1 The Discovery of the Citrasūtra

It is around the first quarter of the twentieth century that some major textual sources, either complete or as fragments, were “discovered,” edited and translated. They began to acquire tremendous cultural significance as carriers of authentic meaning. One such text was the Citrasūtra of the Viṣṇudharmottara, which emerged on the stage of art history in India when it was first translated into English in 1924 by Stella Kramrisch, the pioneering historian of Indian art. This art historian from Vienna chanced upon this text soon after her arrival in India in 1919. The first printed text of the Viṣṇudharmottara Purāṇa, edited by Pandit Madhusudhan and Madhavaprasada Sarma and published by the Venkateshwar Press, Bombay in 1912, is the one that caught attention of Kramrisch. It was her English translation (Kramrisch 1928 [1924]) which brought the text into the discourse of art history and Indology.

It was almost a decade later that another pioneering art historian, A.K. Coomaraswamy, turned his attention to this text, singling out one of its chapters, adhyāya 41, for translation and commentary (Coomaraswamy 1933, 13–21). This chapter of the text deals with the classification of painting into four types—Satya, Vaiṇika, Nāgara and Miśra. According to Coomaraswamy, the first two types corresponded to the pictorial tradition of the Ajanta caves.

By now, the two most eminent art historians of Indian art who played a seminal role in establishing the discipline of art history in India were involved in the interpretation of the Citrasūtra. This, in turn, exalted the status of the text and transformed it into an Ur-text for a wide-ranging extrapolation about Indian art and aesthetics, which continued into the first decade of the 21st century (Nardi 2006).

Almost twenty-five years separate Coomaraswamy’s commentary on the text and its first critical edition, brought out by a Sanskritist, Priyabala Shah in 1958 and 1961. Shah’s edition broke fresh grounds in textual criticism when she incorporated the readings from six new manuscripts. However, her attempt at theorization was restricted, interestingly, to the same chapter selected by Coomaraswamy, which dealt with the classification of paintings. She was far too involved with connecting the types of paintings with types of architectural styles to pay attention to whether Coomaraswamy’s metaphysical readings of the terminology were borne out by the edited text. Exactly two decades later, interest in this text was renewed when another art historian, C. Sivaramamurti published a translation and commentary of all nine chapters of the Citrasūtra (1978). A Sanskritist and an art historian of post-independent India, Sivaramamurti assigned a diametrically opposite significance to the text, one validating “Indian naturalism” as opposed to the manner in which Coomaraswamy used this text as a support for his claims of transcendentalism during the colonial period.

2 The Citrasūtra Turn in Early Indian Art History

The interpretation of the Citrasūtra is further complicated by the fact that it was “discovered” in colonial India; this implied that its interpretation would be caught in a comparative framework involving the art of the colonizer and that of the colonized. The central question around which the art of India and that of the West were compared and judged was that of ‘naturalism,’ a term that was seen as intrinsic to the identity of western art.

When the Citrasūtra came to light in the early decades of the twentieth century, cultural nationalism was at its height and art history as a discipline was being introduced into Indian universities. Naturalism supplied a key term, often as an antithesis to ornamentalism, around which debates on the worth of Indian art and craft were being carried out amidst the ascendency of cultural nationalism. Transcendentalism and naturalism are dialectically connected as part of a discourse largely constructed by A.K. Coomaraswamy to place Indian art on a morally higher plane. The former stands for Indian art and its spiritual interiority, defined in contrast to the naturalism ascribed to Western art. There prevailed a climate of contestation rife with orientalist binaries that pitted the ‘rational,’ ‘scientific’ West with the ‘mystical,’ ‘irrational’ India (Masson-Oursel 1925).² If Indian art is believed to be the result of the artists’ power of meditation, Western art is assumed to capture only the surface of nature or the visible world but never its inner essence. Hence the imbrication of the naturalism debate within nationalism. In the battle between the superiority of naturalism in art as a hallmark of Western control over representation and its rejection by Indian nationalists for the higher civilizational ideal of transcendentalism, the latter won the day. Its victory placed Coomaraswamy at an advantageous position, and today this discourse has come to assume truth value, constituting a common sensibility concerning Indian art (Dehejia 1996).

