Distinction in Indian Iconography [Introduction]

---

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

 

Preface

The present article emerged from a short paper “On the relation of form and contents in Indian iconography“ read at the last International Oriental Conference at Munich. Since then the material has been reorganized, and the résumé of my talk as published in the transactions of that conference ^[1] does no longer represent my views. To call the new version final would of course be presumptuous. The system, as it now stands, is no doubt capable of improvement. It is also conceivable that in the course of time a different scheme is developed which does better justice to the material than this one. Nevertheless some new problems have been brought into focus, and others have been discussed with more energy than hitherto. And although no scholar is likely to adopt my system for his own presentation, I do hope that the article may serve as a stimulus to similar enquiries.

I feel much obliged to the Archaeological Survey of India for their ready help in supplying a great number of photo prints. The references to these photographs will be found either in the text or in the “List of Drawings“. Some of the drawings were prepared after printed works and the references to these publications will also be found in the “List“. Photographs without reference were taken by myself. Those of my photos which are mentioned in the text but not reproduced as drawings have been quoted along with the serial number they have in my collection.

I also wish to express my indebtedness to Mr. J.F. Oats and especially to Mr. S.A. Srintvasan who were kind enough to check the English text. Finally, I have to thank Miss I. von Auer for the careful preparation of the drawings.

 

Abbreviations

Introduction

I

While studying the iconography of some medieval monuments in Central India I came across numerous cases where the identification of a figure was difficult, if not impossible. There could be little doubt that these difficulties had to be attributed at least partly to the artists' peculiar way of presenting the motifs, and that only an examination of the methods of the artists would solve the problems involved. A description of those difficulties which I came across in the beginning of my iconographic studies, and which in fact stimulated the present enquiry, may open the discussion.

Most of the figures on the outer walls of the Pārśvanātha temple at Khajurāho which I had to study some time ago defied all attempts at identification. They had no common features with the better known gods of the Jain pantheon, nor any with those that are less known. Many of these figures showed not only iconographic features of different gods combined but it was even impossible to classify them as Hindu or Jain. Therefore the conventional method of identification proved useless here and it became necessary to find a new way for the description of the material. Those figures which were more or less similar to Śiva, Viṣṇu, and Sūrya I marked in my article on the temple with the symbols “I“, “II“, and “III“ respectively, thereby merely indicating the predominance of the Śiva (or Viṣṇu, or Sūrya) element. The remaining figures (as far as they were furnished with attributes) I partly classified according to sporadic similarities (marking them with further Roman numerals), and partly left them without any definition (marking them with a question - mark). Only very few figures could be identified as recognized Jain deities. Figures without iconographic characterization (e.g. lady removing a thorn) or with very faint characterization (e.g. lady with a lotus as the only attribute) were marked with a zero. But in the second case (faint characterization) the question - mark would have been more appropriate, because there is no clear-cut line of demarcation between “full“ and “faint“ characterization.

Besides icons of hybrid character there were a few hybrid attributes. Such attributes, too, could not be identified; they had to be described by artificial names composed of the names of the blended elements: śakti - ghaṇṭa (spear - bell), padma - sarpa (lotus - snake).

On a later occasion another difficulty occurred. Some of the Jain icons at Deogarh (Jhansi District) which seemed at first sight to represent Jinas, turned out on closer examination to be images of Bharata and Bāhubalin, sons of the first Jina Ṛṣabha. The Jain cakravartin Bharata is recognized by the so-called ratnas, the attributes of the cakravartin, whereas Bāhubalin is recognized by the creepers which cover his legs because he is standing motionless in meditation. If all these peculiarities come out well Bharata and Bāhubalin can easily be distinguished from the Jinas although there is otherwise no difference in the representation (both brothers are shown as ascetics like the Jinas, and this is correct in so far as both renounced the world after their mundane career). But only one or two of the Bharata and Bāhubalin images at Deogarh show the distinctive features mentioned in a clear way (§ 4). In the other cases the specific characters appear on a microscopic scale rendering proper identification impossible for the less attentive observer.

The indistinctiveness of characteristic features was much more tangible when I came to examine the small figures which are carved on theriomorph Varāha images. Up to several hundreds of these miniature figures can be seen on a single Varāha, and although the variety is not very great the identification is by no means easy. The execution is often careless, the specific characters of the different groups (Aṣṭamātṛkās, Daśāvatāras, etc.) are not always clear, and no “spacers“ (such as pillars etc.) are employed to separate the different groups, although this simple method would have facilitated the identification considerably. Identifying such sculptures is like reading a text without the words being separated and without punctuation.

