-American Hovd Archae-ology Project, 2007.” The Silk Road5/2 (2008) (see above).

Wright 2006

Joshua Wright. “The Adoption ofPastoralism in Northeast Asia:Monumental Transformations inthe Egingol Valley, Mongolia.”Unpublished PhD dissertation,Harvard University, 2006.

Wright 2007

Joshua Wright. “Organizationalprinciples of KhirigsuurMonuments in the Lower Egiin GolValley, Mongolia.” Journal ofAnthropological Archaeology 26/3(2007): 350-365.

Wright forthcoming

Joshua Wright. “A Guide to theCeramics of North and CentralMongolia and the Gobi Desert”(forthcoming).

47

Food as Culture: The KazakhExperienceAlma Kunanbaeva1

Stanford University, USA

The Kazakhs have an amusinganecdote often told at the table:

Once a Kazakh came to townto buy food for an upcoming toi(feast) celebrating theexpected birth of a child. Hemet there his Uzbek friend.The Uzbek was delighted at thechance to display his hospitalityand invited the Kazakh homeas a guest. It was impossibleto refuse, and the Kazakh,even though he was in a hurryto return home, agreed. TheUzbek brought him home andbegan to entertain himgenerously. There was flatbread, grapes, honey, driedfruits, unusual sweets and agreat deal of green tea.(“Thank God, it’s hot,” theKazakh thought, “but it ’sso...pale, it hasn’t steeped verylong.”) At the end of the daythey brought in a plate ofgreasy pilaf. (“Why on earth isit made with raisins and nuts— maybe they don’t haveenough food?” the Kazakh saidto himself.) After the meal theKazakh, to the surprise of hisUzbek hosts, stayed for thenight. The wish of the guest islaw. In the morning, they hadtea, and again at the middaymeal; in the evening theyserved soup. (“Fairly tasty, buttoo full of potatoes.”) Andagain the Kazakh stayed for thenight, although for somereason he began to worry. Thehost was already unhappy thathe had invited him home.Groaning, he went off to themarket to buy something forthe table and in his heartlamented that he simplycouldn’t get rid of the Kazakh.But a wise man said, “Did youfeed him?” “Fed him.” “With

what?” “Everything!” “Did yougive him beshbarmak?” “Whaton earth is that?” “Go buy asheep, boil it whole in water,dump it in front of him, andhe’ll leave.” The Uzbek had nochoice; so he did it. And hehad no sooner served themeat, when the Kazakh beganto hurry, washed his hands,thanked everyone, tasted themeat, drank the broth andimmediately began to say hisgoodbyes. “I apologize, friend,but I’m in a rush — it’s possiblethat my wife has already givenbirth. I simply didn’t realizethat your pasture is so far fromthe city. Well, so be it.Farewell.”

When publishing this tale, oneought to give it a title — forexample, “The Dumb Kazkah,” orthe “Dumb Uzbek,” or even “Twoworlds — Two Civilizations”….2 Amore academic title would beappropriate: “The Role of StewedMeat in the Hospitality Ritual of theKazakhs,” “ On the Question of theSemiotics of Food,” etc., etc.

All of the well-known literatureon the ethnography of nomadsstarts with the premise about theirbeing deprived of any agriculturalproducts, and, in connection withthat, the impoverished nature ofthe nomadic diet. It is normal tostate laconically that the basis ofthe nomad’s diet is meat and milk— that is, those things which canbe obtained directly from theiranimals. Meat and milk productsare distinguished for their highprotein content and indeedmaintain the energy of individualsengaged in heavy physical labor.However, the diet of the nomad infact never was limited to them.One need but recall the numerous

observations of travelers con-cerning the abundance of festiveentertainment at weddings andmemorial feasts of the nomads,and also the rich materials offolklore. The portrayal of elaboratefeasts is infinitely varied andevocative, for example, in theKazakh epic “Koblandy-batyr”, theKyrgyz epic “Manas”, and in anyother epic tales of the CentralAsian nomads.

One of the basic reasons forsuch misunderstanding regardingthe diet of the nomads is, if youwish, historical and ecological. Letme explain. The fact is that thefood values of the traditional(steppe) and contemporary(farmed) products of animalhusbandry hardly bear com-parison. Above all this isconnected with the nature of thevegetation cover of the pasturesand the grazing pattern of theanimals. Even today, the qualityof the sheep in Kazakhstan andthe USA, for example, cannot becompared. Here is an example.

In 2002 in Washington a Kazakhdelegation of 18 people, invited toparticipate in an internationalcelebration of the countries of thehistoric Silk Road (The Silk Road:Connecting Cultures, CreatingTrust. The 36th Annual Smith-sonian Folklife Festival On theNational Mall, Washington, D.C.),despite the perfectly satisfactoryfood both in the hotel and at thefestival grounds, decided to find away to eat “normal” (that is,traditional) beshbarmak3 — theKazakhs’ favorite meat dish. Ourhospitable American friendsstraight away found a farm locatednot far from the capital, took usthere, allowed our experts toselect the best sheep, slaughterit on the spot and themselvesbegin to prepare it. The entireprocedure was videotaped.Everyone was bursting withenthusiasm and taking pains toensure the best possible outcome,but in the end they weredisappointed. The meat wasdifferent. The conoisseurs quicklydiscovered the chief reason: the

48

grass in that area was too watery,resulting in meat that was“watery,” lacking in flavor, and“porous.” In addition, the sheep,which for the most part werepenned up, had not developedenough muscle mass. OurKazakhs carefully considered allthe factors. Thus it became clearthat even today people can tell alot from the taste of meat: thesex of the animal, whether it wasfree-range or penned up, whetherit was a lamb or a mature sheep,and how exactly it had beencooked.

