Fig. 74 娵 月示 , Pär'vanätha temple, KhaJuräho, Madhya Pradesh, A. D. 975. Fig. 75 畩 月韲示 , Laksmana temple, KhaJ uräho , Madhya Pradesh, dated A. D. 954 ー 55. 128 about probable ongmal placement 0 日 oose fragments. The single essentlal image in any temple is the deity within the sanctum, the g 韲励〃乢 The god may, accordmg to the dedicatlon of the shnne, appear ln amcomc or seml-lconic 応 n11 , such as Siva as litiga, the pillarlike "mark" Of the deity in his generatlve aspect (see Fig. 24 ). Likewise, representatlons of the divine may appear in human, 1COnic form, depictmg the god in h1S or her most hlerarchic/formal/absolute manifestation—often standmg stlftly frontal with feet planted together (see No. 70 ). EIaborate frames, sometimes with garlandlike arches above, ensconce the deity ln a f0n11 remlnlscent Of a door—like a po al 伝 0n1 unmanifest tO mamfest. lt is 伝 om thß sanctum lmage that all other images are denved. The exterlor wall elevations of the sanctum dlsplay most lucidly the process of estabhshing locatlons and their " 10 cs ' ' that were worked out pnmanly between the eighth and tenth centunes. The frleze po れ 10n Of these exterlor walls, between cornlce and bindlng moldings, will be exammed in detail to glve an idea of the fundamental process of fixing lconographic order throughout the monument. The central buttress or offset ( 励記 ra ) 伝 om monuments with figured walls om the sixth century (n) holds reflectlons of the sanctum's dlvinity (see Fig. 39 ). These include actlve manifestatlons of the deity, close 信 nuly divinities, or the sanctum deity repeated in iconic f0n11. They are the 1 れ OSt important lmages on the wall and are placed on an axls with the center of the r 励わー the point of

mtegral part 0 「 what appears tO have been a rather spaclous commercial- manufactunng and residential center. There were temples dedicated to deities called Näräyapabhättäraka ("aspect of Vi»u"), Sivabhattäraka ("aspect of Siva") , Bhaillasväm1, Sigäkiyadeva, and so on. lt is unlikely that at Siyadovipa!!ana, or Other settlements Of this kind, sacred space was exclusive and separated 伝 0n1 commercial and residential space. The admixture of different types of space ln areas Of settlement suggests a close physical proxlmity between deities and their devotees. This was, however, not the only pattern. As previously mentioned, major temple structures with subsidlary structures or clusters of temples constltutlng a sacred landscape could also be segregated 伝 om mundane space. Remote pilgnmage centers ( 庁れ 4 め required that devotees cover wide stretches Of inhospitable terraln ln order tO achieve closeness tO a deity (see Fig. 12 ). Nevertheless, ln order to understand how crucial the temple had become in the practice of everyday rehglon (even When all sectlons Of SOCiety did not have access) , it iS necessary tO keep ln rmnd the vanetles ofspatial contexts in which it figured. The linkages that the temple represented were many. These linkages were made possible at one level by the nonexclusivity of sacred space—a feature demonstrated by the fact that at any particular settlement or S1te, structures with dlfferent sectarlan afflliatlons could coeXISt. This was also made possible by cases ln which a deity Of Brähmanic afflliatlon could assume Jaina charactenstics. The Osian complex in Rajasthan bears evidence to both conditions. Temples dedicated tO deities of different sectanan afflliation coexlst at th1S center. Further, Saccikädevi (popularly known as Saciyä M and regarded as the titular deity of the OswaI community) occupied a promment place at Osian (Fig. 13 ) and exlsted ー 42 alongside her Brähmapic prototype Mah1säsuramardim. Such centers could thus enlarge the network of linkages among communitles and forge a considerable measure Of SOCial lntegration. Linkages mediated through temples operated at other levels as well. The early medieval penod saw the emergence of royal cult centers in different parts of lndia. The royal cult center, like monarchlcal power representlng lntegration of different polltical segments lnto a state structure, could represent lntegratlon of local cults and their centers. The symbiotlc relatlonship between a monarch and what grew to be a royal cult center (and the growmg perceptlon of this relationshlp as perhaps the foremost cultural symbol of a regon) is well illustrated by the cult of the