For transcendentalism to emerge as an effective discourse to counter colonial presuppositions about Indian art, such as the denial of the existence of fine arts in India, an alliance of Art History with Indology was the need of the day. Already by the middle of eighteenth century, the discipline of Indology had been founded, having as its main objective the study of Indian culture through texts. Indology, right from its inaugural moment, operated with the assumption that it was ancient and classical Indian texts that were the most authentic means of recovering India’s past (Halbfass 1981).

Around the early 1920s, art history came to be introduced in an Indian university when the Vienna-trained art historian, Stella Kramrisch, began to teach this discipline in Kalabhavana, Visvabharati University, Santiniketan, at Rabindranath Tagore’s behest. In other universities where it was taught, such as Calcutta University and Banaras Hindu University, it was as an adjunct to ancient Indian history. It was in 1950s that an independent department of Art History and Aesthetics came to be established in the Faculty of Fine Arts, MSU, Baroda. This discipline was mobilized for the reconstruction of India’s past, part of the nationalist project to regain the lost essence of Indian-ness as a means of reclaiming subjectivity (Guha-Thakurta 1991).

3 Coomaraswamy’s Reading of the Citrasūtra: A Founding Moment in Indian Art History

So strong was Coomaraswamy’s commitment to transcendentalism that it impelled him to interpret the traditional śilpaśāstras through the same lens. I have discussed his negotiation with the anomaly posed by the śilpaśāstras elsewhere (Dave-Mukherji 2008, 132–134). Let us see how the term sādṛśya gets inflected by his translation. As a case in point, let us turn to Coomaraswamy’s translation of adhyāya 41 of the Citrasūtra, which primarily defines four categories of painting: Satya, Deśika, Nāgara and Miśra. He particularly focused on the first category, as it seemed to pose a challenge to his assumption that there is no place for naturalism in Indian art. In fact, the very term Satya may be translated as “real” or “naturalistic” in light of its definition—that any painting bearing resemblance to the world belongs to this category:

yat kiñcil lokasādṛśyaṃ citraṃ tat satyam ucyate

I have translated this as follows (Dave-Mukherji 2001, 159):

Whichever painting that bears a similarity with the world [that painting] is called Satya (“Naturalistic”).

This is in fact in wide variance with Coomaraswamy’s translation (1933, 13):

Painting that represents any of the worlds (kiñcilloka-sādṛśya) … is called Pure or Sacred (satya).

How satya comes to translated as “Pure or Sacred” is not a matter of willful mistranslation. Rather, we need to historicize this mode of interpretation and ask, under what conditions may satya be taken as “Pure or Sacred”? Satya could no longer be accepted as “True to life”, as interpreted by Kramrisch (1928, 51). Another term, which intensified this problem, was sādṛśya, which could not be allowed to retain its usual sense (Coomaraswamy 1933, 21):

If we understood sādṛśya then to mean “illusion” or “realism,” verisimilitude of any crude or naive sort, we should be contradicting all that we know of the oriental conception of art.

It was the co-existence of satya with sādṛśya (resemblance/likeness) which compounded Coomarsawamy’s problem of avoiding literal translation. Hence Coomaraswamy was compelled to connect yatkiñcilloka- with sādṛśyam, against the rules of Sanskrit syntax, in order to yield the meaning “any of the worlds” in place of “any painting.” His translation (Coomaraswamy 1933, 13) continues as follows:

Painting that represents any of the worlds (yatkiñcillokasādṛśya), that is elongated and has ideal proportions … [emphasis added]

In fact, the description of the painting as “elongated” (dīrghāṅgam) and having “ideal proportion” (supramāṇam) belongs to the next category of painting, termed Deśika. By misreading in this manner, he could explain away the naturalistic potential in the term sādṛśya and impose his view of idealism being the essential feature of Indian painting (Coomaraswamy 1933, 13):

So also in verse 2 above, where kiñcilloka must include devaloka [“abode of the gods”], sādṛśya cannot be interpreted as “naturalistic.”