In traditional society, the sheepto be slaughtered was first shownto the guest, then laid on its leftside, its legs tied in pairs, its throatcut without injuring the spine, andthe blood drained. Then anincision was made in the skin andthe carcass “undressed” — theskin was peeled off by hand,starting with the abdominal cavity,then the legs and finally the spine.The abdomen was cut up, thestomach and innards carefullyremoved and cleaned off to theside. The stomach was emptiedof its contents, turned inside out,washed and then the liningscraped off, care being taken notto harm the walls. The intestineswere washed, thrice stretched onthe spout of a tea kettle with coldwater and then turned inside out.The small intestine (ashy ishek —“sour tasting intestine”) wascrushed and boiled together withthe meat, as were the kidneys (atreat for children). The animal’slungs were not highly valued asfood, but were important in thehealing practice of the shaman orbaqsu. Held by the windpipe, theywould be used to flagellate the illperson in order that the diseasebe transferred to the sheep’slungs. Then the lungs would eitherbe given to the dogs or buried.There was one other way of usingthe lungs, in what we might callsilent speech, a traditional way ofexpressing an unspoken reproach.In such a case, a piece of boiledlung would be served to a guestas though by oversight, acci-

dentally, “the mistake of someonewho was helping,” and thusexpress a sense of insult. (Hereis a case of a play on words usingabsolute homonyms—lung, okpe,and insult, okpe). The guest wasthen expected to explain how hisaction in the past or unfortunateconduct had insulted the hosts andinsofar as possible correct hismistake.

The slaughter of animals,carried out according to Kazakhcustom (in contrast, for example,to the Mongolian), requiresmaximum removal of blood: theblood should be carefully collectedand buried in a safe place. It isthought that blood, if it remainsin the tissue, is the first thing tospoil. The division of the fleshrequires a knowledge of animalanatomy, since all bones need toremain unharmed. In the case ofa ritual meal, the bones would becarefully collected and buried, inorder that the numbers of theherds be restored. Meat whichremained on the bones is slicedand salted, whereby, according toKazakh belief, the curing, theripening of the meat supposedlycontinues. It would then be driedfor at least one night, with the aimof removing any moisture, firmingit up and concentrating its flesh.The head and extremities areseared on an open fire, the innardswashed and used to make kuyrdak(boiled innards with onion), a dishthat can be quickly prepared. Theremaining meat is divided into 12muscle groups, parts whichcorrespond to the 12 parts of thehuman body or the 12-yearcalendrical cycle. Each part has itsown social status and significancein the subsequent sitting. Eachguest, depending on his place inthe social order, is obliged toreceive his share. To the son-in-law, for example, always goes thebreast (tyos) , and to the youngbride the kuiymshak (the lastvertebrae).

The head of the sheep, whichhas been cooked separately,occupies a special place in themeal and is given to the oldest or

most honored guest. Beforeserving it, the lower jaw is cut offand the host makes a cross-shaped cut on its forehead. Theperson who receives the headmust share with all those present,accompanying his action withformulae of good wishes. Thus,they give out the ears and eyes tothe young (so that they learn tolisten and be observant), thetongue and palate to the poets (asa symbol of eloquence) and to thesingers (as the emblem of asilvery, virtuosic, and ringingthroat), and pieces of the singedskin (kuika) to the others. Brainsare not given to the young, sinceit is considered that only the wiseare worthy of that honor. A dishprepared from finely choppedbrains mixed with meat and dough(mipalau – pilaf of brain) is givento toothless seniors, in the beliefthat this mixture is especiallynutritious.

It is clear that the ritual has amagic function, in the course ofwhich, by partaking together, asingle social group is created andits structure made apparent.

A special place in wedding ritualis occupied by a dish around whichis distributed boiled and thinlysliced tail fat and liver (kuiryk-bauyr), where the alternation ofthe dense and dark with thetender and light symbolizes thepaired opposites of male andfemale, and the liver, as the organwhich generates blood, sym-bolizes kinship and birthing. If, atthe time of the betrothal, bothparties to the union receive thisdish, then the betrothal isconsidered to be legally bindingand a subsequent refusal by oneof the sides to marry will creategreat unpleasantness.

During the wedding, when themovement of the young couple isaccompanied by showering themwith grain and sweets, at theentrance to the yurt or house ofthe groom the bride feeds the fire— that is, throws on fat or pourson it melted butter. While doingthis she invoked the goddess

49

Umai, the protector of women inchildbirth, whose name has thesound of the word mai (fat), andasks a blessing on her married life.In this way the bride joins the kinof her husband, since the burningof the butter also symbolizes thefeeding of the ancestors of thegiven clan.

Having crossed the threshold,the bride meets her in-laws, who,holding hot, freshly-baked flatbread in the bosom, embrace thebride. In this ritual the breadsymbolizes the placenta, and thewhole procedure the making of thebride into a daughter.4 TheKazakhs say that a bride selectedby god is created at birth from theremains of her future mother-in-law, who then discovers as a resulta closer relationship with her.

