Illll r ~ II I 11,1 r i TOC"'~"'T"«,", ""ltlcm"nl5dndc<"",~[i"'()f'hc:SibaculfUre.3nE ..lyBronz<Asecul,ure who«phy.i"allYpeWOUld3ppeortc,hc:Mongolo1d,lntheS<lmererri,ory where we would lare, expectthe Yuc>hi. The only wa~ we can d'5<1W Yuozhi "",<l<,end tbe name over much hro.der area. of ,heT.tlrn and Turpan !>,,,,n' ,han 'he h..ro",:,,1 record,aUow. We.re engaged In mOl'mgMme. aero" a map wlIhou, any evid=of 'he people. If 'he Chine.., histo,,,,s had never mtnnoned rhe YueUl1 on the bord... of Ch,n., noorchacologi" wouldhavehadtheshgh'''''reasonropo'lUlarc,heirexiSfenco TheYue,h, "re'ghosr.".uromonedupbyh"rorian,rorormentarchaeolosm,.'Weh.ve no need to look for ghost>; we have the remain' of ",.1 ~ople w!th ade"", ,omeeviden"eof,hn,m.to,ia]cuJEbre. Lingui.ticPrehi.wry Embedded m a"" language is evid""ce of iu p"-'t, The prec<ding,enrenc. of word. contain. five word. whICh are denved from GermanIc (i". any,IS,of",.).dueewluchden vefromult1nvlaF"'nch(ianguage...·,dence. p",') and one which combm~. a latin prefIx wlfh • GermanIc roar (emb.dded).Tharrhemos,ba.icv<rbs"rnde.andpreposlfIonsa",all GermanI' would ,usse" a ",)a,iv. chronology In which Engh'h w•• "riguully a German,c bnguage ,har w,," lar~r heavily mtluenced by a Romance language. Anexam,n"tlOnoi rhe Tocharian language. (we can now come up for airandrev.."orlu.d."gnanonr.lfherthanKA)alsoreveal..ometh,ngof th.irpo"althoughwec.nno,workbockt""f"h~"'.Cu"o ... lyenough.,ru. impacrofChin"",onToch.rl.n"norv«y.,gnific.nt(e.g.lochan.nB cdne 'pro,e' denves from the Ch,nese word for 'money', Ibe (modern) qia~ WIth the medieval pmnun";a"on 'd~"an; a)'o me.suremen{$ of capoClt!\ and,a.D. Q. Adarru;h •• recenrly Indlca'M"he word for 'flee') nor does TurklShprov.dca.ignifiean'KlurceofToch.roanrocabula'l'On,heorher hand,M,dclle lran,anl.nguagescurrent III dteTarim Ba"nand.urmundmg arta. in rhe hrmillennium ADprov,ded.ma)orandpred"table,ourceoi 10anwordsroth~Toch.ri.nl.nguage•. FromKhorane.. S.ka"helanguage of ,he ,ourhern Tarlm Ba'Jn,wefindIoanword•• ucha.dsa~a- 'worthy' borrowed",Toehan3l1B"1arr>'wortby',whoietheSog<iian.who<ICrecla' one of the principalmtddlemen m ,he transpo"""on of good,along,he S;Ik Road contriburecllhe word mwll 'wlIle' (cognate with OUI Engh.h m.ad') wInch wa. taken up a, Tocharian B mOl 'alcohol'. Somewhat earlie, conracts With I,anl.n languages"on be...,nin word. ouch a' Isam, 'arrow', which waS app.renrlyborrowedfromano"bea'tlranianI3ngu.g,'"eb .. A~".n ("""""-'we.pon'). Not all borrowmgs came by way of I,aru'n Slncethere "al.oe.,denceof IndJc borrowing. (a, well a, thed,rect impOrt3tIOn of San,l<rlrln Tochar",nrehgwu,li"rarure). These ,ermS refe,ro.,pecrs of "in. Buddhl"hfewhu:hwoujdha~b..,nfo""gn,orhep",-Buddht"Tocha'''''''' e,g. Toclt.tlan B airy4 'nun" asarj're<IChe,·.nd samane 'monk' whICh all fmd ,heir ongm. in word. f'om Gnndhlrl (rhe language of ASok.