Coomaraswamy’s departure from normal Sanskrit syntax and semantics did not go unnoticed, but invited ridicule from his contemporary, a well-respected Sanskrit scholar (Raghavan 1933, 905):³

Satya cannot be interpreted as a Sattvika picture. One can as well derive it from Sat and say it is the picture of the Upanisadic Brahman.

In response to Raghvan’s criticism, Coomaraswamy revised his translation of satya from “Pure or Sacred” to “intellectual and ideal” (Coomaraswamy 1933, 26), and rejected its literal meaning, which was however subsequently reinstated by Priyabala Shah (1961, 120) in the first critical edition of this text:

Whatever (yatkiñcid) painting depicts semblance of the world (Lokasādṛśyam citram) is called Satya.

Coomaraswamy evidently felt in need of more support for his elimination of the concept of resemblance than the Citrasūtra could provide. He began to look to the idealist schools of Upanisadic philosophy and epistemology (Coomaraswamy 1933, 26) to arrive at a metaphysical reworking of the term satya:

Satya is “real”, “actual”, “intelligible”, “ideal”; nāmarūpa, “form” as distinct from the natural disorder (anṛta) of the sensible world (loka) (Bṛhadāraṇyaka Upaniṣad, 1, 6, 3: Taittiriya Upaniṣad, II, 6).

In his recasting of sādṛśya, Coomaraswamy turned to a rather unexpected source for an alternative definition of imitation in art that could preserve the term, but be free of the unwanted association of naturalism. He found what he was looking for in western medieval sources. In my view, this detour to the western medieval sources was not fortuitous but compelled by Coomaraswamy’s confrontation with the Citrasūtra. This is confirmed by evidence of the period for his new interest, as given by his biographer, Roger Lipsey. According to Lipsey (1977, 73), the beginning of the 1930’s marks the period of transformation for Coomaraswamy: “… it is just at this point that the relation to medieval Christian thought changed from an enthusiastic but transient acknowledgement of the value of Meister Eckhart in particular to a permanent and scholarly interest in the entire scholastic tradition from its origins in the St Augustine to its flowering in the thirteenth century.”

For Coomaraswamy, medieval Christian thought best illuminated how to interpret sādṛśya without any ‘naturalistic’ connotations (1933, 26):

Now as to Visnudharmottara, III, 41, 1 and 9, kimcit is not “any” but “somewhat”; loka not here “worlds” but the sensible (not alaukika) aspect of the world; sādṛśya is not “resemblance” but consonantia, adaequatio rei et intellectus, …

Combining references to western medieval sources with the terms from Indian epistemology, Coomaraswamy (1933, 26) arrives at a new definition of the satya type of painting as

… sāhitya, sādhārāṇya, all of which have reference to unity self-contained in art, and not to likeness (sādṛśatā) to a model. Kimcilloka-sādṛśya is then “the unity of which is only somewhat as to the world, …”.

For Coomaraswamy, sādṛśya with its naturalistic implication and its coexistence with satya, translated as “Pure,” was contradictory. Here, purity is taken as a “unity self-contained in art,” and for that reason the potential of referentiality to the world outside implied by the term sādṛśya was viewed as posing a threat to his construct of art as the realm of the mind. In another instance, he qualified this term as “a consent (sādṛśya) of pictorial and formal elements in the substance (śarīra) and essence (ātman) of the work” (Coomaraswamy 1933, 27). While he conceded to some presence of pictorial realism in the other three types of painting, it is the first type or Satya through which the ethical notion of art as “Truth” could be admitted as that which had to be elevated beyond any reference to this realism (ibid.):

… and this contrasts with vaiṇika and nāgara painting in which it is to be understood that the realistic, pictorial (sādṛśa) element is much greater, where accordingly an adaequatio (tadākāratā) rei et intellectus, sādṛśya, is only partially attainable.