A no less magical substance isthe dish zharys kazan (literally, a“cauldron-contest”), in whichmeat is heaped up and put to boilat the moment when labor inchildbirth begins. The readiness ofthe dish, the completion of itscooking, is supposed to influencethe readiness of the fetus to beborn and facilitate a natural andsafe birth.

Apart from mutton, Kazakhsuse beef, goat, camel and horsemeat. The last of these is the mosthighly valued. Only specially fedand tethered mares areslaughtered, ones which havenever been ridden or used in otherwork. The special regime offeeding such mares is called zhylkibailau (the tying of the mare), andthe slaughter of the mare in earlywinter, sogym. Sogym becomes aholiday for the whole community,each member of which receives aportion. The first meal after theslaughter is one in which allmembers of the communitygather at one table (dastarkhan).

In a somewhat analogousfashion to the white and dark meatof a Christmas turkey, horse meathas dozens of parts which differin taste, texture and use. Thus:kazy, whole, salted and cured ribsin a casing of intestine; shuzhyk

— a sausage, in which largechunks of meat and fat aresprinkled with salt, pepper andwild (nowadays ordinary) garlic;karta — the tenderest part of thelarge intestine which has beenturned inside out; zhal — the veryjuicy part under the mane; zhaia— the firm, cured and dried softpart of the haunch.

The techniques of preparationof horse meat include salting,marinating, and also drying andcuring with juniper smoke.Depending on the time taken foreach procedure and thecombination of methods,innumerable taste combinationsmay be imparted to the endproduct, horse meat. This productof itself has endless variety,depending on the type of pasture,feed, the age of the animal, etc.etc. From this follows thecombination of different kinds ofmeat into one dish, in a wayanalogous to how seven notes andthree chords provide the basis forall European music of the lastthree centuries. Thus, Kazakhsperceive a symphony of aromas,tastes, colors and textures in themyriad embodiments of a singledish — et, which is known in theliterature as beshbarmak (literally,“five fingers”). It is served inconjunction with dough which hasbeen rolled out and boiled in brothand on which are placed pieces ofstewed meat which has been cutinto large chunks. Poured over thiscombination of dough and meatis an onion sauce, tuzdyk, whosequality determines the taste of theentire dish. Rings or half rings ofjuicy, crisp onion are covered withfat skimmed from the broth,salted, peppered and thensimmered on a slow flame.Heaven help the person whoovercooks the onion — it will allbe spoiled! The onion shouldretain its crispness and having justbegun to secrete onion juice,replaces it with concentrated brothand becomes transparent. Onlyonce have I tasted French onionsoup which could be comparedwith the Kazakh tuzdyk sauce.

Tuzdyk can even rescueunsuccessful meat. For example,if the meat had been stored toolong in the freezer, it is possibleto freshen it after the initial boilingby braising it in tuzdyk.

The quality and the texture ofthe dough also has an infinitenumber of variations and dependsat very least on five basic factors:1. on the type of flour (finely orcoarsely ground, from milled orwhole grain, from hard or softwheat, home-made or com-mercial); 2. on the nature of theliquid (whether the dough is mixedwith icewater, warm or hot water);3. on the amount of salt (lightlysalted or saline solution); 4. on thepresence or absence of eggs in thedough; and, 5. on the amount ofkneading (soft or firm dough).

In the preparation of the dough,its feel is extremely important.While working it, one paysparticular attention to the quantityof liquid ingredients, adding flouras needed and achieving therequisite firmness and springinessby comparison with human skin —for example, they say aboutdough: it is light and soft like achild’s cheek (or a baby’s behind),or it is springy and firm like thebreast of a nursing mother, etc.The mixing of the dough alwaysalternates with letting it rest —until, as they say, the dough is“sated” and ready to be rolled out.It is rolled out with a long rollingpin of narrow diameter, to thepoint where the dough forms acircle of the thickness andtransparency of a sheet of paper.Such dough when boiled will notfall apart nor will it stick to itself.

Such tender and white doughgoes well with tender and light-colored mutton. Camel and horsemeat, on the other hand, requirea darker and thicker dough, piecesof which are cut up into smallerportions and boiled longer so thatthey absorb the broth.

Beshbarmak is served onseveral dishes — in earlier timesflat wooden ones, and nowenameled or ceramic ones — so

50

that one dish will serve three tofour individuals of the same agegroup. Each dish has a namedepending on which cuts of meatare on it. For example, bas tabakis the main dish, zhambas tabakhas the cuts from the abdomen,k’iuieu tabak is the dish with thebreast for the son-in-law, etc.

Beshbarmak is normally eatenwith the right hand. Eating withthe hand allows one not only tosense the temperature of the food(and thus be certain that themouth never will be burned) butalso assemble and combineaccording to taste the dough, apiece of meat and some onion intoan appetizing portion. Moreover,while sitting on the floor, it isdifficult to handle a fork with layersof dough and pieces of meat on ittaken from the gravy withoutdripping on the table cloth, the rugand one’s own clothing. In usinga fork, it is necessary to hold outthe left palm and inevitably scorchit with dripping gravy, which canthen only be licked off or endlesslycleaned with napkins. Kazakhs areconvinced that the right hand isespecially suited to eating thisdish. There is even a specific joke.They ask: “Why is the thumbshorter than the rest of the fingersand off to one side?” And theyanswer: “So that it will be easierto shape the food into a portion,lift it to the mouth and then placeit in the mouth with a singlemovement of the thumb.”(Kazakhs know that it is especiallyimportant to have a thumb wheneating pilaf, since rice easily stickstogether in a lump.)