Smva deity Ekahfiga, the patron deity of the GuhiIa rulers of 43 Mewar. This relationshlp was forged over centunes by the Pä'upata Saivas who 44 were in control of the shrine of Ekahfiga. The Guh11a rulers also functioned as the chlef offcers ( 市ルめ of EkaIifiga, the deity being the sovereign of the realm. ln addltion tO the symbohc signlficance that was attached to major centers like the royal cult center, the regular social linkages of a center s network were contmgent upon the way ln which resources were provided tO it by vanous elite groups. This point can be illustrated by citlng evidence for tWO centers, one 伝 0n1 the kmgdom of the Kalacuns of Tnpuri and the other 伝 om that of the NadoI Cähamänas of the Marwar reglon in Rajasthan. The KaIacun record, which dates to the close of the tenth century, comes 伝 om Bilhan. Beg1nning with an lnvocatlon tO Siva, one part Of the inscnptlon records the 仕 s Of at least nlne villages t0 different Saiva ワ bestowed by a Kalacun queen, Nohalä, who was related paternally to CäIukya rulers. 45 The final part of the record specifies a number Of levies that 、 vere lmposed on different ltems Of merchandise at the marketplace ( 24 な 4 加川 d 叩の at Bilhan. The levies were resources allocated to a 川酊 called NohaIe'vara (after Queen Nohalä) and presided over by a Saiva ワ 4 Aghora'iva. They 、 vere lmposed on such items as salt, products 伝 0n1 011 presses, betel nuts, black pepper, dned gmger, and so on. They were also imposed on horses and elephants, suggestlng the presence of long-dlstance, speciahzed merchants at the marketplace. The evidence for a temple and monastery center in the kingdom 0 「 the NadoI Cähamänas is available 伝 om a set of twelfth-century copper plates 佇 om 41 Fig. 13 Entrance tO Saciyäル 1 朝 temple, Osian, Rajasthan, 11 th century, with later addltlons. 43

Beyond the Temple Walls The Scholarly Fate of North lndian Sculpture, A. D. 700—1200 ⅱロた N. Desai or a worshiper in lndia, the temples, along with the images on their walls and in their intenors, regardless Of their scale or age, articulate the spint Of the enshnned deity in the sanctum ln prescnbed ntual ways. Ascending the steps (Fig. 1 ) and walkmg through an archway, the worshiper enters the sacred intenor space ( 川 4 卩 4 叩の and begms the personalized veneration 0f the central deity in the innermost, darkest part 0 「 the temple, the "womb chamber" @韲励の . BY performing some 0f the ntual actlvitles¯rmgmg the bell; chanting; offermg flowers, 応 od , and milk; or lighting the lamps—the worshiper enlivens the spint of god and hopes to forge an actlve personal relationsh1P with the central deity. A イロ r. れ Of the deity—to see him and tO be seen by him¯remams one Of the fundamental pnnciples of Hindu worship to this day (see Fig. 2 ). Another important ritual for the worshiper is the circumambulation Of the temple extenor, an act through wh1Ch one expenences some Of the broader conceptual and architectural meamngs of the temple and the role of the plethora Of sculptures that grace lts extenor (see Fig. 3 ). Even a casual, nonworshiping VIS1tor is struck by the elaborate arrangement Of sculptures on temple extenors and their less profuse presence in the darker inner halls. 嶬市 ile one may not grasp a11 Of the COSn1010 cal meanlngs, lconographic orders, or aesthetic subtleties Of the forms, many 0f these charactenstlcs are foregrounded by the very placement 0f the sculptures. ln direct contrast tO this expenence Of temple sculptures in S1tu, are the mstances ln which a VIS1tor encounters the as entltleS ln a museum. He or she is baffled by the complex iconography as well as the figures exaggerated limbs and contorted positlons that defy the laws 0f human anatomy. Many forms Of art become decontextualized in the museum settmg. However, when viewed discretely, temple sculpture 伝 om north lndia, dating 伝 om about A. D. 700 to 1200 (often referred to as the "medieval" penod) , suffers most gravely. This is due to the fact that, t0 a greater degree than almost any other type, this type of sculpture was designed as an integral part 0f an architectural whole' t0 be seen and understood as a component performmg a COSn1010 cal function, enabhng one to visualize the "manifested multiplicity' of the divine (see Fig. 4 ). lndeed, it is difflcult tO dO 」 ustlce tO these elaborate carvmgs Of divine, semldivine, human, and natural forms When seen ln isolation. ー 40 , dO the context Of the whOle tO bear on the part that ℃ see ln a museum setting? How do we develop a methodology for the study of such sculptures that can take 1ntO account their onglnal placement and function in the temple architecture as 、 vell as their enormous visual and styhstic diversity? These are difflcult but answerable questlons. The fact that they are rarely raised is reflective Of the dearth Of comprehensive scholarship on the subject. Worshipers at the entrance of Kandanyä Mahädeva temple, Khajuräho, Madhya Pradesh, about A. D. 1025 ー 50. 19

