Home > Cultural Informatics > Visvarupa > The Visvarupa Iconographic Traditions - The Archaeological Reports of Professor T. S. Maxwell 

Viśvarūpa

Vaikuṇṭha-Viśvarūpa Vol. I

D:alphasitereports-2/Ti

FELDFORSCHUNGSPROJEKT 1989/90

DFG-AZ.: Ma 1069/3-l

Kennwort: Vaikuṇṭha-Viśvarūpa

 

 

BERICHT

 

UBER DIE IKONOGRAPHIE, CHRONOLOGIE UND INTERPRETATION

DES ARCHAOLOGISCHEN MATERIALS

 

 

T S Maxwell

 

Alle Rechte der Vervielfaltigung und Weiterverwendung des

Inhaltes dieses kataloges sind dem Verfasser vorbehalten

Copyright (c) T. S. Maxwell 1990

 

 

D:alphasitereports-2/Ti

FELDFORSCHUNGSPROJEKT 1989/90

DFG-AZ.: Ma 1069/3-l

Kennwort: Vaikuṇṭha-Viśvarūpa

 

INHALT

 

i.

Einleitung zum Bericht

ii.

Forschungsbericht:

Einseitige Kurzfassung der Hauptergebnisse

iii.

Englische Zusammenfassung mit

zusatzlichen Schlubfolgerungen

1.

Teil I:

Bericht uber die Ikonographie, Chrono-

logie und Interpretation des archao-

logischen Materials

1.1

Landkarte, Region 1

1.2

Skulpturen des Vaikuṇṭha- und Viśvarūpa-

typus in Region 1

1.3

Chronologische Entwicklung

1.4

Teil 1.1:

Sonderberichte

2.

Teil II:

Bericht uber die Ikonographie. Chrono-

logie und Interpretation des archao-

logischen Materials (Region 2)

2.1

Landkarte, Region 2

2.2

Ursprung und Evolution der Viśvarūpa-

Ikonographie in Region 2 und die Weiter-

entwicklung in Region 1 (3.-6.Jh.)

2.3

Die spatere Entwicklung der Viśvarūpa -Ikonographie

in Region 2:Evolutionen A und B

2.4

Teil 11.1: Sonderberichte

 

EINLEITUNG ZUM BERICHT

Das vorliegende Feldforschungsprojekt und seine Fortsetzung ist auf eine ausfuhrliehe Dokumentation des fur die Entstehung, Entwicklung und Ausbreitung der Vaikuṇṭha- und Viśvarūpa- Kunsttraditionen relevanten archaologischen Materials in Kashmir, Himachal Pradesh, Uttar Pradesh, Rajasthan, Gujarat und Madhya Pradesh angelegt.

Abschlubziel der Arbeit ist die Darstellung einer Gesamttheorie fur die Analyse und Interpretation dieser komplexesten kosmologischen Ikonographie Nordindiens des 5.-12. Jarhrhunderts und deren Integration in den existierenden Wissensehaftskorper der Disziplin Orientalische Kunstgeschichte, wobei die chinesische und eventuell die sudostasiatische und japanische Ikonographie in Betracht gezogen werden konnte.

Die Auswertung der Dokumentation wird an der Universitat Bonn durchge-fuhrt. Zum Teil wird dieses Forschungsvorhaben als Erweiterung der van Wlllibald Kirfel, dem ehemaligen Ordinarius fur Indologie an der Univer-sitat Bonn unternommenen Arbeit (Die dreikopfige Gottheit, Ferd. Dummlers Verlag/Bonn 1948) betrachtet, zum Teil als Fortsetzung der an der Universitat Oxford begonnenen verwandten Arbeit des Verfassers selbst (Thomas S. Maxwell, Viśvarūpa, Oxford University Press/New Delhi 1988).

Die Feldforschungsarbeit wird zeitlich und geographisch in mehrere Forschungsgebiete aufgeteilt, wobei Region 1 aus Rajasthan und dem nordlichen Teil Gujarats und Region 2 aus dem sudwestlichen Uttar Pradesh und dem nordwestlichen Madhya Pradesh besteht.