The traditional concluding ritualof the meal is now becoming buta memory — asatu, when theelder feeds with his own hands andfrom his own dish all the youngunmarried participants in themeal, thus bestowing on them ablessing and wish for long life andsharing his social status.Children, who compete for theright before the meal to pourwater on the hands of guests forwashing and then offer a towel,

await asatu as a reward andpromise of good fortune in life.

At the end of the meal theyserve hot broth (sorpa) which,according to the elders, con-tributes to easier digestion of themeat.

The meal, which begins with aninvocation, concludes with theritual bata beru (the bestowing ofblessings) and dastarkhan kaiyru(closure, literally the “return” ofthe dastarkhan), after whichpeople engage in relaxedconversation and music making.

Any leftover food is taken awayand sorted, the onion removed,the dough and meat storedseparately. The broth is poured off,the bones cleaned and everythingleft for the next day. Some peopleeven prefer day-old beshbarmak,when it tastes so good to eat thecold meat with traces of aspic fromthe broth in its folds along withhot bread. Alternatively, one cancut the cooled and hardeneddough into narrow strips, fill it withsmall pieces of meat, and heat itin a little broth to prepare a specialdish, naryn, or fry the dough inbutter and eat it separately. Whendiluted with water, the con-centrated broth serves as the basefor soups called kespe, containingfreshly rolled out dough which isthen wrapped into a roll and cutinto narrow strips.

I have always been struck bythis endless transformation andreworking of one dish into others.The basic principle is completelywaste-free production. Literallyeverything went into the food.Even the congealed fat removedfrom the broth was melted andused for frying or in making doughfor bread. This fat, like butter, maybe stored for a long time in a driedsheep’s stomach, from whosewalls it is believed some beneficialsubstances are absorbed whichimpart medicinal qualitities. Thenthe fat may be used to treatillnesses of the skin and joints.

Just remembering the quantityof cholesterol I consumed in the

past now makes me shudder. Butnowhere and never have I seensuch strong, healthy and hardypeople as I did in my childhood.Anyway, that already is anotherlifetime…

I have heard tell of specialdishes which were prepared onlyby men when hunting. Along withordinary meat cooked on a spitover an open fire, my imaginationis stirred by so-called esip. Eldersinsist that the size of the stomachof any being equals the volume ofits soft tissues. Thus the stomachof captured game would be stuffedwith its meat (with salt and wildedible herbs), sewn up, buried insand and a fire lighted over it. Themeat slowly braised literally forhours. The walls of the stomachturned into a solid container, insideof which the aromatic and softmeat stewed in its own gravy.Each time I heard these tales Iregretted that I had not been borna boy and thus could not gohunting.

Among the mythologicalhunting tales connected with foodis the story about the great huntsof the past when a giant wild bullwas cooked on a spit. It sup-posedly happened in this way. Thevarious kinds of game weredressed, the heads and ex-tremities removed. Then theybegan to place one animal insidethe other in a specific sequence.First they sewed a pigeon(kogershin) into a steppe grouse(ular), then the grouse was sewninto a goat (eshki) the goat into awild sheep (teke), and the sheepinto a bull (buka), which had notbeen skinned. And all thistogether was roasted slowly allnight over the fire, having beenhoisted on a truly gargantuan spit.Toward morning, the bullpractically was burned to a crispand came apart like a ceramicdish, its contents being cooked toperfection and ready to eat.

The confused details of thistypically male dish — detailswhose symbolism already hasbeen half forgotten — such as the

51

selection of animals, their number,the cutting apart of the joints ofthe animals and birds before thestart of the preparation process,and in an even greater degree —the lighting of the fire and the allnight vigil until dawn — connectin certain ways with an analogousritual of preparation of nauruzkozhe (a new year’s soup) bywomen in the night of the springequinox (March 21-22, the Nauruzholiday). Women cook soup allnight in a huge hemisphericalkettle, rythmically and withoutpause tracing the path of the sunin the way they stir it with a paddle(pispek) (from the verb pis —cook, to be cooked, to stick into,to stir up with a paddle). Thepaddle is a stick with a cross-pieceat the bottom (analogous to thestirring stick used for agitatingkumiss). In addition to the heapedup meat, generally they put intothis soup five different grains —millet, oats, wheat, rice and corn.However, the selection ofingredients may vary dependingon what is available (for example,beans may substitute for corn),and even the inclusion of waterand meat in the number ofcountable ingredients may beoptional. Only women of child-bearing age may participate in thisritual. At dawn, they give to theelders the separately cooked headof an ox (who cooks it, where, andhow are not clear from thestories), and then all eat nauruzkozhe. There is an obviouscorrelation in the symbolism ofthese two somewhat parallel ritualdishes, the hunters’ and that fornauruz. They both are preparedat night, in both there must be fiveingredients. However, the one, formen, is distinguished by the fixednature of its components, even asthe second is marked by theirfluidity. Whereas the first isentirely restricted to the closedmale society of the hunt, thesecond involves exclusivelywomen, although in the finalreckoning it is in principle open tothe whole community.

