Distinction in Indian Iconography [Part II, 2]

The paper was published in Bulletin of the Deccan College Research Institute 20 (S.K. De Felicitation Volume), Poona 1960, pp. 164-248.

Distinction in Indian Iconography

Part II: The Theory

IX. Transformation in Art and Transformations in Literature

§ 57. “Discrepancies between texts and monuments“ can be understood in a literal and in a general sense. In the first case, the specific statements of the texts are contradicted by the monuments (see the Introduction on the Pārśvanātha temple). In the second case, the monuments are only at variance with the general spirit of the texts: these are scholastic, stressing the differences rather than neutralizing them, while the artists obliterate the differences (see the Introduction on images of the Jain cakravartin). Roughly speaking, discrepancies of the first type are due to amalgamation, those of the second type due to stereotype representation. In this chapter we are only concerned with the literal discrepancies.

§ 58. While emphasizing the differences between texts and monuments we should not lose sight of two facts which are relevant to the problem. Firstly, the difficulty of tracing in each case the particular text which determined the iconography of the period and the province concerned. Unless such a text is available we are, strictly speaking, not in a position to call in question the correctness of a representation. That a northern monument for example does not conform to a southern text is but natural.

Secondly, we have to take into consideration the existence of a body of symbolical, speculative, or even mythological knowledge which is not reflected in written documents. Art may reflect in certain cases old traditions which did not reach us through the channel of literature.

But even though we admit that the non-literary tradition has contributed to the development of iconoplastic art we do not think that art expresses religious ideas wherever we may find a remote relation between both sides. No doubt, in the case of “syncretistic icons“ for example we have reason to believe that iconography reflects certain speculative conceptions (§ 33; Banerjea, p. 5 f., 540 ff.). But in other cases (for example if we compare the association of cakra and lotus in art with the association of cakra and lotus in literature) we should not rule out the possibility that the same objects are associated with each other independently and for different reasons in different spheres. This is only natural when the general mental outlook favours associations of all kinds (see § 69). Even general similarities (e.g. assimilation of attributes as compared with speculative identifications) should not be explained by a direct influence of religion on art but by the fact that Indian religion and Indian art are both expressions of the same national genius.

§ 59. If a monument does not conform to the relevant iconographic text we can assume that the artist departed from the tradition. But agreement between iconographic texts and monuments is neutral with respect to the question of artistic innovation or dogmatic tradition because the monuments themselves have influence on the texts (i.e. on the iconographic texts, rarely on mythological literature in the strict sense). Still it may be useful to mention a few examples of stereotype representation and amalgamation which we find in our documents.

Not only do we come across names for cult amalgams (Hari-Hara, Ardhanārīśvara, Sūrya-Nārayaṇa, Hari-Hara-Pitāmaha) and amalgamated attributes (vajraghaṇṭa), there are also explanatory myths which justify the existence of such forms. Banerjea says (p. 548):

“Features of the Sun god are traceable in the representations of many of the cult deities. This is naively explained by the myth-maker by saying that from the “parings“ of the resplendent body of the Sun (the effulgent body of the god had to be trimmed by his father-in-law Viśvakarmā in order that his daughter Saṁjñā, the principal consort of Sūrya, could bear her husband's company) many characteristic traits by way of weapons, attributes, etc. were made for the other Brahmanical Hindu deities.“

Besides this myth which is remarkably abstract we have others which deal only with Hari-Hara (Rao, II, I, p. 332) and Ardhanārīśvara (Rao, II, I, p. 322 f.).

Stereotype representation is expressed by iconographic terms like jaṭāmaṇḍala (besides prabhāmaṇḍala) and jaṭāmukuṭa (besides the normal mukuṭa). The terms refer to the arrangement of jaṭās in a circle (fig. 48) or in the form of a mukuṭa (figs. 59-61). Broadly speaking, our four drawings are demonstrations of jaṭāmaṇḍala and jaṭāmukuṭa, but that does not mean that the textual descriptions enjoin stereotype representation in the extreme form in which it is found on the monuments.

According to the Uttarakāmikāgama (Rao, I, I, terms, p. 28), jaṭāmukuṭa is a complicated arrangement of jaṭās, but we are not told that it should look like a tall crown (the normal sense of mukuṭa). Stereotype representation may also be expressed in the form of a simile. It is said that the elephant-skin spread behind the Gajāsurasaṁhāramūrti of Śiva should look like a prabhāmaṇḍala: prabhāmaṇḍalavac cheṣaṁ gajacarma[ṁ] prakalpayet (Aṁśumadbhedāgama 70, quoted after Rao, II, II, pratimālakṣaṇāni p. 77).