Weitere Forschungen zur Untersuchung des Themas sind fur die als Region 3 (Himachal Pradesh), Region 4 (Kashmir), Region 5 (Zentral-GuJarat), Region 6 (Zentral-Madhya Pradesh), und Region 7 (ostliches Uttar Pradesh) bezeichneten Gebiete geplant.

Die erste Stufe des Projektes, Untersuchung des archaolologischen Materials in den Regionen 1 und 2 einschlieBlich einer Auswahl der dort beflndlichen hinduistischen Tempel. wurde 1989 mit Untersttitzung der Deutschen Forschungsgemeinschaft durchgefuhrt. Fiir diese Gelegenheit, die beiden Themen Viśvarūpa und Vaikuṇṭha weiterzuerforschen und die theoretische Basis dieser Untersuchungen zu verstarken, bin ich der Deutschen Forschungsgemeinschaft sehr dankbar. Der Rektor Magniflcus der UniversitKt Bonn hat mir zu diesem Zweck ein Forschungsfreisemester (das Wintersemester 1989/90) genehmigt.

Die Anordnung des Bildmaterials ist im Februar 1990 in Bonn begonnen worden: diese Arbeit konnte noch nicht abgeschlossen werden. Die Schreibarbeit wurde im Verlauf des Sommersemesters 1990 unternommen. Mit Ausnahme des Sonderberichtes uber Suhania wurde der nachstehende Bericht in den Monaten Marz bis Juni 1990 in Bonn aus den ikonographischen Feldforschungsergebnissen und einem Teil des Bildmaterials vorbereitet. Der Haupttext bleibt in englischer Sprache fiir eventuelle Publikations-zwecke; Schlubfolgerungen sind, soweit dies mir in dem vorhandenen Zeitraum und ohne Mitarbeiter moglich war, auf Deutsch aufgezeigt.

T. S. Maxwell

Bonn, den 30. Juni 1990

 

FORSCHUNGSBERICHT: EINSEITIGE KURZFASSUNG DER HAUPTERGEBNISSE

(vgl. englische Zusammenfassung mit zusatzlichen Sehlubfolgerungen)

 

1.

Die literarische Basis fur die Entstehung in Mathura sowie in Samalaji des Viśvarūpa- Bildwerks ist das Mahābhārata; im Bhagavadgītā wird diese kosmische Offenbarung Kṛṣṇas van drei Personae des Epos funfmal beschrieben.

1.1

Im 10.Jh. wlrd die Identitat der Hauptflgur der Viśvarūpa-Ikonographie. namlich Kṛṣṇa, durch andere Erseheinungsformen Viṣṇus, z.B. Rama. ersetzt.

2.

Die asthetlsche Basis des Hauptcharakterlstikums des Viśvarūpa-Bildwerks, d.h. der Vereinigung selner ikonographisch unterschiedllchen Kompositionselemente, besteht aus Weiterentwicklungen der Emanationsikonographie des kusanazeitlichen Mathuras.

2.1

Diese asthetische Basis hat als Grundlage aller Viśvarūpa-Bildwerke in Regionen 1 und 2 bis in das 10.Jh. uberdauert.

2.2

Im 10. Jh. ist eine andere, architektonische Vereinigungssttruktur in die Viśvarūpa-Ikonographie in Region 2 sudlich des Ganges eingefuhrt worden; im 11.Jh. ist diese Struktur in Region 1 ubertragen worden.

2.3

Bis zur EinMhrung der architektonischen Vereinigungsstruktur gelten Viśvarūpa- Bildwerke auf Grund ihrer alleinigen Fortsetzung der kusanazeitlichen Emanations-struktur als eine besondere Nebenschule in der Geschichte der hinduistischen Plastik.

3.

Erst nach der Entwicklung der archltektonlschen Vereinigungsstruktur konnen Viśvarūpa- Bildwerke als Elemente van ikonographischen Programmen an Tempelwanden betrachtet werden, und zwar erst ab dem 11.Jh. in Region 1.