·, m,criptlom) of thelnd century BCtO rhe4,hcenrury AD, Tochnn.n'pocinli.ts.rege ner.llyre!uc,an,ro.dm",har,h.",wa. any flow of Indo-Iranian vocabulary Into Tochan.n e.rlier than the 1" millenmum Be,. mnewhen words such as Tochanan A porat, Tocha"an Bperel'a",,'ruayhavebeenborrowedfmmKlmelranlanlanguage(wehnd O,serle fariit 'axe')_ Tochari.:!n al,o employs a word for 'oron' (A ancu, B "ricUIlIO) whICh 'Ome compare wah .imilar word, .uch os O.. otic ~~d4n 'stU:I'.While.u,hev,deoceeanonlybeemployedpo.invely,i.o.i,cannor demonstrate'h.t~on ..c.. didnotexl..,rhe",ha~beenfewto.ugge",ha, ,hoancestorso[,hoTochar""uW<:relndo>econ",,,with,ho,,lndo-lranian netghboun;beforec.l000--500ac.Togobackanyearlierrake.u,inroone oflh.mostd"pured~<ea.ofTocharian"udies Real 'Europeans" The ",Iationsh,pberweenToch.rr.nand theothe,Indo_European otod. or.ubfamlh.. ,nd ,be i..ue of how 'Europe.n-the Tochanan language. a"u.lly orc hn. dogged Toch.rr.n "udle, ..nee i,.mcep,ion. If.,.., ",!l"", ogamon,hevartou,lndo_Europeanlanguagestock.,w< ... ,hat,ha,whi,h embraces ,he IncilC (and Nuri",ni) and lraman languages i. ,hemo" ..r",m higher-order group inlndo"European. By ,his we meon ,horwe c;n recon.rrue,ahng",,,jc,,.gehe,weentbedevdopmen,of l<anian.nd Indo ­ Ary.non rheonehand ••nd Pro,o_lndo_Europeanon ,heorher, namely, Proro-lndo-lran,.n.lnshorr.wegetfromProto-lndo-Europeanro5an,l:m orA,...,anonlybvw:>yofProro-lndo-lrani.n,.nd.nydi<cussionoflndo­ Aryan Or Irani.n orig,in.will 'ndud. the geographical sr.gingareaof Proro-Indo-lr~nian. The malorilyof,<11rhough noralI,lndo-Europeau15 .. wouldal.oaccepraBalro-51.vic"agebefo'"theind.:pendenrde~lopmenr of rhe Bal{tc and 5lovlClanguages. Now borh lhe.e.ubgroup. of lndo­ European are.atem languages (I.e. rhey changed a 'hard'gorkroj,.or.) andlheexJ"enceofodteruruquelingnlSticfea,u",.,ha",db.twttn,he.. ,atem languages.uggests ,omedegree of proximity berween rho 'WO group., I.e. in ,orne way there was acontlnuum of Balro·Sbv,,-lndo-lr.ruan which expenencedanumberofcommoninnoyltfionsnor.xpeflencedby,heorb", Indo-Eumpean.ro,k•. ToehatJsn,a,wehave •..,nlnChaptl',Throe.belong. '0 the cen'um langu.ge, whicb are dcfmed by 'h"'rhordpronunciarion of wh.rwereconsrruc,.,palaral vetars ('g, "/f.,erc.). I,i. generally, alrhough norun,~"ally,pro,um.d,ha,rhecen,umlanguage,dldno'partic,pa,ein the ,nnovatlon. exp~rrenced by ,he .arem language.; being a cen'um langu.ge,however,doesno,te))youanymoreabout,"preh1Srotlc ",Jation •. Forrhar,JjnguJS"ba~.oughrorhercm.riaamongrheV3"o",centum Iangu.gcs,oderermlIlerbe!,cotrtcrpositton,onthelndo-EuropeanfamJly '",e, And "uly ama>ing atr"l' of modd. 