Notice a distinction created between sadṛśa (which allows an element of pictorial realism) and sādṛśya (a consent of pictorial and formal elements) was crucial for his interpretation to avoid contradicting “all that we know of the oriental conceptions of art.” Rather than unpacking “the oriental conceptions of art,” Coomaraswamy ended up congealing the old binaries between a rational occident and a spiritual orient. His ‘cultural unconscious’ is betrayed in a footnote in which he claims to have captured the voice of tradition through his alleged fidelity to the textual sources (Coomaraswamy 1933, 27).

What follows is derived from all these sources without the addition of any thought or phrase of my own. Verbal authority could be cited for any statement.

What relevance this kind of interrogation has for us today, this reaching back to the cultural politics of the early-twentieth century, also needs to be addressed. The transcendentalist claim made by Coomaraswamy on the basis of a “found” textual tradition can be explained as a political exigency for combating the colonial representation of Indian art. However, when such reliance on ‘scriptural evidence’ is perpetuated in postcolonial times, its inbuilt essentialism can have serious implications today in an India that is witnessing an unprecedented obsession around nationalism.

Foregrounding the claim that rejection of naturalism by Indian artists was intentional, Coomaraswamy effectively undercut the criticism that Indian art was deficient in naturalism. He suggested, instead, that it was never the intention of the Indian artists to imitate the visible world but to create a symbolic image based on a supramundane ideal which transcended the world of appearance. The charge leveled against Indian art of crude execution and a lack of anatomical accuracy (Ruskin 1905, 347) in the rendering of human and animal form was answered not by questioning the criteria of execution and cultural knowledge of bodies, but by claiming that these so-called “deficiencies” were deliberate, a result of a “specialized technique of vision” (Coomaraswamy 1934, 166):

Technical production is thus bound up with the psychological method known as yoga. In other words, the artist does not resort to models but uses a mental construction and this condition sufficiently explains the cerebral character of the art, …

The vision of the classical Indian artist that emerges from Coomaraswamy’s early writings appears to be modelled upon the idea of sculpture of dhyāni Buddhas, with artistic activity relegated to the domain of pure mental contemplation (Coomaraswamy 1908). The eyes of the artist appear to be visualized as half-closed, focused more on the inward, contemplative space of the mind and less on the world “out there,” the domain of naturalism. With an ingenious reversal of binaries set up by the colonial critics, Coomaraswamy constructed a compelling counter-narrative, but in the process, he reproduced the very colonialist logic he had set out to resist.

Certainly, Coomaraswamy’s transcendentalist claims for Indian art and aesthetics did not remain uncontested, as demonstrated by V. Raghavan’s (1933) criticism of Coomaraswamy’s interpretation and by the translators of the Citrasūtra. However, such voices of dissent were relegated to the margins and rarely made an impact on the disciplinary framework of Indian art history, where Coomaraswamy’s theoretical framework has retained a canonical status. Most curiously, Kapila Vatsyayan, in her very Foreword to my edition of the Citrasūtra has cited the same verse that I have used to critique Coomarawamy’s transcendentalist standpoint in order to justify it. The tenacity of this view can be gauged by the manner in which Coomaraswamy’s misreading of pratyakṣa, direct perception, as the least important pramāṇa or source of knowledge in Indian epistemology, is ratified by Vatsyayan (Dave-Mukherji 2001, xiii–xiv). In fact, all the six classical darśanas of Indian philosophy consider pratyakṣa the most salient source of knowledge (Matilal 1986).

4 The Citrasūtra in Postcolonial Art Historiography: C. Sivaramamurti

The return of interest in the Citrasūtra in the work of C. Sivaramamurti in the late 1970s signals a re-appropriation of this text within his agenda of cultural nationalism. Ostensibly, his turn to this text was to eradicate false readings and restore it closer to its original state (Sivaramamurti 1978, xv). Underscoring his interpretation of the Citrasūtra was the claim that naturalism was accorded a positive role both in Indian art and in the text, which resonated with the stand taken by him more than four decades earlier (Sivaramamurti 1934, 189):

Questions of perspective and foreshortening do not appear to have baffled the old artists of our land as in evident from their talks on such technical details as kṣayavṛddhi and the actual conformance of the pictures to those rules so elaborately discussed in books on theory.