To tell about meat products may

create the impression that theKazakhs eat only meat, literallyfrom morning to late at night.Meat is indeed consumed by allmembers of a Kazakh family, theonly exceptions being for pregnantwomen, who should not eat camelmeat or certain parts of a sheep.But in general to characterize theKazakhs as exclusively meateaters would be a mistake. Freshmeat is prepared primarily on theoccasion of receiving guests; inthe daily diet, they rely on grainand milk products.

Unlike with meat, whose basicqualities to a considerable degreeare predetermined by natural andseasonal conditions, productsmade from flour are the mainstage on which the cook maydemonstrate her talents. There isa wide spectrum of breads andother dishes made from flour,ranging from kuimak (a thin blin,rather like a French crêpe), cookedon a flat frypan and made of liquidbatter mixed from whey andintended for immediate con-sumption, to flat breads baked oncoals in two frypans, which holdbetween them the sourdough.

Even today the most popularfood made with flour is balls ofdough fried in boiling oil (to adegree similar to small donuts).The so-called baursak or kyikshaare not round but cut from rolledout sourdough. Leavening doughwith commercial yeast is a ratherlate invention. Instead, peopleused as a leavening ashytky, theremains of the previouslyprepared bread. The art of makingthe initial ashytky was kept secret,although it is possible to developa certain general understandingabout the stepwise process ofcreating such a sourdough starterat home. The first stage is theliquid leavening (ashytky kamyr)with salt in sour milk; the secondstage is the fermentation of thisleavening in a warm place; thethird, the feeding of the leaveningwith sugar or honey; the fourthstage is an inelastic dough, mixedwith flour until it is about halfprepared (but still sticky); the fifth

stage is the addition of flour andthe mixing of a soft, non-stickingdough. At that point part of thedough is set aside in a coveredcontainer (best of all one made ofclay or ceramic) as a reserve, andfat is added to the remainder, afterwhich the dough is formed intoloaves. The leavening remainsalive for 7-10 days. Generally it isabout then that it is time to bakemore bread for the family.

The attitude of Kazakh womento the ripening dough is indeedmaternal: it is protected fromdrafts, from sharp drops intemperature, even from un-necessary agitation. It is commonto cover the dough with a speciallysewn blanket. It is even forbiddento praise the rising dough, so asnot to jinx it, etc.

The cult of ancestors, which isstrictly observed by Kazakhs,requires regular preparation ofshelpek or zheti nan (sevenbreads), round, sourdough flatbread fried in fat. (The numberof these breads should be amultiple of the magic numberseven.) This is to be done onThursdays, or when one hasdreamed about deceasedrelatives, or in connection with avisit by guests who knew andloved the deceased. It is believedthat the spirit of the ancestors(aruakh) is fed by the aroma orsmoke from the frying of shelpek.

The best recipe for shelpek isto mix into soft dough one cup ofover-ripe, even bubbling, sourcream, which had been leftovernight in a warm place, alongwith a pinch of salt, a sprinklingof sugar and two pinches of soda.The dough made in that way iscovered and left in a warm placefor an entire day. Then in theevening one must carefully andtenderly knead it and divide it intoseven portions, roll out the ballswith a large rolling pin and letthem rest, then fry them on bothsides in a deep frypan filled halfway with vegetable oil. It isimportant to have the fat at theright temperature, which can be

52

determined by tossing into the oila bit of flour, which should sizzlebut not burn. Fat that is too hotchars the bread and does not allowit time to breathe; insufficientlyhot oil begins to absorb into thedough and makes the bread toogreasy and heavy. The breadshould be enjoyable and light—itis eaten hot, while declaring“Kabyl bolsyn” (“May it beaccepted!”).

The process of the stepwisetransformation of the end productcan be seen especially clearly indairy dishes.

The Kazakhs use milk from alltypes of animals in varyingcombinations. Nowadays themajority of milk products derivefrom cow’s milk, since it is themost readily obtainable. Kazakhsobserve a single rule for all typesof milk: before beginning anypreparation, the milk has to bescalded. Otherwise it is consideredraw, unripe, and not brought tolife by heat. Of course this reflectspopular experience, since raw milkmay contain harmful bacteria.

The cream is removed from theboiled milk and butter made fromit in wooden churns.5 At one timethe butter was preserved in thestomachs of sheep, in which it notonly did not become rancid, buteven breathed through the poresof the skin, absorbing healingenzymes from the tissue of thestomach. Over time such butterbecame as well a healingointment, which helped in skindiseases and in healing wounds.

In working with milk, theprimary task is the gradualelimination of superfluous liquid.Here again we have an obviousprocess of thickening, con-centration, and condensation —the change in the consistency ofthe end product, achieved bymeans of fermentation. Sourcream is added to warm milk (thetemperature level is easilychecked — the hand should feelneither heat nor cold), mixed, andthen wrapped in a specially sewnfelt cover or woolen scarves and