3.1

Im 5.-10.Jh. sind alle Viśvarūpa- Bildwerke als alleinstehende Kultbildnisse fur Einrlchtung in eigenen Sanktuarien zu verstehen, und zwar auf Grund ihrer Form, ihrer komplizierten und in sich eingeschlossenen Symbolik. und des ausdrucklich kosmischen Charakters des symbollsierten Themas.

4.

Die Entwicklung der Viśvarūpa- Ikonographie weist im 8.-10.Jh. zwei verschiedene Evolutionsrichtungen auf (in diesem Bericht als Evolutionen A und B gekennzeichnet).

4.1

Diese beiden Evolutionsrichtungen entstehen nlcht zeitgenossisch mit der Entstehung der beiden ursprunglichen Versionen des Viśvarūpa-Bildwerktypus im 5. und 6.Jh. in Mathura und Samalaji, sondern im 8.Jh. in Region 2. als die ikonographischen Einflusse der letztgenannten beiden Fruhversionen schon vereinigt gewesen waren; es handelt sich urn eine zweite Verzweigung der Tradition.

5.

Das dreikopflge, auf Garuḍa thronende Vaikuṇṭha-Bildnis mit acht Handen entsteht im 9.Jh. in Region 1.

6.

Im 10.Jh. in Region 1 entsteht eine sitzende. halbmeditierende Vaikuṇṭhaform als Entwicklung des Yoganārāyaṇathemas.

7.

Im ll.Jh. in Region 1 entstehen 8 weitere Valkuṇṭhaformen, die in diesem Berieht dokumentiert werden.

8.

Die Vaikuṇṭhaform wird im 11.Jh. in einer Gegend van Region 1 als Kultbild des Śāktakultes weiterentwickelt.

 

ENGLISCHE ZUSAMMENFASSUNG MIT ZUSATZLICHEN SCBLUSSFOLGERUNGEN ENGLISH SUMMARY AND CONCLUSIONS

Viśvarūpa is a violent cosmic vision of the Hindu god incarnate Kṛṣṇa, which Kṛṣṇa himself reveals to the warrior Arjuna on the eve of the final battle in the war of the Mahābhārata. This vision is described in the Bhagavadgītā. The tenth and eleventh books of the Gītā contain five descriptions of Viśvarūpa of which two are spoken in the first person by Kṛṣṇa, another is given in the third person when Saṅjaya describes Arjuna's vision to the blind king Dhṛtarāṣṭra, and the other two are in the second person, being spoken by Arjuna' to Kṛṣṇa as he experiences the vision. Several hundred years after the composition of these descriptions, sculptors at the artistic centres of Mathura and Samalaji began the attempt to depict Viśvarūpa in sculpture as a cult icon.

There were obvious aesthetic and compositional problems involved, and at the same time fairly fixed iconographic rules had to be obeyed. At Mathura, an existing image of Kṛṣṇa as Viṣṇu, having the extra heads of two of Viṣṇu's animal incarnations, as a man-lion and as a boar, was used as the central figure. This tricephalous image was later developed in Kashmir and the western half of North India, to represent a different cult form of Viṣṇu known by the abbreviated title of Vaikuṇṭha. Both types of image - Viśvarūpa and Vaikuṇṭha - continued to be made and worshipped in northern India at least until the 11th century. This research project is the beginning of an attempt to understand these various iconographic forms, their development, and their influence during the approximately 800 years of their existence.

At the basis of the earliest Viśvarūpa iconographies (5th century in Mathura, 6th century in Samalaji) are the Kuṣāṇa-period emanation iconographies of Mathura sculpture (ca. 3rd century). These provide the aesthetic framework for the iconographic depiction of the main characteristic of the Viśvarūpa concept, namely its unification of multiple forms in a single image (see Part 2 of this Report, Origin and Evolution, Sections A, B, E). The development of these Kuṣāṇa-period Mathuran iconographies at Samalaji is explained by the exportation of early sculptures having emanatory forms from Mathura southwestwards along the edge of the Thar Desert, by way of Bairat and Pushkar, into the northern part of modern Gujarat.