It is largely the presence of terms such as sādṛśya and kṣayavṛddhi (literally “decrease and increase,” i.e. “foreshortening,” a technique used to show apparent depth in naturalistic art) that indicated to Sivaramamurti that the traditional Indian artist was no less proficient in representing the world “naturalistically” than his western counterpart. In a similar move, he looks beyond the context of the śilpaśāstras, examining classical Sanskrit literature, and finds further evidence of imitation in art, as for example, in the Mṛcchakaṭikā (Sivaramamurti 1934, 189):

The tendency of imitation in an artist given in the Mṛchhakațikā is healthy sign of progress at a particular stage provided it is directed in the right direction.

Thus by placing the śilpaśāstras in this wider context, Sivaramamurti retrieved key terms such as sādṛśya, anukṛti and kṣayavṛddhi from the framework of transcendentalism. Sivaramamurti reinstated naturalism in post-colonial India as an index of ‘progress,’ as proof that India could measure up, and more, to the categories and standards of art set up in and by the West. Indeed, he writes naturalism back into the art-historiographical demands of post-colonial India, mapping the term powerfully onto nation building.

For me, the rejection of naturalism in Indian art on grounds of its western identity is as problematic as its nativist retrieval in Sivaramamurti’s interpretation of the Citrasūtra. Interestingly, Sivaramamurti’s espousal of an Indian naturalism met with deep skepticism from Stella Kramrisch, who was invited to write the foreword by the author (Sivaramamurti 1934, 10):

If as the text shows and Dr. Sivaramamurti stresses, realism was a main consideration with the painters, their criteria of verisimilitude were, no doubt, met in practice, although no object painted in the murals of Ajanta, which are roughly contemporary with the Chitrasūtra would strike a spectator today as being realistically painted. The realism is in the eye of the beholder and pious stories told, though not in the Chitrasūtra …

Kramrisch almost poses as a paradox the gap between the ancient painters’ intentions to accomplish realism and their actual practice, and argues that the premodern criteria of representation do not match up with our modern expectations of verisimilitude. Today, it is possible to see an overlap between Kramrisch’s skepticism about Sivaramamurti’s literal reading of the mimetic terms in the Citrasūtra and David Shulman’s (2012) recent problematization of ‘the real’ in the context of his study of the South Indian models of mind.

In his reading of a seventh-century playwright, Harṣadeva’s classical Sanskrit drama, Ratnāvalī, Shulman foregrounds the role of painting in the way the plot unfolds. About the question of verisimilitude, which emerges when the characters recognize each other in their respective portraits, Shulman arrives at a paradox similar to Kramrisch’s, but resolves it by stressing the realm of imagination (Shulman 2012, 28):

Apparently, the painted image is recognized not so much by virtue of how it is painted but mostly as a prop for projection, …

At one stroke, if Kramrisch evacuates realism/naturalism (taken interchangeably in this context) from both the walls of Ajanta and the text of the Citrasūtra, Shulman edges it from the painted surface to the space of literary imagination. Naturalism or ‘the real’ is discursively edged off to the realm of the mind via the “eyes of the beholder” and the physical entity of a painting reduced to a “prop for projection.” Despite her reservations about claims of verisimilitude in the Citrasūtra, Kramrisch had earlier accepted cultural specificity in visual representation. This view implicitly allowed for “Indian naturalism,” even if her example draws from Chinese and European naturalism (Kramrisch 1922, 25):

The Chinese naturalism to the European eye appears as an idealistic abstraction, while the European naturalism strikes the Chinese as ugliness.