WhiIe lndia's archltects expenmented with forms to define the temple as a shelter for both the deity and the worshlper, they began to extend the measure of the sanctum's sacred central spaces ln ways that could be made visible on the temple's outer walls. TO this end, they first apphed a broad central buttress to each wall on which a sculpted image representing one aspect 0f the inner deity could be placed (see Fig. 39 ). By the early seventh century, applying the grid plan of the レ示 tu 川 4 ″ d ロ信 , architects began tO glve one plane Of projectlon on each 、 the width Of the inner sanctum and a second plane (the central buttress, beanng a divine lmage in an archltectural frame that was designed tO suggest a doorway) the 16 measure 0f the intenor sanctum's わ ra んれ 4 (see Fig. 40 ). These stepped planes in effect added tO the square altar/sanctum a 、 cross plan" Of masomc piers on which images could be placed. (See also Danelle Mason's essay. ) Within the sanctum s r 励 4 , the 川反ⅲ acted as the focus for a human s approach to divinity. such a sculpture made it possible for an otherwise mvisible and all—pervading divinity tO seem present for worship at this chosen location. AS a t001 for meditation, such images made the deity actual—not as stone, but as a 17 VISion tnggered by the image in the worshiper's mind. ln like fashion' the temple created an lmage 0 「 the cosmos. ()n ・嶬/ ・ estern sem10t1c terms, it bOth ル and 覊イ the cosmos. ) Vanat10n A number of expenments with architectural forms for sheltenng the dlvinity contnbuted to the final curvihnear ( 〃畩韲の temple tower typical 0 「 northern lndia. small, "flat-roofed,' PIIIared pavilions ( 川叩 d 叩汀 pandals) were used as votive and probably funereal shnnes ln central lndia for a number 0f centunes. (Their superstructure, in fact, suggests an altar platform. ) Some other shnnes still used large wedge-shaped pent-roofs called 〃沁汀 , sometimes for the sanctum but more frequently as the roof for the pillared hall in front (see Fig. 41 ). Rectangular shnnes were sometimes built tO house certaln types Of divinitles (sets 0 「 mother goddesses, Visnu ln a rechning posltion, the nine planets). They occasionally also had ⅲ 5 汀 r00 民 but often instead had a barrel-vaulted 19 Fig. 34 Temple no. 17 , Sanchi, Madhya Pradesh, about A. D. 400 ー 25. 98

36. 37. 38. 39. 41. 42. 43. Their Doctnnal and Sectanan Background, Patronage and SpatiaI Distribution," Ph. D. Dissertation, Centre for Histoncal Studies, JawaharIaI Nehru University, New DeIhi, 1990. See D. N. Jha, "State and Economy ln EarIy MedievaI Himalayan K1ngdom 0 「 Chamba," (paper presented atthe 52d sesslon 0 「 the lndian History Congress). V. V. Mirashi, ん立朝研〃 K ロ信 cl Ⅲ示 C 市 Era ( C ( ) 甲ん立行凹Ⅲ〃 ん市 ca Ⅲ川 , V01. 4 , pt. 1 ) (Ootacamund: Department ofArchaeoIogy, lndia, 1955 ) , CLI. DebaIa Mitra, B “靃んⅳ M ( リⅢ川閉な (Calcutta: Sahitya Sansad, 1971 ) , pp. 85 ー 89 , 225 ー 32 , 24 ( ト 46. One must keep ⅲ mind the 信 c い h although by now Buddhism had lost its dominance and its distinct identity, ma. 」 or Buddhist centers like N 引 an 市 , Vikrama'iIä, Somapura Vihära, Ratnag1n, and Mainamati, of which reveal massive structural and sculptural activltles, continued to exist untilthe close ofthe early medieval penod. The shnne ( Ⅷ ofAravyaväsini, the 。、 forest deity, was established at Aravyaküpag1n; R. R. Halder, 、 'Samoli lnscnption of the Time of Siladitya; [Vikrama-Samvat] 703 , ' ' E 巨叩 / ん市朝 , V01. 