§ 60. Iconographic terms as mentioned in the last § call for a few remarks. Since modern scholars sometimes add new names to the old ones (e.g. “Hari-Hara-Sūrya-Buddha“, Banerjea, Pl. XLVIII, 1), one is not always sure, whether such a name occurs already in the ancient texts or is a recent invention. The same is true of similes. Banerjea, while describing a Gupta image of Skanda, says that “the outspread tail of the peacock serves as his prabhāvalī“ (p. 366), and it is clear from the context that this is an expression coined by Banerjea himself. But we do not know whether Rao or his sources describe Agni's hair as “standing on end and forming a sort of prabhāmaṇḍala“ (Rao, II, II, p. 523 ; the texts quoted by Rao do not use the word prabhāmaṇḍala in this connection).

At the outset iconographic terms have to be taken as merely describing certain elements of the representation. We can neither infer from the existence of the mere word that such a thing (person, etc.) exists in reality or in mythology, nor can we treat each term with suspicion and explain the underlying representation simply as a caprice of the artist. The “reality“ of such terms is always a matter for further investigation.

It is hardly necessary to add that iconographic terms are not restricted to cases of stereotype representation and amalgamation. We mention only mahāmbuja or viśvapadma “a double-petalled lotus, one set of petals pointing upwards and the lower set gracefully drooping down“ (Banerjea, p. 299). mahāmbuja serves mainly as a pedestal, but is also found as a hand-attribute (fig. 85).

mahā- and viśva- are common adjectives in iconographic terminology and distinguish a more comprehensive or a more elaborate form from the normal form. E.g. Mahālakṣmī (Banerjea, p. 496), Mahāgaṇapati (Rao, I, I, p. 55), Mahāviṣṇu or Viṣṇu Viśvarūpa (Viṣṇu with three or four heads), Maheśa (Rao, II, II, p. 379), Mahāsadāśiva (Rao, II, II, p. 373 f.).

§ 61. In the case of most types of transformation (e.g. explicit amalgamation) we find neither full agreement nor complete disagreement between texts and monuments. It is therefore equally possible to believe in the absolute authority of the texts and to regard the artistic development as autonomous. In the first case one would refer all transformations to the texts and explain the absence of literary evidence by the loss of the respective documents. In the second case one would explain the transformations as artistic phenomena and consider the textual references as the reflex of the artistic development.

It goes without saying that both solutions are one-sided. In reality an “osmosis“ of artistic and literary transformations takes place; the same tendencies can be observed in the field of art and in the field of pure imagination so that it becomes often difficult to decide which side is responsible for a particular transformation. Mythologists and iconographic experts contaminated different gods, but the artists themselves added to the general syncretism, transferring attributes from one figure to another and so forth. Examples for this osmosis are numerous and have been mentioned repeatedly in the course of the article. See §§ 7, 33, 40, 44.

§ 62. The relation between monuments and texts has not yet been critically studied.

A vast iconographic literature has come down to us, and it was tempting to regard this literature as a sort of never-failing dictionary which would enable the student to “read“ without difficulties the iconographic “text“ in the form of the monuments of the medieval period. Thus it became customary to proceed from the study of the texts to the study of the monuments and not the other way round. Under these circumstances, differences between texts and monuments were not likely to be viewed in the proper light. Scholars like Gopinath Rao noticed again and again such differences. But normally everything that was not in harmony with the descriptions in the available texts was either automatically considered as representing some unknown literary tradition or it was neglected. In other words not only the identification but also the selection of the material to be studied was based on the iconographic literature.

This limitation in the scope affected especially two sections of the iconographic material. On the one hand, one did not study carefully figures other than icons proper; scholars did not investigate statues appearing on outer walls, door-frames or as subsidiary figures on greater images. On the other hand, the elements of a composition were examined only to the extent to which it was necessary for the identification of the main figure itself. If this figure answered in broad lines to the description of the iconographic texts it was not found necessary to check every detail. One was satisfied to demonstrate some general correspondence between texts and monuments but did not ascertain in each case and for each element the exact measure of agreement or disagreement between both sides. Cf. Also § 54.

Nowadays it is admitted that the sculptures can no longer be studied as illustrations of the texts. Still the old method was in vogue for some time, and these critical observations may therefore be useful for the assessment of the earlier literature.