However, it is instructive to recognize the difference between Kramrisch and Shulman’s skepticism about the claims of the real. Kramrisch, to some extent, was able to accommodate naturalism as a culturally specific phenomenon that manifested in art theory, if not in practice, while Shulman relegates the question of painterly resemblance solely to the domain of imagination. However, such a categorical dismissal of verisimilitude assumes a wide gap not only between discourse and practice but also between how the Indian mind thinks, how the Indian eyes see and how Indian bodies act; thinking, vision and agency are assumed to fall outside of any rational intentionality.

What needs examination is the persistence of a belief that ancient Indian art willfully rejected any empirical interest in the world that surrounded the artists. In my view, the theoretical framework of what this belief (i.e. in transcendentalism) was once a part of has faded into the background, turning the belief and the problematic produced by that theory into a widely-shared yet questionable sensibility concerning India’s past and art traditions. Rather than a simplistic retrieval of an “Indian naturalism,” as attempted by Sivaramamurti, the very semantic framework of the real operating in the śilpaśāstras and literary texts needs to be unpacked and thought through.

5 The Śāstra-Prayoga Debate

Since the inception of art history in India in the late eighteenth century, the core of the discipline came to be defined by a “detailed study of archeological evidence using a stylistic approach” (Maxwell 1989, 5). Despite the momentous “discovery” of the śilpaśāstras inaugurated by the coming to light of the Mānasāra in 1834, and by subsequent work on the Citralakṣaṇa, Citrasūtra, Mānasollāsa and other sources around the early-twentieth century, these have still not been adequately explored for developing interpretative frameworks of art history in India. As pointed out by Tapati Guha Thakurta (1991, 170) in the context of Bengal, a sharp divide arose between history and aesthetics, as between archeology and the study of śilpa texts, when modern commentaries on ancient texts came to be written by the late-nineteenth and early-twentieth centuries:⁴

Art history and aesthetics, even as they grew out of the same institutionalized scholarly sphere, would henceforth always be marked out as a discretely differentiated field.

The separation of aesthetics and archeology got mapped onto the śāstra (theory) and prayoga (practice) divide, and led to a division of labour in which the question of meaning and interpretation was seen as the realm of specialized Sankrit-based scholarship dissociated from the empirical study of artifacts, with the latter thought to lie in the domain of archeology.

Reflecting on Indian intellectual history, Sheldon Pollock touches upon the imbrication of power in the Indian intellectual tradition around śāstra-prayoga (Pollock 1984, 499–519). Seldom has modern art history in India responded to the issue of the caste hierarchy that has existed between the Brahmin authors of the śilpaśāstras and the low caste śilpins or artisans who practiced art (Misra 1975). Kramrisch (1985, 61), in fact, was one of the first art historians to examine caste in the context of art practice and point out contradictory views about labour and caste often held by the Dharmaśāstras:

Manu says that the hand of a craftsman engaged in his work is always ritually pure. The Gautama Dharmaśāstra postulates that a Brahman may not accept food from an artisan. The law books thus distinguish the craftsman in his social position on one hand, and in his state of grace on the other—when he is engaged in his work, when he creates, and, thereby, gives effect to his being an embodiment of Viśvakarma.

Despite Kramrisch’s attention to caste and its implication for the system of patronage and art practice, the model created by Coomaraswamy of the śilpī or an artist/artisan aligned with transcendentalism remained dominant. Even Kramrisch later began to follow this line of thought (1985, 57):

As artists and magicians, the Indian craftsmen transmuted matter into form, vision into concrete shape, and man beyond his earthy competence … Charged with this power their work became concretely real, subtly effective, and transcendentally existent ….

6 Reading Sādṛśya through Ethnography

Since the time of the modern discovery of the śilpaśāstras and particularly with reference to texts on architecture, ethnography has emerged as a valid method of interpreting difficult terminology. As early as Ram Raz’s attempt to make sense of obsolete terminology around 1830s, recourse to extant practice was sought as a way of checking the extent to which śāstric terms had purchase with craftsmen. The question that I would like to raise here is the usefulness of ethnography in the interpretation of terms that deal with the ‘resemblance’ or ‘truthfulness’ of visual representation. Where are we placed today in terms of our understanding of this terminology, and to what extent is combining ethnography with the study of śilpa texts useful for shedding light on its slippery semantics?