The outstanding piece of evidence for this migration of symbols is the Kuṣāṇa-period populated column at Nand near Pushkar. Although the outskirts of Nand village are today being pushed westward by the erection of new temples, the column appears originally to have been erected at the boundary between cultivated land and desert, only a few hundred yards from the periodic course of the Luni River in Rajasthan. The date of its erection at Nand is unknown (it is regarded locally as having been born from the earth at an indeterminate date in the recent past, and was not remarked by Tod when he passed the very spot in 1819); the reason for its establishment there may be connected with the importance of Pushkar as a tirtha and with the related significance of surrounding villages as stations on the pilgrimage route which defines the Puṣkara-kṣetra. Further research is required to discover more evidence of the exportation of Kuṣāṇa-period sculpture from Mathura along this route.

The Viśvarūpa fragments at Mathura and the series of Viśvarūpa images and fragments from Samalaji, in their present condition indicate that the emanation iconographies were first developed further, not at Mathura itself, but at Samalaji after a time lapse between the 3rd and 6th centuries. The time lapse requires further investigation to account for this dormant period in the activity of the migrating symbols. The Mathura - Samalaji connexion is demonstrated by the vertical and branching emanatory forms comprising the nimbus of the Samalaji Viśvarūpas. Such interconnected forms are scarcely detectable in the Mathura Viśvarūpa fragments, which from fragmentary evidence appear to have concentrated on representing the innumerable devouring faces of Viśvarūpa by a peripheral chain of Śiva-Bhairava heads around the nimbus, plus rows of ascetics upon its enlarged surface, depicting the hosts of sages and saints who disappear into the universal fire. The literary basis for both Viśvarūpa types was the series of five descriptions of Kṛṣṇa as Viśvarūpa contained in the 10th and 11th adhyāyas of the Bhagavadgītā as it now exists (see Part 2 of this Report, Origin and Evolution, Section E. Gītā Index-1 and -2, and Sections F and G).

The design of the Samalaji Viśvarūpa series, as local interpretations of the Bhagavadgītā or an allied text, was clearly based on an equally localised re-employment of Kuṣāṇa-period emanation motifs. The Samalaji Viśvarūpas are not logical developments of Gupta-period iconographical constructs, from which they were apparently isolated. Gupta elements relate to style, not structure. This shows that individual sculptures of Gupta style were influential in northern Gujarat, not iconographical combinations such as the proto-'Vaikuṇṭha' form in which the heads of Varāha and Nṛsiṃha were added to define the Viṣṇu image.

The Samalaji sculptures depict the multiple heads ('mouths') of Viśvarūpa, not by means of Bhairava masks as at Mathura, but by a conversion of caturmukha Brahmā imagery, in effect using a triple-crowned Brahmā, with three heads in human form, to represent the central figure of the Viśvarūpa image. A possible relationship between this redeployment of Kuṣāṇa Brahmā iconography (complete with the vertically emanating form reinterpreted as Hayagrīva) and the Pra-Pitāmaha concept of the Bhagavadgītā is marginally noted. The seated posture of this figure at Samalaji has female precedents in the Kuṣāṇa Mathura iconography and a terracotta proto-'Vaikuṇṭha' of Gupta date is also known, showing perhaps a parallel development in Vaiṣṇava iconography at both sites, which was taken up chiefly at Samalaji. The main themes of the nimbus are an evolutionary cosmogony (the Vaiṣṇava axis) and the generation of gods and heroes. The base of Nāgas, perpetuated as an important element of later Viśvarūpa images across North India, seems to have been a Samalaji innovation.