20 ( 1929 ー 30 ) , Repnnt (New Delhi: Archae010g1cal Survey oflndia, 1983 ) , pp. 97 ー 99. F. Kielhorn, 、、 KhaIimpur Plate 0 「 Dharm 叩äladeva,' E 叩 / ん市 ca , vol. 4. ( 1896 ー 97 ) , Reprint (New Delhi: Archaeolog1cal Survey oflndia, 1979 ) , pp. 243 ー 54. SimiIarIy, when Dharmapäla, PäIa ruler 0 「 eastern lndia ⅲ the late eighth t0 early ninth century, was making a grant 0f villages for a deity, a small temple shnne ( Ⅷ kt 市たの for Goddess Kädambari—probably a 10Ca1 goddess absorbed intO an ever—expanding BrähmaVIC pantheon—was mentioned as a boundary mark 0 「 one 0 「 the villages. F. K1eIhorn, 、、 Siyadoni Stone lnscnptlon," 石巨叩ん信ん市 ca , vol. 1 Reprmt (New Delhi: ArchaeoIogncaI Survey oflndia), pp. 126 ー 79. The evidence Of this inscnption has been analyzed in some detail in B. D. Chattopadhyaya, 、、 Trade and Urban Centers in EarIy Medieval North lndia,' 7 加ん市 Historical R Ⅲ , vol. 1 , no. 2 ( 1974 ) , pp. 203 ー 19. Devendra Handa, op. C1t. , pp. 14 ー 17 , 52 ー 55. A. Eschmann, H. Kulke and G. C. Tnpathi, 7 加 CuIt 9 リ川第んイ〃 R ( ) 〃行ⅲ ( 川〇 ns (DeIhi: Manohar PubIications, 1978 ). According tO the cult 0 「 Purusottama—Jagannätha 0 「 Onssa, the deity became the ruler ofthe reg10n and the monarch his representatlve. 44. 45. 46. 47. For a recent study Of the relationship between the Guhilas, the Lakuli<a Pä<upatas and the shnne of Ekalifiga, see Nandini Sinha, 、、 Guhila Lineages and the Emergence of State in Early Medieval Mewar," M. Phil. Dissertation, Centre for Historical Studies, Jawaharlal Nehru University, New Delhi, 1988. V. V. Mirashi, op. C1t. , pp. 2 ( ) 4 ー 24. D. C. Sircar, 、 'Stray Plates 佇 om Nanana, 石叩 / ⅲん市 ca , vol. 33 , pp. 238 ー 46 ; S. Sankaranarayan, 、、 Nanana Copperplates 0 「 the Time 0 「 Kum 計叩引 a and AIhana,' 石 ra ⅲん市 , V01. 39 , pt. 1 , pp. 17-26. For detailed analysis 0 「 the Nanana plates 丘 om this perspective, see B. D. Chattopadhyaya, 月平“な ofRural S Ⅲ lel 怩〃なイ RuraI & ⅵアⅲ E ロ町 Me 市 6 Ⅷ / ん市 a , chapter 3. 47

Ha み ara : NP. ofa conjoined 応 rm of Siva. H ル 4 4 : NP. 0 「 a demon killed by ・ Visnu in hiS Varähävatära. イ ra : NP. Vedic king 0 「 the gods, God 0 「 Storms and CIouds; 市 a 0 「 the east. His レ〃 4 is an elephant. イれ i : NP. of a た行 oflndra; one of the Sap tamätrkäs. I 命れ 4 : NP. 0 「 a 応 rm 0 「 Siva; 市ロ of the northeast. His レな / la 〃ロ is a bull. 6 : 、、 Deity chosen ()y the worshiper himself). " GeneralIy refers [ 0 a family deity. J 鰓 4 れれみ 4 : "Lord of the World"; epithet of a 応 rm of Kma partlcularly popular in Onssa; NP. of a た attendant of the eleventh [ 所 / ta ⅱたロ ra. Jai れ 4 : A religlous doctnne re. 」 ecting the Vedas and revenng twenty-four enlightened beings ( 扉〃 za ⅱ ras or アめ . 襯〃町 4 : A headdress formed ofpiled, matted hair; attnbute, especially Of Siva. J ーれ 4 : " 、行 CtO ロ ous " ; in Jainism, an enlightened being and teacher, a 扉 ra. Ka 4 , 4 ー 0 財れた A mythical mountain ln the HimaIayan range, home of Siva and Pärvati. KäIi : 、 'Power of Time, The Black One" ・ NP. of the great Goddess in a destructive 応 rm ; a た行 of Siva. 4 〃ア〃 ga : 、、 The Käli Age"; the final and most degenerate age of the world, the present age. Ka 縢 : NP. of the tenth incarnation of Visnu, the future world savior who appears on horseback, one Of the イ a 、ⅲ盟 ras. K 川 4 : NP. of the god 0 日 ove. 叩耘 : Oval cup made ofthe top of a skull; attnbute, especially Of deities in their temfying 応 rm such as Bhairava or KäIi. ara れ 44 襯 : "Basket crown"; a conical headdress that resembles a pile offlattened pots, attribute 0 「 Gane'a and mlnor deities. K 嘉 4 : NP. ofthe god of war, a son of lva and Pärväti, rldes a peacock (Kumära, Skanda). Ka 〃襯 i : NP. 0 「 a た行 of Kumära (Kärttikeya), one of the S 叩い m rk . 叩 0 な 4 響 4 襯イ : Att1tude ofstanding with feet together and arms hangmg straight at the sides, charactenstic 0 「な / za ⅱたロ ra 立 K に : NP. 0 「 the descending node ofthe moon, one 0 「 the 〃ロレロ ra ん . Often shown as 2 serpentlne tail. ル体ルなり : 、、 Fame, g10 . K ( 4 ( K れの : 、、 The Dark One"; NP. 0 「 the eighth 酊汀 ra ofVisnu, hero 0 「 the epic M / ra K ァ 4 : Member of the m ⅲい or ruling class. K 川 4 4 NP. 0 「 a beneficent form of the great Goddess, Durgä. Kubera: NP. of the god ofwealth; 市汀信 0 「 the north. His レ汀んの is a man. Ku 襯 4 : "A youth" ・ NP. & 6 a ム 0 K な 4 K 川 4 : "Tortoise ”・ NP. of the second lncarnatlon ofVisnu, 0 「 the d ロレローな ra レの la of the Goddess Yamunä. : A sweet. & 6 4 な 0 襯 0 4. I 襯 i : NP. ofthe goddess ofwealth and good fortune; consort ofVisnu and Other gods. La 川 Näア 444 : Representat10n of Visnu with his consort. I 4 : "The Lord with a CIub" ・ NP. 0 「 a reformer Of the Pä'upata cult; considered to be an avätara 0 「 Siva. 耘〃ー a54 れ 4 : A seated posture of relaxation with one leg pendant and one tucked under the body. 