Samuel Parker (2003, 5–34) has been at the forefront of embracing the ethnographic approach in the study of texts mainly on architecture. Can this method be equally illuminating in the study of the śilpaśāstras? Do critical terms that relate to visual representation—anukṛti (imitation or mimesis), sādṛśya (resemblance or verisimilitude), viddha (literally “pierced,” or capturing resemblance), aviddha (that which lacks resemblance), etc.—hold resonance for traditional idol makers? Here perhaps we need to distinguish between technical terms for parts of a building and terms that deal with aesthetics (beauty) and visual representation. Terms that deal with complex semiotics such as sādṛśya, beauty, etc., are too culturally loaded for them to travel well across time. Perhaps in this regard it is instructive to turn to British art historian, Michael Baxandall (1972, 26), who has brought to our notice how some art terminologies common in fifteenth-century Italy, which had made perfect sense to their contemporary public, resist easy comprehension for today’s audience.

Ethnography, as employed by some Indologists, assumes continuity of meaning across time such that by recourse to extant practice by modern sthapati (traditional architects) the meanings of traditional treatises can be unlocked. Terms such as sādṛśya are ensconced in a whole web of meanings and practices of viewing that make them intelligible, and in the absence of the latter, they often give rise to skepticism around śilpaśāstras—as if what they say fails to coincide with what they really mean.

Besides, do terms like sādṛśya or satya, no doubt used in classification of paintings in terms of degrees of correspondence to the visible world as far as the Citrasūtra is concerned, have the same valence as technical terms that classify a building? Or, as demonstrated by the survey of the historiography of textual interpretations, terms dealing with visual representation are seen by cultural nationalists as too invested in civilizational identity and the marking of cultural difference to yield an ‘objective’ meaning. More recently, Isabella Nardi has ventured into this fraught zone and attempted to employ the ethnographic approach to analyze how a present-day traditional idol maker engages with visual representation and beauty. Nardi has interacted with local sculptors in Orissa and Rajasthan to explore the relevance of the śilpaśāstras to their current practice (Nardi 2006, 58–63).

In her conversation with a local sculptor, Ram Prasad Sharma, Nardi was struck by Sharma’s comparison of the sculpture of Lakshmi with a European Renaissance master, Giovanni Bellini’s Young Woman with a Mirror (painted in 1515), only to demonstrate the greater beauty of the former (2006, 60):

Applying his rules and ideals, the sculptor explains that the major defects of this Renaissance figure are that the woman has a huge belly and very big arms.

Likewise, Nardi (ibid.) attributes this sculptor’s failed attempt at copying Michelangelo’s sculptures to “different vocabulary and aesthetic ideals of the Indian sculptor and [the] Renaissance [sculptor].” Such an ethnographic exercise, while shedding light on the gap between text and practice, serves to underline cultural difference and tends more towards truisms about how each cultural practitioner responds to aesthetics through what is locally familiar than to clarify meanings embedded in texts.

7 “Mistranslation” of Sanskrit Terms: A Paradigm for Rethinking Visual Representation in Indian Art

There is no simple, pure unproblematic naturalism in the Indian tradition waiting to be recuperated. The very fact that there is no one-to-one translation possible in Sanskrit, for example, opens up an area of interrogation and also a paradigm for interpretation. Perhaps we have neither the words nor the concepts to capture the domain of “naturalism.”

Kapila Vatsyayan insightfully problematizes the question of naturalism in the context of premodern India thus (Dave-Mukherji 2001, xii):

For a fuller examination, it would be necessary to reopen the question of what constitutes as Nature in the Indian mind-scape, more precisely, the mahābhūtas and prakṛti. This comprehension has to be clearly distinguished from the emergence of ‘naturalism’ as an art movement in Europe in the 18th and 19th centuries.