The combined influence of the Samalaji Viśvarūpa series and that represented by the Mathura fragments, is seen in the iconography of later Viśvarūpa sculptures father east in Region 2 between the 6th and 9th centuries. In Region 1, only the dying echo of Samalaji is definitely known for this period, in the damaged Viśvarūpa sculpture at Kathlal, although this Report shows some evidence of later forms of Viṣṇu having three human heads. Western Indian Viśvarūpa sculptures (i.e. in Region 1) dating from the 7th and 8th centuries therefore must exist, and research to find them and to chart their iconographical evolution is required. The development of Viśvarūpa iconography in Region 2 during this period, and beyond, is better known and has been traced through two evolutionary processes in Part 2 of this Report (see Evolutions A and B). This shows a definite trend toward increasingly architectural structures, which appeared in the 10th and 11th centuries south of the Ganges, as the Kuṣāṇa emanation-aesthetic was forgotten and finally superseded.

Between the 5th and 9th centuries, the Viśvarūpa sculptural tradition represents a persistent minor theme within the major stylistic traditions of Northern and Western India. Because an unusual aesthetic structure (basically consisting of the populated nimbus) was required to compress many miniature images together in a single icon, the relatively rare representations of this theme almost amount to a sculptural sub-school. It was not until the rise of the architectural structure for these icons in the 10th century that Viśvarūpa iconography was finally integrated with the main North Indian traditions. The architectural frame of these later Viśvarūpa icons closely resembles a temple doorframe, so that only from the 10th century onward does it become possible to consider Viśvarūpa images as panels for the bhadra-niches of temples.

No Viśvarūpa image dating from before the 10th century has been found within the iconographic programme of a temple, and only the very latest of the 9th-century images (Bhuili: see T. S. Maxwell, Viśvarūpa, Oxford/New Delhi 1988, p.269 and P1.70) is sculpted on a rectangular slab suitable for placement in a bhadra-niche. The possibility therefore arises that in the first 500 years of their production, Viśvarūpa images were not intended for installation on temple walls, but were independent cult icons which may have been set up in separate small shrines of their own. This would reflect not only the negative or potentially destructive character of the theme, but also the self-contained imagery of the icons, which makes them unsuitable as units in an extended iconographic programme. This was not true of Vaikuṇṭha icons, which were definitely designed, from the beginning of their appearance in the 9th century, for placement in temple niches (see Sculptures 15 and 16 in this Report, Part 1).

The architectural construct evolved for the later Viśvarūpa images re-entered Region 1 in the 11th and 12th centuries, when it also became involved with later Vaikuṇṭha iconography (compare the shape of the stele of the 10th-century Suhania Viśvarūpa with that of a 12th-century three-headed Viṣṇu published by Pal, Indian Sculpture 2, Los Angeles/Berkeley 1988, p.56, no.56).

The development of Vaikuṇṭha iconography in Region 1 and its various cult contexts has been traced in Part 1 of this Report. Discounting the images of Viṣṇu with the side-heads of lion and boar (Proto-'Vaikuṇṭha') created at Mathura in the 5th century, the archaeological evidence examined shows the following major evolutionary stages.

Stylistically, the earliest three-headed and eight-armed Vaikuṇṭha image that was found in Region 1 is datable to the 9th century (Sculpture 16, from the region of Jhalavadh in eastern Rajasthan, also published by K. Desai, Iconography of Viṣṇu, New Delhi 1973, p.44 and Fig.39), which probably follows closely on the adoption of this deity as the royal god of Avantivarman at his first temple, the Avantisvamin at Avantipura in Kashmir circa 850 AD. The Avantisvamin bronze is a standing figure, while the Jhalavadh sculpture shows three-headed Viṣṇu seated on Garuḍa.

The texts relating to these icons and meditational images are discussed in K. K. Dasgupta, "The Pāṅcarāta Tradition and Brahmanical Iconography" (Shastric Traditions in Indian Arts, edited by A. L. Dallapiccola, Stuttgart 1989, 71-91), and in T. S. Maxwell, 'Vaikuṇṭha' (Festschrift for Debala Mitra, pp.1-29 and Sanskrit Appendix, in press since 1988). This basic difference in the iconographical depiction of Vaikuṇṭha in separate regions at the same historical moment suggests two spontaneous reactions to a single religious impetus. The precise nature of the connexions, and differences, between Vaikuṇṭha iconography in Kashmir and Rajasthan in the 9th century requires explanation through careful re-search in Kashmir itself, where a number of such images have been found.