'ha' ,hey havt: proposed. We hsr rheuputatlvt:,e!at!on,hip' below on a wesr roea,taxi.. I,.." ! I I~j <{I ~, I~I 1:;0 ~! '" v; 'v. li: .~ :;s" ""5" ~ 0 ~ "" ~ iill ,he,. could ent.. tho Tarim B..,n from eIther the we., or the n"rrh. Of tWo potentlal,ou=sthe""'''l,.urelytheprefe,redopuonbecau,ethemain .ourceofSak.,ox"i.Khotanonthe,outhemS,lkRoad,whichwosnoE rW'\Y3pproacltabled"""tlyfromthenorrhotherlhanalong.de.enn>ad W'.,h""".lso,eenth.!rilephY'1Calevldencealongrheenrj"""urhem rou," ofth.$J!kRoadback,ntoBacm33ppe;l,..tobem.rkedbyth''''''''~neraJ phy,ical type. Tho ,raJ=<>ry of th,.lndo·Afghantype,the",fon:, i.mo,r ...ily ... ignedtoth""e",hocarnedlnd1Co'I'.ni~n(S.kal\.nguage.fmm the we&!, and Ih"" f"rrh.. con["'t1ed by rhe .urvlv~1 of Sob'. doses, lingu"u< rdations in rhePa,mr•• ThU5, by 0 crude proces. of ehminatlon, rh .. leav<. ,heearlieratteo'edPrmo_Eur<>pold. (orHemphll!"peflpheral Qawrighul people) .s the mor~ ltkely <andida,.. to he ldenl1fied wlm the TochartanC.ubnr.to.Thcrei..,.dmi,tedly.ach,onologic.lgaplxtween thelaro", mumm,es.ndKrotani3!l tex" but we should rec.ll 'hat the mummie< from. fur example. Zaghunluq ('Ur·DOI'ld·, tM baby in 'he blue I>onnet.,he,dpartllewoman"he'ScreamBaby·••!C.jdarotoc.60011C and wo are only,ugge>,ing,ha"heybeassodatedwilhalmgu1Stlcsub,rrale th.tmu"ha.ebeen ..",bh.hedb.foretlte3rdcenturyAD.AsTochanan loartwor<l. a,. atru,ed in ChInese. few c.ntutles before fhe Kroranian docum.nrs.wecanperhap.narrowtltegapstlllfurtlterlxtweenrhep,ehUit_ oncmummies.nd,heTocharianlanguage.. TheTurpangroupwitb".bu[]al,danngtoal>outmeStltor4thcentun.. scfaU.wi,hlnarernrorywherehi.to!lcalevl<kncoreveal.l>otltToch.r1.n AandBdOCllmen,.. Wlnlei,,,diff,,,ul'toknowhowfarbaokwecanextr.­ polare eithu of ,he,e langt1'g., a' 'M vernacularlangu"P' of the region. TocharJan at le.sr supplies a plau51bl. candidate for the Turpan group 0 f mumrni..... nceweh""'noOfherlanguagem'h.. regKIn'h.toffe.. acred,ble alrern.t1ve. In ,.rms of generol phrsical'yp«.hereroollreProto·Europoid ,ype.ppe."bur.long.;Jde,helndo·Afg!tantype.with,MPamir_Ferghana type apparendy the most re.:en'. Lop"u, ~r~:F PhyslulTyp. und"""mined Proro-EuroprndiQiiwrilllml"",I",: Indo.Afghan Qumul Pr",o-Eu"'l""d Tu'pan Proto.Europ",cl Indo-Afghon ""mr~P<:rgh.n. _ alsonow",caUWernerWin="'uggestionrh",thelanguall"whlch~merged In wrltt.n ,ources 3S Tochan.n A may bave boen .poke. ra"her to the norrheast.i.e.inthedifOcllonoftbelaror T"rki,hpen.",monofEasr Cent<aIA.ia.hefo<enhad heentaken up~.a'd""d· htuWcall.nguage When_"""minelhe general physical type of the moreeasurlypopula,j"". iuch a. tho.efrom ¥anbulaq. It is,he •• me ,IS ,hat found "' Q:iwrighol. The chronologtcal gap b.twet'n the Qumul mumm", .. dared rO abou, 1000 