〃 4 : "Sign, S1gn of gender"; amcomc, pillarlike 応 rm of Siva, ongmally his phallus, often placed in a yoni-shaped base. Ma : "The Great God"; epithet of Siva. Ma ″ ira : "The Great Hero ”・ NP. ofthe twenty-fourth Jaina 扉〃Ⅷ i ra ; founder 0 「 Jainism who lived about 540 ( ? ) to 477 B ℃ . 4 みに血 ara : "The Great Lord"; epithet 0 「 Siva. 4 äみに i : NP. of the 'akti 0 「 Siva ぉ Mahe'vara, one Of the Saptamätrkäs. 4 ー 4 んホ ra 川 4 ーれ : Representation 0 「 the Goddess Durgä as slayer of the bu 10 demon, Mahisäsura. 川 4 4 ′ 4 : A mythical composlte sea creature. 襯 4 耘 : A dlagram often used as an aid to meditation or as the basis for an architectural plan; a two—dimensional representatlon Of the cosmos. 4 ′ ( i : NP. ofa . Marici: NP. ofthe Buddhist goddess of the aurora WhO a CharIOt SW1ne. 川 4 物 4 : Cloister, monastery. Ma なア 4 : 、 Fish" ・ NP. ofthe first incarnatlon ofVisnu, one 0 「 the イロⅲ [ ra 立 襯叩ä : "lllusion. 外ーに川 , 4 ー 0 〃れた A mythical mountaln in the Himalayan range, center of the world, location 0 「 lndra's heaven. 川 i れ 4 : "(Loving) pair,' couple in erot1C posture and gesture. 襯 0 面 4 : A (usually round) sweet, attribute 0 「 Gane'a. & 6 ロな 0 耘財 川財イ : An attitude of the body, especially ofthe hands; a hand position with specific relig10LlS meaning. ″ー 4 れわ . A sage, seer. 川朝′〃 : A manifestation or anthropomorphic representatlon Of a deity; a sculpture. れ畩 4 : "Snake" (). ); attribute 0f2 snake; a snake deity or SP1nt. れ畩 i れ i : A female snake deity or spint. Ⅳ 4 れ市 : 'Re. 」 oicmg"; NP. ofthe bull レロ of 、ⅳ a. N ん 4 ( Ⅳ ( ん 4 ) : Man-Iion"; NP. of the fourth incarnation ofVi»u, half man and halflion, one 0 「 the Ⅷ ra 立 Na : "Lord ofDance ' ; epithet of Siva as the cosmic dancer. れ叩鰓 ra 43 ( れ 4 ″鰓の : The nine planets, including the sun (Sürya), moon (Chandra) and the ascendmg and descending nodes of the moon (Rähu and Ketu). N 襯ä 4 : NP. 0 「 the twenty-second Jaina 庁れ / ⅱたロ ra. N ー : 'Misery, dlssolution" ・ NP. of the 市た 2 ロ 0 「 the southwest. HiS レわの ta is a C01 ア se. ゼ 4 畩ⅲアäル i' : Epithet of the ascetic Pärvati performlng the penance Of standmg amld five fires. 4 4 襯“み 4 〃 4 : "Five-faced 〃心 ; ″ロ representing four aspects Of Siva as pro. 」 ecting heads,the 6 h and final as the 〃ロ itself Same as ca ル rn Ⅲ勦ロ〃ロ . ara 血襯 4 : "Räma with the Ax ”・ NP. 0 「 the sixth incarnatlon ofVi}nu, one Of the d ロレ 4 ーな ras. ゼ 4 れみ 4 : NP. 0 「 the twenty-third Jaina 扉〃 za 黻ロ ra , shown shaded by れロ hoods. 277

20. today, 100Se sculptures displaymg chunky Durgä with a Lion bodies and wide, square faces shOW the areas tO have exhibited a styhstlc contlnulty. Probably extenor wall, central offset This relief of Mahisäsuramardini, like provenance unknown, probably south-central Uttar Pradesh the related pieces in the Government (Madhyade'a) Museum, Mathura may have once About mid-l ()th century constltuted the 6 ん adra niche image on a very Sandstone; 48.9 x 36.2 cm small shnne, although it could also have acted The Denver Art Museum, Gift ofElizabeth B. Labrot in memory of as sanctum lmage for a (NO. MarJone Mclntosh Buell; 1983.266 1 ) or even as a 100Se VOtive sculpture. If it did indeed rest ln an extenor wall, it may have been placed in a north-facing niche, as IS usual when Mahi$äsuramardmi appears on the temple of a male deity (Siva or Sürya). On a goddess temple, however, she may have been placed in the 励 ra niche opposite the entrance, thus either on the west wall or, as many goddess temples are onented toward the north, on the south. Sensuous and delicate yet formidable, the great G0ddess Durgä sits enthroned in the posture Of royal ease. Her eight arms fan out around her 伝 om the elbows, beanng a plethora of weapons and ritual 0bJects. prormnent are a mace with gnnmng skull and a dnnkmg cup formed 0f a human cramum that emphasize her connectlon 、 the darker side of the deity. Although carefully arranged curls adorn her forehead, she is crowned with the ascetlc s matted COils of hair like the God Siva. BeIow her knee her lion レん 4 〃ロ t 、 MISts upward, banng its teeth. Durgä is the name 0f the great G0ddess in her supreme 応 rr れ . She is た広 the energy that anlmates the umverse and the gods; without her bOth are inert. She protects agamst adversity and, tO her devotees, the Goddess is the universal creator, all is within her. Like Siva and Visnu, she has many aspects—Käli is her destructive force, Päハ or Laksmi her beneficent, but as Durga alone she cames all stages Of the umverse within her. The ornaments and anatomy Of this sculpture, such as the high globelike breasts and short wa1St, as well as the treatment Of weapons, seem tO relate her tO sculptures 伝 om Madhyade'a, including the somewhat later アⅲ i ( NO. 30 ) 伝 om Kannat1J. R 旧 1 : 刀肥月れ研ん市 4. 1969 : 24 , pl. 10 ; The Asia Society 1984 : 129 , no. 14. 21 < 20 172