 

II

Every student of Indian (especially medieval Indian) iconography will agree that lack of distinction is one of its most conspicuous features. We have however to ask what exactly constitutes the peculiarity because assimilation in one form or the other is found everywhere in art. Louis Reau remarks in his Iconographie de l'art chrétien Tome I (Paris 1955) on p. 298:

“Just as a painter must be aware that the local colours of an object are modified by the reflections of the adjoining objects, it is necessary to take into account in iconography, much more than has been done up to now, these penetrations, which provide the key to a great number of problems without recourse to the texts being required.“

It seems that three factors distinguish the situation in Indian iconography. If we regard the reduction of distinctions as the result of a transformation ^[3], we can describe these factors in the following terms:

1. Non-decorative elements (e.g. hand-attributes) are subject to transformations which elsewhere are confined to decorative elements (e.g. floral patterns).
2. The transformations are not restricted to formal characters (composition, postures, scaling), but affect also the objects themselves. The transformations have an effect on the iconography, i.e. on those elements which are essential for the identification of the figures.
3. The transformations are not linked up with the migration of the forms in time and space but are produced instantaneously. The process is accelerated.

In other words: not the transformations as such but their intrusion into iconography is typically Indian (see below section III on “iconography“). This follows not only from (a) and (b) but also from (c). The speed of the transformations has the result that similar forms representing different motifs are seen side by side, i.e. on monuments of the same style and often even on one and the same monument. Here a confusion is well possible whereas nobody is likely to confuse similar forms which belong to different regions and periods. A comparison with language is perhaps not quite out of place: confusion of similar words (having different meaning) is possible only if they belong to one and the same language, not if they belong to different languages.

The lack of distinction cannot be demonstrated with the help of a single case of similarity. What matters is not the absolute but the “relative“ degree of similarity between different forms: the general fact that the similarity between forms which stand for different motifs tends to be relatively great compared with the measure of resemblance which remains in the case of forms which stand for the same motif (cf. § 70 b).

It goes without saying that the rendering of a motif may become so unrealistic that nobody will expect exact definitions. Distortion of the natural forms (§ 48), decorative transformations (§ 49), excessive syncretism (§ 65), ignorance of the iconographic traditions (§ 66), and non-epic representation (§ 73) are among the factors which may ultimately obliterate the demarcation - lines of the objects and facts represented. In the present enquiry however only such cases of obliteration have been discussed where the character of the representation is in general realistic; a partial absence of distinctions is here unexpected and may therefore create misunderstandings.

 

III

We propose to deal in this article with three questions: In what forms does the reduction of the distinctions manifest itself? Who is responsible for the reduction (the artists alone or mythologists and intermediate authorities as well)? What are the reasons for the reduction of the distinctions? In the second part of the study we shall try to give a final answer (chapters VIII, IX, XI); here we mention only the main points which underlie the description of the material in the first part.

(a) The reduction of distinctions manifests itself broadly speaking as “amalgamation“ and “stereotype representation“ (see section I). In order to make the difference as clear as possible we use two elementary examples. A hybrid animal (e.g. śarabha, a bastard of a tiger and an elephant) is an example of amalgamation: separate objects are joined together. But if a bull is rendered in such a way that it may be mistaken for some other quadruped (e.g. a buffalo) we speak of stereotype representation; in this case the distinctive features of the animals are neglected and different motifs are moulded upon the same pattern. A third category, multiplication, has been classified as “related phenomenon“ and will be denned later on (§ 51).

(b) The reduction of distinctions goes sometimes to the credit of the artists, sometimes to the credit of the “theologians“ (the authors of the systematic treatises on the subject), and sometimes to the credit of the mythologists themselves. Assimilation of different attributes is an artistic transformation, the association of various conceptions like Varāha, Rāmacandra, etc. with Viṣṇu - Kṛṣṇa is a mythological and theological transformation.