The partial integration of Yoganārāyana iconography with that of Vaikuṇṭha occurs in the 10th century, to produce a Vaikuṇṭha image seated in a near-meditation posture, but with the front hands still holding two attributes loosely in front of the folded legs. This form is evidenced at two sites, Nilakantha and Candravati, indicating a fairly widespread phenomenon. The Nilakantha material is removed from its original context and has almost perished due to erosion; the slightly later Candravati image appears on the north bhadra-niche of a small tantric Vaiṣṇava temple.

In the 11th century at least eight new types appear. Three of these occur on a single temple, the sāsbāhu at Nagda. Of these, two occur within architectural frames at the base of the main doorway śākhās; they are both standing images which hold the standard eight Vaiṣṇava attributes, though in differing order, and they are associated with different avatāras and gods, namely: Nṛsiṃha. Vāmana/Trivikrama, and Śiva on the one hand; and Varāha, Viṣṇu Gajendramokṣa, and Viṣṇu on the other. Specific interpretations of the avatāra-heads seem to be implied, but this requires further research. On the south wall of the same temple, a version of Garuḍa-mounted Vaikuṇṭha appears which holds the attributes of Yama and Brahmā in addition to those of Viṣṇu.

On the 'Mira' temple at Eklingji a self-contained iconographic programme on the north, east, and south walls of the sanctum exterior show forms of Vaikuṇṭha with eight, twelve, and sixteen hands respectively, which may correspond to the three icon types described in the Aparājitapṛcchā as Vaikuṇṭha, Ananta, and Trailokyamohana.

The other two 11th-century developments stem from the Undesvara temple at Bijoliyam. More obviously than the Candravati image, these might appear to relate to a Tantric or Yogini cult; on the basis of the iconographic programme of the Undesvara temple, I have suggested in fact that they were taken over by a Śākta cult to represent aspects of Pārvatī and Śiva - a total cult reversal of the original icon. Their most striking feature is the replacement of the central face of Viṣṇu with that of a horse. One of these images is male, the other female, and it is the iconography of the goddess which is the more elaborate. These images are discussed, with their iconographic context, in an attached Special Report. Another, severely damaged version of horse-headed Vaikuṇṭha, seen in Cittaurgadh, has the position of the avatāra heads reversed and appears to have been seated (the legs are broken) like the Nilakantha and Candravati images. This was first reported by R. C. Agrawala (Note 3 in the attached Special Report) but not interpreted. It appears to represent the same god as that now affixed to the Mahakala temple at Bijoliyam, in a different posture. All three sculptures stem from the stretch between Cittaur and Kota, and none of the same type were found outside this predominantly Śaiva region.

In conclusion, certain iconographical tendencies are worth a brief notice. All the Vaikuṇṭha images datable to the 10th century and later are represented without a nimbus. Three of the images have the usual position of the animal heads reversed, the lion appearing on the left and the boar on the right; in these examples all the hands are destroyed, making the analysis of any further iconographical variations impossible. But they are of variable date (9th/10th, 10th/11th, and 11th centuries) and posture (seated on Garuḍa, standing, seated on a throne), suggesting no consistent variant form of Vaikuṇṭha. The disappearance of the nimbus from Vaikuṇṭha images approximates in date (10th century) to the disappearance o f the populated nimbus from Viśvarūpa images and the introduction of an architectural framework for the support of the many minor figures. At the same time, a growing tendency is noticeable in succeeding centuries to identify the main figure, in images of both the Vaikuṇṭha and Viśvarūpa types, with forms of Viṣṇu other than Kṛṣṇa notably with Hayagrīva or Rāma, and even a coalescence with Śākta imagery was discovered in one area.

D:alphasitereports-2/Region1.vk

TSM 1989/90

 

[ Previous Page | Next Page ]

 Report Volumes  | Viśvarūpa Multimedia Project

HomeSearchContact usIndex

[ Home | Search  | Contact UsIndex ]

[ Sutradhara | International Dialogue Unit | Multimedia Development ]


Copyright (c) T. S. Maxwell 1990