ln the context Of the resultant diverse pattern of patronage of religlous establishments, two aspects Of this dimenslon of early medieval society need to be highlighted. First, the process 0 日 ocal state formation, which has been identified in this essay as the maJor process through which "classical" lndia changed to early medieval lndia, resulted in the incorporatlon Of many local cults intO what is 32 generally understood as ″叩な Hinduism. This process may explain how purely local cults and deities came tO be regarded as manifestations of such deities as Siva, Pärvati, or Visnu, who all symbolized suprareglonal Hinduism. This may also tell us Why Täntr1C practices, WhiCh were essentially esoterlc, came tO permeate not only Hinduism but also the erstwhile heterodox systems like Buddhism and 33 Jainism. At one level Of Hinduism as well as ln Other systems, there was tremendous opposltion tO Täntnsm and such nonconformist sects as the Kaulas 34 and Käpälikas. At another level, however, there were elements that came to permeate different sects and rituals. ノ cur1011S convergence Of emotional devotlon ( 励ロた行 ) and adherence to pnmordial Tantra is recognizable in all major religlous systems of the penod. The presence Of common elements in the mldst of diverse relig10LIS practlces can alSO be seen in the second m or area Of the cultural-ideologlcal dlmenslon of the penod. This is reflected in the growth of the institution of the ワロ or イⅢ , the central figure in the 川酊 / za , which often in combination with the temple 35 constituted the sacred complex in a mundane landscape. Royalty almost ー 36 lnvanably subscribed to Brähmanic ideolo ; 施レ - イⅢ宅 a , or the worship ( 2 可の of the deity ( Ⅷ ) , the Brähmapa ( ) , and the preceptor イⅢ ) , was expected. Elsewhere, there could often be opposition to the authonty of the Brähmana and the ideology of varna, but 励黻広 which reflected the bond between an individual and his deity, は S unlversal, and initlatlon 1ntO the correct path was conducted by the 汀ワロ or 4 川 . The growth of Hindu 川口 in the postclassical" penod (which somewhat paralleled the Buddhist and Jaina monastlc orders) , particularly among the Saiva sects, was Of momentous significance. Apart 佇 om being geographically widespread and close to temporal power, the 川酊 / za centers evolved a system Of successlon 0 「汀汀り , as , WhO figure ln eplgraphic records ⅲ terms akin to genealogrs of royal families. The 川酊〃 functioned as dissenunatlon centers and networks Of sectanan and phllosophlcal traditions. Their growmg importance ln early medieval society is attested by the fact that in the kingdom of the KaIacuns of Tnpuri (near JabaIpur in Madhya Pradesh) alone, Saiva temples and 川酊 / la controlled by the saiva Mattamayüra sect were created at Gurgi, Masaun, Chandrehe (Fig. 11 ) , BiIhari, Bheräghät, and other 37 places. Relig10us Centers: Spatial Contexts and SociaI Linkages Relig10us centers—the m or loci for ritual and other actlvity, including architecture, art, dance, and mtlSIC—were, in keeping with the cultural— ideologlcal complexity of the penod, also of diverse type. However, it was the temples which (among a vanety 0 「 religlous centers) had become the common structural features along the contlnuum of geographical space. A modest temple structure might be erected in the density 0 「 a forest or in a hamlet. The range 0 「 spatial contexts Of temples would further encompass villages and urban centers 0 「 vanous dimenslons. ln such places as osian in Rajasthan, Bhuvanesvara in and Pattadakal and AihoIe in north Karnataka, temples dominated and constituted a sacred landscape. lt can be assumed that when a shnne to a deity called "Dweller 0 「 the Forest" @ra 卩 ) , 示ⅲ / ) was put up by a guild in south Rajasthan in the 39 40 seventh century, lt was located far 佇 0n1 a settled area. One can contrast this with the evidence 佇 0n1 a m or urban center of the Gurjara—Pratihära penod, located, in all likelihood, in the LaIitpur distnct 0 「 Uttar Pradesh. The set 0 「 eplgraphic records related tO this urban center called Siyadopipa! い na range in date 佇 om 907 to 967 and show that temples, 0 「 different sectanan afflliations, were an 38