blankets and left to fermentovernight. It is essential not to letit cool or to shake it. By morningthere forms in the pan a productwhich compares with thick yogurt,which Kazakhs call katyk. Itsconsistency is such that when aspoonful is removed, whey quicklycollects in its place. If the katykis passed right away throughcheesecloth, the remainingcoagulate forms so-called suzbe,from which all the acidity has beenremoved with the whey. Thisdelicacy is a special favorite ofchildren and elders. In order tostore suzbe, one must suspend itand allow the remaining liquid todrain, then mix it with flour andsalt which have been roasted,form it into balls and dry it on theroof of a porch in the sun. Theresult is tushy-kurt — mild cheese,usually rather high in fat contentand easily chewable. If katyk isstirred, so as to activate theprocess of fermentation byintroducing air, and over the next10-15 days new katyk is added,the result is a different product —so-called airan, which one candrink as it is or, if diluted by halfwith water, make into a drinkcalled shalap. When airan hasbeen drained, salted, rolled intosmall balls and dried in the sun,one gets ashy kurt — salty sourcheese, which then is broken intopieces and will keep as long as onewishes. Kurt which has beenbroken up may be added to broth,which, in addition to sourness,acquires healing qualities and thusis used to treat head colds. (Sinceit makes one sweat, the brothmade from kurt eliminatescoughing and fever.) Airan also isdiluted with sheep’s milk, whichthickens in this mixture (two kindsof milk) and adds lightness as ifof beaten eggwhites, the resultingdish known as akta.

Even the whey which remainsafter straining, rich in proteins andminerals, is not thrown away.When boiled down to dryness in acauldron, it forms an orange-colored grainy mass, irimshik,which when mixed with melted

butter creates a new delicacy —kospa. (The more delicate irimshikis obtained from fermenting ofsteamed milk in a calve’s stomachand by subsequently boiling itdown, although this technique isalready becoming a thing of thepast.)

Yet one more dairy dish, whichis especially appreciated in the hotsummer, is a broth called kozhe— a drink fermented from airanand boiled grain, There severalvarieties of summer broths, whichdiffer depending on the kind ofgrain: tary kozhe — millet; bidaikozhe — wheat; zhugeri kozhe —corn; arpa kozhe — oat. Thesecold soups even compete withwhat the Kazakhs consider to bethe royal drinks — kymys (mare’smilk) and shubat (camel milk).Well, maybe not quite…they areso highly rated only if kumiss andshubat should not be available.

Starting even as long ago asHerodotus, much has been writtenabout fermented mare’s milkbeing an inseparable part ofnomadic culture. Here I wouldnote though two characteristics ofkumiss which are not generallyknown. First of all, it is a triedand true means of treatingtuberculosis. Even today there arespecial kumiss treatment clinics,like sanatoria, in the Kazakh cityof Borovoe (Burabai) in Kokchetaudistrict and also in Bashkiria. Asecond quality of kumiss is that itis an aphrodisiac and remedy forimpotence and nervous disorders.Such giants of Russian culture asLev Tolstoi and Sergei Rach-maninov in their day underwentsuccessful treatment with Bashkirkumiss.

There are certain misun-derstandings generally connectedwith the neophyte’s first attemptto join the clan of kumissenthusiasts. Its taste is unusualand unexpected for the tastebuds, which anticipate thesomewhat sweet delicacy ofsomething like milk. It would bea better approach to expect ataste which is reminiscent of beer,

53

whose varieties are indeed infinite.So it is with kumiss, which varies,for example, with the season —summer, autumn and winter —and with the length offermentation — young, one-day(saumal), two day, i.e., alreadyripe and somewhat intoxicating(tunemel), and finally, aged —truly intoxicating (ushkundik).When first trying kumiss — this isadvice for kumiss neophytes —one should start with smallportions and not on an emptystomach, in order not to upset thedigestion.

One of the secrets of goodkumiss is careful stirring. In thedays when kumiss had a specialplace in a traditional yurt, a highleather bag of kumiss with apaddle (pispek) in it usually stoodto the right of the entrance on awooden platform. Each personentering the yurt and passing byit was expected to stir the kumisswith this paddle. When kumisswas poured, it was done in asteady stream using a special ladleand raising it high to allow thestream of kumiss to fall freely,absorb oxygen and foam.

Shubat (fermented camel milk)is lighter, softer and has a higherfat content than kumiss. Thisdrink is not so alcoholic and is verynutritious. Shubat is especiallyvaluable for exhaustion andstomach or intestinal problems aswell as for diabetes. Whenfermenting, camel milk does notbecome more concentrated orsolidify, but retains a naturallythick, almost velvety uniformity. Ifone boils this milk, it begins tocurdle and the resulting mass,which is sweet and light, becomesa special dish, balkaimak (“honeycream”).

Natural kumiss and shubat arepractically unobtainable outsidethe areas where they traditionallywere produced, just as themajority of Kazakh dairy dishescannot be prepared other than onthe Kazakh steppe. The apparentexplanation is the odd fact that thewhole milk with which one begins

contains from its inceptionsubstances which allow it toremain unaltered — according tothe Kazakh culinary under-standing, it does not age quickly— despite the fact that in freshmilk quite naturally every day andhour there are living processes atwork. I remember well howKazakh elders greeted withskepticism the first refrigeratorsand freezers. Although they hadsuch great economic benefits, atthe same time, as was believed inthe villages, they forced foodproducts in a somewhat be-wildered fashion to freeze instrange, uncomfortable andunnatural poses. It was as thoughthey were seized suddenly andunexpectedly by freezing andforcibly detained…

In considering the current stageof the development of Kazakhfood, I can appreciate more clearlyhow far the cuisine of the nomadshas departed from its traditionalsources. Possibly just as far asthe nomads themselves havemoved… Instead of interactionwith food — interaction which isvery like basic human stages ofdevelopment such as recognition,attachment, touching, conception,birth, growth, transformation,disappearance and resurrection —there is the consumption of pre-prepared foods, prepared bystrangers. That is, it is con-sumption of products which aperson obtains without observingand even knowing the processesof their preparation. Moderncuisine has become impersonaland part of the inexorable streamof mass industry and culture.Food (tamak, as) for oneself, one’sfamily and one’s guests has beentransformed into a product (azyk-tulik, produktylary!), identical foreach and every person.