Since the “form“ (art) as well as the “contents“ (mythology and theology) are subject to transformations we cannot reduce all the phenomena to the antagonism between the two sides. That may be possible in the case of the biblical motifs in Christian iconography where the artistic innovations stand out clearly against the unchanging background of the scripture. But in Indian iconography we are often not sure who changed the original motif since art is only the last member in the chain of transforming factors. ^[4]

(c) It seems that the reduction of distinctions has mainly three reasons: “association“, “conventionalism“, and “variation“. These three categories may be called “psychological“ in contradistinction to the “descriptive“ categories in (a). We speak of association where different mythological conceptions (e.g. Hari and Hara) or different objects (e.g. weapons and vegetable forms) are contaminated. Conventionalism means economy of forms and economy of motifs, a tendency which is amply testified by medieval and late medieval icons. Variation we call the tendency to modify again and again a given form, a process which, may result in the reduction, loss or misrepresentation of the specific characters. Thus the braids of hair can assume the shape of a halo (fig. 48).

Our classification of the material is based on the descriptive categories mentioned in (a) and on their subcategories which will be defined in the course of the article.

The term “iconographic“ (see chapter X) is used in contradistinction to decorative and epic compositions. The student of iconography is concerned with iconoplastic art alone, not with decorative patterns (geometrical, floral, animal or human) and not with representations of particular historical or mythological events (e.g. Jātaka scenes). No doubt, iconoplastic art preserved to some extent the “epic“ element, but broadly speaking there is a vast difference between the familiar type of icon and the representation of epic scenes as they are found at Sāñcī or Halebid or Ābu. We define this difference with the help of the terms “epic“ and “non-epic“ (chapter XII) and regard only compositions where the epic element is missing or reduced as the subject of iconography. In its narrower sense, “iconographic“ denotes all those features which help in distinguishing one figure from another, e.g. objects appearing in the hands or on the heads. We call all these distinguishing features and elements “attributes“, using the word in its widest sense.

 

IV

Our plan, which has been described in the previous section, is very abstract. We might as well have studied critically a number of monuments with special reference to the various processes of assimilation. We might also have dealt separately with the individual elements of the representation (objects in the hands, head-gears, etc.), demonstrating assimilation with respect to each of these categories. The second approach is no doubt more systematic than the first, but even this method would have been simple as compared with the programme of our essay: a classification of the material according to the various forms in which the assimilation manifests itself. We would have hesitated to arrange the facts according to such an abstract system if the article were meant as a self-contained monograph. This is however not the case. The article has been written in order to provide such enquiries as are restricted to a single monument or group of monuments with the necessary theoretical foundation. We are fully aware of the inadequacies in our system, but we have avoided a detailed discussion of questions relating to the classification. Our categories are only meant as guides for the student and not as the elements of an ideal system of Indian iconography.

The knowledge of Indian iconography is still imperfect. In Europe the literary background of the monuments had been studied carefully before the relative independence of the form was duly recognized. Here one has to study the “autonomous“ and the “heteronomous“ aspect of the forms simultaneously, running the risk of attributing to the text what is actually an innovation of the artist and of demonstrating the liberty of the artist with the help of features which later on may turn out to be based on ancient literary traditions. We have availed ourselves everywhere of the works of Gopinath Rao, Jitendra Nath Banerjea, and Umakant Premanand Shah, but we feel that in some cases our treatment may have suffered from the difficulty of coordinating in a satisfactory way literary and archaeological evidence.

A technical problem is the lack of published photographs. It would be beneficial to iconographic studies if complete sets of figures from structural temples and caves were reproduced in the archaeological literature. So much more so as the processes of assimilation are most easily observed in the case of a series of figures.

We cannot enter into a discussion on the relations between iconography on the one hand and art and religion on the other. It seems however necessary to emphasize that differences in the degree of iconographic correctness and logic do not affect in any way the intrinsic aesthetic and religious qualities of a representation. The standards of iconography should not be applied to beauty and religious value.

* * *

In the Khajurāho-article we have shown that the artists were to a large extent independent from the texts (see also chapter IX). In the case of the temple under discussion, the artists had split up or alterated existing gods in order to create new deities. Similar cases will be surveyed in §§ 29 (alteration) and 41 (derived figure repeats a single feature of the original figure). It is however unusual that the artists create new figures on such a vast scale. - On a later occasion we shall try to demonstrate that the description of the “iconographic types“ does better justice to the material than the identification of individual figures (see §§ 46 and 54). A fuller treatment of the phenomenon described here as “variation“ (§ 69 c) will also be included in that study.