by the SoIa Khambi at Badoh. Built on a knoll overlooking a lake, the structure 叩 parently served as a pleasure pavllion. lt consists 0f an open pillared hall conformmg to the type placed before temples; no mscnptions pertam to the structure, but it may be dated architecturally tO about the tenth century. Our overvle 、 Of ninth- and tenth—century lnscrlptlons suggests that the pratihära monarchs, while actlve in providing land grants, were not involved in comnusslomng temples or lmages. This is supported by the inscnption 0f Bäuka, a pnnce of Mandor (ancient MäVdavyapura), dated Vikrama year 894 (). D. 837 ー 38 ) , which glves a long account of Bäuka's 信 mlly and culminates with a 63 ghoulish descnption 0f his victory in a battle. VenIy when Bäuka danced in the battlefield, putting down his feet on the very entrails and corpses, His fnghtened enemles, like antelope, became ever SO quiet 64 with the strain 行立 / la , 行立 / la. Th1S was strange indeed! The celebration of this victory and the praise of Bäuka's lineage is the sole purpose 0f the inscnption; no temple was built and no grant 0f vlllages made. Bäuka was a Pratihära pnnce but did not belong to the mam impenal line. Dunng the setbacks wh1Ch marked the reign of Rämabhadra (ca. A. D. 833 ー 36 ) , Bäuka asserted independence and laid claim to the 応Ⅱ range of kmgly titles. Though his political ambitions were soon suppressed by Mihira Bh0Ja, Bäuka's inscnptlon IS 65 lmpenal in aspratlon and general character. Records 0f this type were probably once common in the capital at Kannat1J before that city was completely destroyed in the eleventh and twelfth centunes. The fairly wide use of these purely laudatory lnscnptions is evidenced by a eulo (pra ) 0 「 the Rästraküta monarch Krsna III (ca. A. D. 939 ー 67 ). This inscnption has no purpose but the 66 valonzation Of the klng. lts political overtones are evident bOth 伝 0n1 the use Of the Kannada language and its location at Jura in the Dähala country, an area traditionally under the sway Of northern rulers. The noninvolvement of the impenal Pratihäras m temple building can be understood in vanous 、 vays. The elaborate Ved1C rituals, such as the horse sacrifice ( レⅢ怩の and royal consecration ( の , crucial t0 early lndian kmgshlp, represent the most anclent aspect Of the situatlon. These rltes, frequently performed in lndia dunng the tWO centurles before and after Chnst, were seen as propelling the kmg into a heavenly sphere and infusing him with divine power, simultaneously g1Ving him a mandate and placing him above numerous sects and classes Of society. Ritual performance was therefore the pnmary and most appropnate focus of royal patronage and made kings dependent on Brähmins because only they were qualified to carry out the rites. Brähmins consequently had to be supported through village grants and other gifts. Royal sponsorship 0 「 Vedic ntual declined after the 6 h century and regal patronage of temples and lmages clearly emerged under Harsavardhana and hiS contemporanes Nadu and the Deccan. ln northern lndia this innovatlon proved tO be short-lived. lnscnptions show that the Pratihäras stood a100f 伝 om temple building. While Vedic performances were not relnstated, grants to Brähmms continued, suggestlng that preservatlon Of Vedic knowledge through recltatlon 68 ( 〃ア叫の was valued apart 伝 om its apphcation to specific ntuals 川アの・ However, the vedas and attendant srauta ntuals had dechned in importance and had ceased to impmge on day-to-day affairs, beyond their theoretic value as the source Of tradltion and the basis Of cosmogomc and SOCiaI order. The reluctance of the impenal Pratihäras to become involved in temple building can be attnbuted tO Other factors as well. As mentioned earlier, a deity could receive and hold 部応 as a bona fide legal entity. A different relationship was created, however, when a ruler actually constructed a shnne and thus established a divine personallty. The relationship between such shrmes and the ruling clan could be extended tO the P01nt where the deity was seen as the true ruler and the pnnce merely a mlnlster or representatlve Of the god. One Of the best known mstances 0 「 this is the LakuIi'a temple at EkaIingi, which contains the patron-deity 0 「 the 4 日 g. 20 Reused pillars at the entrance Of the Män Mandir, Gwalior, Madhya Pradesh, 9th century. 58