There are, however, house-wives who try to preserve, if nottraditional food in all its richnessand original form, in practiceimpossible, then at least somecomponents of it along with theprocedures for their preparation athome. Among these still popular

homemade dishes are, forexample, pilaf, beshbarmak,shelpek and baursak.

One of the dishes which hasbeen part of the requiredrepertoire of my family cuisine andwhich accompanies me from cityto city, from country to countryand from continent to continent isdairy suzbe, analogous to theAmerican farmers’ cheese“Friendship.” The variant of therecipe adapted to Americaningredients is remarkably simple.Pour into a heavy saucepan6 threeone-liter cartons (three quarts) ofBulgarian cultured buttermilk,place it in the evening in a coldoven, turn on the heat set at 350degrees (F), and cook for exactlyone hour. Then remove and cool;let it rest and settle until morning.In the morning, holding the masswhich has formed with the hand,carefully pour off the whey (whichcan be used for preparing breador bliny or for cosmetic purposes)and serve — the suzbe is ready!Lovers of more solid suzbe candrain the mass using cheesecloth.At this stage of preparation, suzbeis an ideal breakfast. One can addto it, according to taste, sourcream or yogurt, jam or dried fruit(fresh raisins are especially good),an apple or a banana. If onestrains it further, one obtains anew and quickly prepared dish.For this one needs half a pound ofsuzbe, one egg, one tablespoonof sugar and one tablespoon offlour. Mix it all together, and makeseven balls from the mixture,dredge them in flour, gently patthem with the hand and fry invegetable oil on both sides until agolden crust forms — thus onegets syrniki ( ir imshik salyppisirgen nan) or cheese fritters.They are very good served withsour cream, preserves or honey.From the same suzbe, yet furthersqueezed and drained, one caneven prepare homemade kurt.For this one needs to mix with thesuzbe, which has been sqeezeddry, two tablespoons of dryroasted flour, two tablespoons ofmelted butter, half a teaspoon of

54

salt. Shape the resulting mass intoelongated “sausages” (as always,making seven of them), and letthem dry in the sun while coveredwith cheescloth. That’s all thereis to making kurt. (Incidentally,Kazakh children like kurt just asmuch as modern city kids likechewing gum.)

Probably the most productivetime for traditional Kazakh cuisine,as I recall, was the period of the1970s and beginning of the 1980s,when new dishes appeared, alongwith the relative abundance ofingredients and access to newtechnology such as gas stoves,ovens and refrigerators. In theprocess of enriching and varyingthe Kazakh table, of course a rolewas played by the interaction ofKazakhs with the foods of a goodmany other peoples of CentralAsia. It was then that manty(Chinese dumplings) became aregular feature of Kazakh food, asdid (Dungan noodles fromstretched dough), Siberianpelmeni, Russo-Ukrainian borshchwith beets, and also a variety ofcanned preserves (marinated andsalted tomatoes, cucumbers,apples, cabbage, mushrooms),and piquant Korean marinadeswhich still fill several aisles in thecity market of Almaty. Theabundance of rice grown in theKazakh south, especially inKyzylorda, has made pilaf,especially, almost daily fare for theKazakh family.

Kazakh pilaf is very simple toprepare. It contains five basicingredients: rendered fat, mutton,onion, carrots and well washedrice, over which is poured boilingwater. Salt and pepper to taste.Preparation time is an hour and ahalf: 20 minutes to clean and slicethe vegetables, 30 minutes tobraise the meat (the meat isbraised separately for 10 minutes,with the onion for another 10minutes, and then another 10minutes with the carrots). Afteradding the rice and water (to alevel of one finger’s width abovethe rice), cook for another 40minutes. All in all, just 90

minutes. Pilaf is cooked tightlycovered. All other details are leftto creativity and experience.

Let me share a few littlesecrets: 1. Pilaf will be tastier ifone adds apart from the meatitself a few lamb ribs (the bonesintensify the flavor); 2. one shouldbraise the meat before boiling offthe excess moisture; 3. it is betterto salt the meat the night beforeusing it; 4. twenty minutes afteradding the boiling water to therice, use a flat spoon to shape therice into the center of the pot likethe dome of a yurt and with thehandle create an opening in thecenter going right down to thebottom — rather like the smokehole in a yurt — in order that therice can breathe; 5. when theplates have been laid and thefamily called to the table, with theburner turned off, cover the pilafwith a clean towel and then onceagain the pot lid in order to allowthe pilaf to rest. The towelabsorbs excess moisture. In fiveto ten minutes remove the lid andthe towel, carefully mix the lowerlayers of meat with the rice, andserve.