Be, .ndTocharian'ext.ofabou'the7Ih~nturyAD",ndeedg<eatbufherew. may rec.1l fhat with respect '0 Toohanan A we an: d.:ahngwJth an already deadlanguasewh,ch,houldh.""prr:.:ededi.. h"ronc.latt~"anonmth" r~glDn by a matter of cenruries, fhus uarrowlng fhe gap be,wc:t:n the mummr.. and our language horizon. Here '00 rhen one migh, p«>pose B north-,outhnwvem.I1Iof.arlyToohanon•. Tho """knee of a crud. ,I.. h from Occam', rator would seem rO ,ugge.. tha' ,he malOnty of our earlre" p<ehlSlotlc mummies may be «garded ao aOC<itou of the I"•• Tochari.n,. The5e conc\u.ions should. however. be '<mpetedwith'heob.. rvationtltat,.~ceptonlh.b3",ofmed,evalw.1I pain,ings.we oanno',poak of th<phYSlcal type of tbo•• popularion, who occupied ,h. terri'ory most do.ely a,""c,aced Wltlt our Tooharian B doc· umen... i,e.aroundKucha.lnOlhe.-wor<l,.w<:canno, ...,lyrelatetheBronze Age populations witlt ,he laroc oa'" ".,", mentIOned in Han documen .. tltatproduc.d the Tooharran documents of the hlOto!1calpenod.Wha' we have proposed i,merdy,ha,on,bo""idenceof phy"cal typo!ogy and geog:raph,cal d;stribu"on, ,be T<>ch..riana would appear to denv<: from Ihe notthand,helndo--lraman,l~rgolyfromth.we",hown>bU5'a"'such conclu"onswhenWe3ddinto,beequatlon,h<"ld enc oofarch.eology? TheFlrstFarmus Mummygroup5,languagesandbacl<groundphyficalcypef Gro" Charcho" Frnally. ahhaugh the Qumul mummies lie beyond aur documen,ary .vtd.ncofo'llre""rnacularl.ngu~geoofEa"C'I1I"'IAsia._.ullhavea f..... hin .... 'awhallangu.geth.ymayhave.poken.Wedo.fo,.xample. h",,< Tocharran documenrs a mere 150 km 193 miles) rO ,h. woo<- We mighr Wehaveseenhowourearhe"arch••oJogical""ideneeforhumanoccuparion of E.stCenffal As!alSexrremely meagre and i,i. lmpoooible ",derermine withanr ce"31nty from wh,chdu-ecrion rheTatlm and neighbouringreglo!!$ we"'lmtlaliy..,rrledifcolon"""ronpre«dedrheBronzeAge.l.e. c .2000BC. Ftonlrhebeginning.;ofth.Sron:oeAge"'OU,MO",:um,tanriaIOYldence tltaf we havc surveyed would rend"r" more hkely tha' it Wa' ,."led from enhertltenortltorlhewesrralher thanfhe"'''th or •• st. The p.udty of a Mongolo,d phYSical type .mong rho e3di." human remains .nd il> conf,nement to fhe far east of East Central As.ia ,ugg~,,' ,ha' MongolOId population,enreredl.feandfrom,heeastand",",heaot.Futtbermore.we h..... goOdidea"fthe'.nlemen.. andcemet..... oftheNeohtlllc¥angohao cuhureofno"h.rnChlRawhlchexpandedwestword"n,onelghbouring cIlIa is U LEE.. ,pn., , II W"OWe"'H'M~"''''''<l l!l:t The Linguiscic Srratigraphy of Tocharian ~I':'~ We Imve ",en in the pre<ecling 'e<tions tnat one "f ,hecmical is,u.. " ,h. «>,,,"ctorderongofourev,deru;e,beir,narofpopul.tioruorarcnoeologlcal :1," mfl"enco •. Thepmblemheforeus n ow15arraoginginC<ltrec<chronologjeal ~ j ,I ;~ orclorthediff.",ntptehi.lOriclangu.ges.pokenlnlhcT.rimBosin.V.rion. deg=' of diff,cultyo<xomp.ny,uch,ne",.tcise bu,!romou'perspecti"" ,he ptimarf goal is to arrangc corre<rJy the ,elotion,hlp between th. To<hari.n.ndlran"",langu.g•• a,rhe•• ar.'h.langu.g~mo"likelyto nave""rvecla,thevernacularscf,t>pr.hj,rorlcp<>pulatlons. As we h.ve.l",ady ..en,.very language COmalO. s<lmething of "'OWn cultural h"tory in "'vo<abuJar), Some oi Ih. rocab111ary will he tnbemea from ProtO-incio·E:ump<on, some wIll b. new word. created [,om olde, Inhe",.dd.men".The.. donorholdthesamemt~n:stfotu.herethat loanword. do, sioce borrowed vocabulary, paruculariy d,ao;;dtlc cultural vocabular)l, rnay poll1rcothet!lneandplaceoffore,gnconta~ ... A.wehav~re<narkedbefore,Chin=appea"toha""had.veryminimal I I Impact on the Tocn.rian language. To !x ,ure, we have th< odd loanword, o.g. T""harian klu 'rice' from Old Sininc 'glow (mo.rern Chinese Mol and IUmo,for units of m....urement,e.g. lank,lOw (=lO,ank)and cak ('"10 tow), all tran.parently fcom Ch,n""". Bm the nnpacl her. IS m'nim~l, <ul,urallyptethcr.bleandcomporativolyr.cent. Otherwise, Ihere" no "idence ,hat the Tocban.n. ga,n.d tbelr'lOc.bula,y rortn.,na"veflora and faun. of the Tanm B.Sln from earlier Chme.e inhbitan", nor do they app<3r to have gamed the" agrl~ulrur.l vocabulary from thi ••o"rce. Till, rend. addinonal.upport 10 the argument that the Chinese were bath la,ecomersintoth.T.nmBa,mandd,dnotha""a"gnlflCantlmp.<tuntti ... Iativelylatelnthoeaswwm<>stregion,. Now many of the other loanword, found m Tochar...n are connet:ted withtherelog;ousand,.,,,,alloleassoc,.tedwithBuddhl5Inancihence,M .ourco.here.reBuddhtstHybridS.n.knt,Prakm,.ndotherlndic languages,o<casionallyperhap.f,lrered,hroughlran,an,ntermedi.ries. The,. ",ay be eaSliy ...ough ... ""c,.ted wlth the n:i.tlvely I.te.p",ad of Buddhl.m amO}ngalrady.stabl"hed popul."on., both TO}ch~".n and lran,on-speaktng. Byfarthegre.testn~mberoflo.nword.inTo<h31"lJ1wo"id.pp ...rto derJVe from S.ka or some other E.... Ir.ruan Jang""ge and Ih• ..,fiU outa larger",m.nncrange.ln,omeln".n=theyappeartohaveb""nloanworo, a,,,,,datedwlthexchange,e.g.Toch31JanBp,to'price'm.yderivefromS.ka piho."price',andTochananA""""To<harianBperi'debt'wouldb.de"ved from Sob pl",_ 'th.t wh,ch ,,'0 be patd'. Th.re are other word. wh,ch nughralsobo.",gnedrothegen."r .....lmofcammerce,e.g.unn.ofwdght andmeasu",. lranlOn m,Utary t.rms al.o had an ,mpacr nn Tomanan and wh.n one co,,,,d.rscl<ptcrion. ,n Buddhi$t temples of Tocharian knlgh" kiue<1 om hke SasMn,.n. with long sword., one mIght recall ,h"the Tocha,ian B word for ,word, kerrlt:, "domed from ,arne East Iran,an 1'lJguagc,e.gAves'anlwr"a-'da~·Bact"anal.o,uppliedlo.nword. fiUingoutSlm,larpo[,tkalconte>tt',e.g.k.amrMo'head','m,ef'(go<!)glves Tochanan A kakmor/,k 'ruler'. It is .1,0 Impo'tant to no", that the TocbOtion word for ',ron' (Toc:h."