Like the solar deity Sürya ( No. 28 ) in the Hindu pantheon, the Goddess Marici ("shming"), Buddhist dlvinity of the dawn, rldes a char10t. Rather than horses, the chanot is pulled by seven SW1ne. The creatures spread outward belOW her on a tnple-stepped base. A lion, frequent accompamment 0 「 goddesses, JOins the SW1ne (far right). Each animal tramples the contorted body of a male dwarf, possibly representing, as Pal has speculated, "the demons of darkness. ” The severed head of Rähu ( No. 51 ) acts as chanoteer. A 朝た ra appears over his head, multivalent as SOlar disk and wheel of Buddhist law. The monumental goddess, knee flexed in the archer's posture, once bnstled with ten arms. She has three faces, the one emergmg tO her le 仕 , a SW1ne. Four female attendants flank her, none nsmg above her wmst. The outer tWO bear 」 eweled garlands, the inner, fly whisks. The attendant by her nght knee nught be carrymg a ml 江 or , perhaps reflecting the hght of her radiant mistress. The figures are placed against prlasters that once formed a pierced frame. The remnants 0f a lotus halo and small pedestal, probably supporting a subsidlary deity, can Just be discerned behind the goddess's crown. The width of the relief and layout of the frame make it likely that this image stood in a temple or monastery intenor rather than within a はⅡ mche. ln north lndia by the tenth century, Buddh1sm ()n both its Mahäyäna and Täntnc manifestations) was concentrated pnmarlly in the eastern regons of Bihar and Bengal under the patronage of the Päla dynasty. Yet scattered remalns and inscnptlons, particularly around sites histoncally significant tO the rehgon (for example, Sarnath or B0dh Gaya) , testify tO the occasional continuation of belief and patronage in central lndia even 3 as late as the twelfth century. Sanchi, near Vidisha in anclent Dafirnade'a, is best known for the Buddhist 立叩 (reliquary mounds) constructed there around the turn Of the millenmum. Yet, 伝 om remains at the site, lt is clear that Buddhist monuments continued tO be constructed at Sanchi into the tenth century. AIthough these remains deserve concerted study, the stylistic charactenstics Of remnants apparently dating to the late ninth or early tenth centunes are not 1nconSIStent With thiS image. FuII, square faces with small eyes and high, short mouths pushed close to the nose above chins marked by a mound of flesh with 256 no body, spatulate feet, and weighty yet tapermg legs with gently defined musculature charactenze the Sanch1 images as they dO this Marici. Even certain details such as the treatment ofthe goddess's lotus pedestal or the leg drops of the attendants, with a single beaded side, can be found in the mutllated lmages at the site. This is not to say that the Marici does indeed come 伝 om Sanchi. Yet, based on 、 vhat ℃ know 0 「 Other images 伝 0n1 anclent Da'ärnade'a in approximately the first half of the tenth century and the obvious contin- uatlon 0 「 Buddhist patronage at Sanchi, the figure's attnbution tO Vidisha does seem li kely. The single standing temple 伝 om approxnnately the late ninth or early tenth century at Sanch1 (Temple 26 ) shows its Buddhist dedicauon through the images in its 励川 offsets. ln other respects—structurally, iconographlcally, and stylistically—it utilizes forms and formats common tO Hindu shnnes in the area, yet another example 0 「 the commonallty Of craft traditions among 5 rehglous sects. R 肥れ : Pal ( 肥 & 如川 0 血な ) 1978 : 69 ー 70 , no. 38 ; Heeramaneck 1979 : no. 89. 1. 2. 3. 4. 5. Pal, 7 e & 如 1 川川 0 加な : 月 & lec ″研 & “ tu 五 0 川〃肥 P 明 - 月豆明 Co 〃“れ , p. 70 , no. 38. ln this 応ロ 11 she is at tlmes considered た行 Of Hayagnva. Generally she is た 0f the Dhyäm-Buddha Vairocana, ongmally a solar deity. See Liebert, ”四叩帰 ( D 加ワ研どん市明 Rel 0 , p. 174. See N. N. Das Gupta in 7 e & ru. 面 r E ″甲 , V01. 5 , 刊ど H ⅳ 0 ヮ and C ル肥研〃肥ん市明叩ん (Bombay: Bharatiya Vidya Bhavan, 1979 卩 95 刀 ) , pp. 421 ー 24. Temple 26 and its adJacent monastery date t0 叩 proxlmately the late mnth or even early tenth century, and scattered fragments in the VIC1mty show carvlng typical ofeven later penods. The area around Vidisha may have been under the pohucal control ofthe Paramära dynasty when this image was carved (Nos. 2 , 68 ).