Yet one more opportunity forcreativity is to add herbs(barberry, cumin, peas, nuts,raisins), but here we already arereturning to the more refined pilafof the Uzbek neighbor, who sogenerously received thatapocryphal Kazakh of theanecdote with which we began ourjourney into the world of Kazakhcuisine…

It goes without saying, that incontemporary conditions, es-pecially abroad — for example, inthe U.S. — the demandingtraditional cook will complain thatthe mutton is not the same, therice is not the same, the water ischlorinated, etc. However, theseexcuses are not important for thetrue master of traditional pilaf;rather, what is important is the potin which pilaf should be prepared.We are talking here of the kazan,the heavy metal cauldron with ahemispherical bottom. That

shape, as is the case with theChinese wok, allows the flame ofthe campfire (or gas burner) toheat the cauldron equally, and thethickness of the wall retains thetemperature well. The closestresult can be obtained by the useof a Dutch oven. However, weadmit that Kazakh students intheir dormitories today preparepilaf in pots and in deep frypans.

Kazakh pilaf varies, and onlywhen all are taken into accountcan one form a kind of idealizedimage of what it should be. To agreater or lesser degree then anygiven version is a reasonableapproximation. The same can besaid about any kind of popularcreation: for example, the idealform of a song is composed of thetotality of all its variants, and it isprecisely this fact which makesone want to listen to it again andagain…

The kazan is the ideal vessel forpreparing yet another popular dishtoday, dymlama. In conclusion,let me share the recipe.

Dymlama is meat braised withroot vegetables. Cut the meat(lamb or beef) into large chunks,quickly sear it in fat, add onion,salt, pepper (or other herbs) andthen add layers of various rootvegetables (lightly salted), slicedto the size of a hen’s egg. This isbest done in the followingsequence, from the heavier andfirmer to the lighter and softer:carrot, potato, rutabaga (orturnip), beets, eggplant (in itsskin), green cabbage cut intowedges including the core, andthen at the top tomatoes, split inhalf, and bunches of parsley anddill (which will be discarded whenthe dish is done). Do not add evena drop of water! Cover tightly witha lid and braise one and a half totwo or even two and a half hoursover low heat. Do not open thelid, and do not stir the contents.When the dish is done, the meatis completely tender and hasabsorbed all the juices, and eachvegetable retains its shape, colorand taste. There is not much

55

gravy, but it is exceptionally tasty.As far as quantities go, one shoulduse approximately the sameweight of each ingredient. Thechoice of vegetables may varywith the season. The best time isautumn when the vegetables havejust been dug out.

Of course it is good to haverecipes already at hand, and themore the better, since that waythere is a choice. Yet I believethat if the marvellous recipes ofthe cuisine of our sedentarizedCentral Asian relatives should withtime become part of world cuisine,that in no way would diminish theculinary creativity rooted in thecenturies old life experience of thenomads. At very least I wouldhope that not only the perfect finalproduct, but the very process ofthe interaction amongst food,people, and society at large willnot be ignored by those who todaytake an interest in our experience.Then it will become clear thatalong with the effort to have analways perfect, easily repeatedand reliably guaranteed dish whichwill be completely satisfying, thereis another pleasure, that ofunpredictability and experimen-tation. There is an impressionisticsubtlety of detail and a minimalamount of reflection in the infinitevariations of one and almost thesame dish created without arecipe. Its essence is to createhuman warmth and even the soul.Thus the culinary arts becomeelevated to mythological stature

as the creation of the world. Thatis, we are talking here about foodas the embodiment of theorganizing principle of the world,like the above-mentioned kumisspaddle, which turkic culturesconsider to be in the category ofthe world tree. It is no accidentthat the process of preparingkumiss is sometimes comparedmetaphorically with a shamanisticact.

Underlying traditional foodculture is the entire traditionalculture of a people. It is not onlya culture of food, but food itself isculture.

About the author

Educated in Almaty, Moscow andSt. Petersburg, Alma Kunan-baeva is a Visiting Professor atStanford University’s Departmentof Cultural and Social Anthropol-ogy, where she teaches courseson the culture of Central Asiannomads. She has written the prin-cipal entries on Kazakh traditionalmusic for the New Grove’s Dictio-nary of Music and Musicians andthe Garland Encyclopedia of WorldMusic, published a book, The Soulof Kazakhstan, and been involvedin a broad range of educationalprograms, among them onessponsored by the Aga Khan Trustfor Culture. She is also Presidentof Silk Road House in Berkeley, apioneering non-profit cultural andeducational center supported bythe Silkroad Foundation.

Notes

1. This article was translated fromRussian by Daniel C. Waugh.

2. Incidentally, I never once heardthis story from Uzbeks, and whenI told it I sensed their be-wilderment and embarrassment.

3. The Kazakhs themselves callthis dish et (meat), andbeshbarmak (five fingers) is alater Russian borrowing, derivedfrom the fact that a dish shouldhave a name which distinguishesit from the ingredients.

4. In a similar fashion childlessparents adopt orphans as theirsons or daughters.

5. The second half of the 20th

century saw the introduction of itsinnovations in the churning ofbutter at home. Hence the uniqueKazakh popular method ofchurning butter in first-generationvertical washing machines with anagitator on the bottom. This hasno effect on the quality of thebutter, which, when collected, iswell rinsed and then, as a rule,lightly salted.

6. I bought mine for 25 cents at aflea market in Berkeley. It is analuminim pressure cooker withouta lid and missing its handle.

56