Of course, the absence of a single term does not signify that the concept does not exist. Rather, the absence itself can be seen as an important clue. Perhaps what we have is a much more complex experience which can be described by using a combination of sādṛśya, satya and anukṛti. In the same way, Sanskrit terms such as kṣayavṛddhi (literally, diminution and expansion) or vartanā need not be directly taken as equivalents of “foreshortening” or “shading.” The lack of a close fit between them and the gaps opening up as we attempt to map one set of Sanskrit terms on another set of English terms may itself signify a rich terrain to explore.

Once it is acknowledged that that no mode of visual representation in any given culture in the West or elsewhere has a direct, unmediated and privileged access to the visible, it makes futile any defensive and anxious search for an “Indian naturalism.” In Art History, Norman Bryson (1983) has engaged with a post-structuralist critique of representation and argued against any unmediated access to the real, a view that helps us to complicate the issue of naturalism even in the context of premodern Indian art.

In more recent times and a context closer to South Asia, the question of ‘the real’ has assumed new significance in the writings of David Shulman, who has added a more specific dimension to this inquiry through his focus on South Indian models of mind. I would conclude by juxtaposing Coomaraswamy’s anti-naturalism with Shulman’s evocation of the ‘more than real.’ At one level, one may discern a certain correspondence between Coomaraswamy’s rejection of naturalism as alien to the Indian sensibility, a construct created under the sway of cultural nationalism, and Shulman’s recognition of a different model of mind for South Asia that allows for an interplay between the real and the imagined. At the same time, it is possible to recognize a distinction between their models. While Coomaraswamy’s view of naturalism arose out of a desire to defend Indian art from colonial denigration, Shulman—despite his focus on South India—complicates any simple binaries of the real and the imaginary; this he does by refusing a separation between an observing subject and observed real, instead posing their mutual permeability. Notwithstanding the fact that Shulman arrives at this model through a close reading of literary texts, it has important implications for notions of visual representation.

Once we overcome the defensive rejection of ‘the real’ and resituate the real at the cusp of the existing and the imaginary, the sāstric terminology can be accessed in new ways. The term vāyugati is a case in point. I would like to express my gratitude to Alexis Sanderson for pointing out the ‘literal’ meaning of vāyugati when I was editing the Citrasūtra as his D.Phil. student in Oxford in the late 1980s. This term appears in the description of hallmarks of a skilled artist, the citravid, in the last adhyāya of the Citrasūtra, and sets up a high standard of verisimilitude for artistic practice: an expert must have an ability to depict parts of the body (or difficult transitional points) normally covered by ornaments like the neck, wrists, feet and ears; a body wounded by an arrow; an old person; he must create differentiation between a sleeping and a dead body, and an illusion of an apparent undulation of a flat surface. It is in this context of technical mastery over representation that vāyugati appears as a further marker of a skilled painter (Dave-Mukherji 2001, 250–251):

taraṅgāgniśikhādhūmaṃ vaijayantyāmbarādikam

vāyugatyā likhed yas tu vijñeyaḥ sa tu citravit

He who is able to paint waves, flames, smoke, flags and garments etc. with the speed of the wind (vāyugatyā) is considered to be an expert.

Previous editors of the Citrasūtra, who include Stella Kramrisch, A.K. Coomaraswamy, C. Sivaramamurti and Priyabala Shah, have overlooked the performative dimension of vāyugati. They have considered its representational meaning and connected it with the objects to be depicted, such as flames, banners and clothes that are windswept (vāyugati). Vāyugati can also connect with the agent of representation, i.e. the citravid or the expert in painting, who must paint flames, banners and clothes with the speed of the wind. If the first sense locates the objects to be drawn, such as flames, banners and clothes, in the lived and observable world, the second meaning implies artistic labour and skill, and above all points to the temporality of execution as an index of mastery over representation. Either way, it helps to recast the image of a traditional painter, not as creating images out of a meditative trance through half open eyes and duplicating formula from the past, but as one who encounters the world with keen eyes, trained hands and a vibrant imagination.