."encuwo or ~ncu) would appear to be relatw '" .ome way to Iranian (O...,tlc). In .ome ms",nceo the loanword. arc connected with exotic .nlmals, e.g. Tocha.,.n B mew,yo '"g.r' f,om Saka m"y<,- ',ige", Or Tom.".n B ,k,mek.. 'pigeon' which ,eem, related toS.ka,wa""k.a·'plgeon'.Bynndla,gc,howeve",twouldbeexceed!ngly diffICult Imgu""caUy 10 make ~ny case fo' the "'mpornl pnomy of Sak. In 'he Ta,im Ba'In WIth "'speCt to Toc:hanan. 01 tho approxi"",,,,1y 2S or 30 loanword., moot could be ea,ily .xpl~med as later bormwmg, pa,sed betweencommunitiosalong,kSllkRoadconnectedwlth,h.,iseofurban orBuddh,stln'tim'ion,.M,,",'importandy,th~re"noe\"idencethatthe vocabularyof.grlculture,n,heTarimBa,,,,,pecificaHydorivesfromSalr.o. A... s,CanaJ~andBricks AlthoughSakaandSogd,ano.rctheear"mo£th.lr.n,anlanguagesdirectly anested In the Ta.,m s.."n, there are ""me ['an'an loanword. In Toman.n tha, may dOti"" from a ,tilleadier period. Two of th .... have only recently com.toligh'mtheformofaTocMrianBttxt,hatdocumentsthetran.fer ofan ..tatt.1XIithinthotexttherc"",f.ten<;etothe oro l.411<Wry,,'gre.t canal'.nd,heiirfe,som.formofwa'e,"ou",o,hat,an,.rvea,.n.state boundary< Douglas Adnm, has recently .uggo.tcd tha' bath of the,e word, areoarlylo.nsfromanEa"bamanlangtlagc(o'proto-langU'ge).nd,whde there an: a numbor "f ways to explain how th ..o loan, could have taken place, only On... not probiomatlC. A. "ngation ag,ieulture " tMonly type of agriculrure ,hat one can p'""",e in tl>c Tartm Ba.in, It ISd,fficu!t to see how tM Tochanan. could ha""ekod out a hYing there bofore they knew of irrigation. Ir"al.., unhkrly thot ,h. Iranl~n;andTocha"an' both am""" at prtc"elyrhe.ametlmc, thelran'~n,p ...ingonto th.Toch.,i.n, the ...chnolagyofagrkulrurcbeforcthelarterh.dstarvedtodeath.The,impleg, expbn.rlOn wculd be topl.ce tho Ironian. in po",ion fi"l and thon have ,heT""harlans,larg.lypastorali.ts,wander,nancladop,agtlcul,u"';thi>1, precioely the type of pattern wo have e.rherencountercd m Wesr Central ASIa where .reppe pa5lorah". moved 'nto ,heoa.e. andadopred the agrkulruralrechmquesofth...rloerinhabuants.But, •• wehave.lrcady seen, we do not ,eally have ,yp'cally West Central A.i.n BMAC farmo" o<cupYlngthe Tanm Basin, nor doe, ollrrevu,w of thcphy.",al eVIdence ,uggest that population' from We" Centra! A.ia .rrived in. ,ho Tarim and Turpan ba,i", earhor tban th""" from ,he north. None of our u,ual model, will do, Adorn' suggests that the most convemem ""planation would Involve rhe ndap,ion of irrigation redtniq""" and E... Iranian tertnJnoiogy by ,ho _SLUL)2 :'LSi . J ;;Iib,,· II 1)1 1 I~I 13 II '"' 6'"0 " >-l \ Jf! ,I ~(1 /,<', \\ /;:lJ