2. 9. RASA AND INSTRUCTION

9. RASA AND INSTRUCTION

De gustibus non est disputandum-"in matters of taste there can be no disputes," runs the old saw.108 On the face of it, taste would seem to be an affair of the heart, not the head, and rasa as "taste" would hardly be expected to pertain to the domain of reason, preeminently moral, social, or other kind of judgment. Thinkers in classical India directly engaged this question, like their counterparts in the early modern West, though the discussion proceeded differently in the two cases, with equally divergent outcomes. As with so much in the history of Indian aesthetics, the conversation begins with Bhatta Nayaka. Since the time of Bhamaha (c. 650) the view had been dominant that the cultivation of literature produces pleasure but also "instruction"-in this context, always instruction in the four "ends of man," love, wealth, morality, and spiritual liberation-with the two outcomes equally balanced.109 This old view came to be embodied in the very definition of rasa at a relatively early stage. For Pratiharenduraja (c. 900), the "enhancement" of the stable emotion that leads to rasa meant its development, in all its complexity and along with all its requisite contextual elements, precisely as a source of instruction.110 It cannot have been much later that Bhatta Nayaka for the first time contested such didacticism. His challenge is implied in the famous differentiation of genres that is almost certainly his own: if scripture commands us like a master and history counsels us like a friend, literature seduces us like a beloved. But it is directly expressed in one of his few surviving fragments: reading literature is about experiencing rasa, not gaining knowledge of some moral precept, something his follower Dhanamjaya was thinking of when at the beginning of his Ten Dramatic Forms he sarcastically proclaims, "I salute the fool who turns his face from pleasure and thinks the point of literature is mere instruction, no different from historical narrative."111 This too is the position Abhinavagupta defends early in his commentary on Light on Implicature: in literature, "pleasure is the predominant element"; it is "bliss that constitutes the final goal of literature, taking priority over even instruction."112 Yet it is a position he would qualify and eventually abandon. In his turn away from Bhatta Nayaka, Abhinava was taking a cue from his teacher Bhatta Tota (c. 975). Again, we have only a fragment that explains the latter's position, but it seems clear enough: "Pleasure is constitutive of rasa, and rasa is simply drama, and drama simply knowledge." Pleasure may still be held to be an essential component of literary art, but not an end in itself; its true purpose is "knowledge" of the four ends of man. This fragment is cited by Abhinava late in his commentary on the Light itself in a closely reasoned passage that seeks, or so it seems, to manage a tension with his earlier statement. Rasa, he explains,

is made possible by virtue of the "conjunction of aesthetic elements" that are themselves inseparable from instruction in the four ends of man. In composing the elements appropriate to a given rasa, the poet's total "self-surrender to the savoring of rasa" that Anandavardhana stipulates is actually what is instrumental in making such instruction fit and apposite. Hence, literary pleasure as such is instrumental to education, as my teacher argued.

To soften the seeming contradiction with his earlier argument, he adds that pleasure and instruction are not two separate things since they converge in a single object, the propriety (aucitya) of the aesthetic elements: the experience of rasa is preconfigured (in terms of which elements are "appropriate" to the production of each rasa) to align with social norms (in terms of

which responses are the "appropriate" ones to engender in the first place). "The source of literary pleasure as such lies, as I have repeatedly said, in the social propriety of the aesthetic elements, whereas 'instruction' is nothing other than the correct understanding of those elements, as being 'appropriate' each in its own way to the given rasa." Instruction through literature, he concludes, should concern itself with practices that conduce to the success of the protagonist or to the defeat of the antagonist.113 In his later commentary on the Treatise on Drama, a far more insistent theory of rasa as social and moral pedagogy is developed. The whole focus of aesthetic experience is shifted to the emotions that pertain to the ends of man: "The end result of the savoring is instruction in morality and the other ends of man"; the viewer of drama "comes to possess a certain form of consciousness of the sort conveyed by the deontic language of scripture-that those who do such and such a thing receive such and such a reward."114 By a virtuosic if no doubt anachronistic reading, he links Bharata's idea of the four primary rasas, the erotic, the violent, the heroic, and the peaceful (replacing Bharata's macabre), with the four ends of man, love, wealth, morality, and liberation. And in his account of the sixth hindrance to aesthetic consciousness (concentrating on nonessential parts of the artwork), Abhinava explains that the "most essential aesthetic components are those several forms of consciousness that pertain to the ends of man," and again correlates with them the stable emotions, desire, anger, determination, and impassivity.115 Literature's capacity to refine our moral imagination is thus continually reasserted as a central tenet of Abhinava's mature aesthetic theory. It deserves noting, given Kant's influential "differentiation" of aesthetic judgment from social and moral judgment, that the domains of these types of judgments overlapped closely in classical India. Equally notable, however, is the actual nature of such "judgment." The pedagogy enabled by rasa experience is not learning to feel the way another has felt, to see the world through the eyes of another, to develop solidarity with another in his suffering-what the philosopher Richard Rorty, for example, understood to be the moral work of literature.116 This may appear to be implicit in the discourse on rasa: how else, one might suppose, could what Bhatta Nayaka called "commonization" or the "heart's concurrence" achieve their effect without an ethical education that made it possible to experience the experience of another as one's own? But Indian thinkers never quite make this explicit; such an interpretation even seems to misconstrue their argument. Commonization is concerned less with positing a broadly human, and humane, way of understanding narrative-like Rorty's (widely shared) view that "by identification with Mr. Causaubon in Middlemarch ... we may come to notice what we ourselves have been doing," our blindness to the pain of others, for example-than with applying the narrative to one's own life by assimilating its notions of propriety. Taking pleasure in that narrative was instrumental to the creation and confirmation of the judgment of a work's moral order, and about that order there was no dispute, since there was no dispute about the social norms with which it was to be correlated. A narrative has an essence, to which there is a "proper" way to respond.117 The pedagogy of rasa was, thus, not a matter of working through the moral ambiguity of literature (was it just that Dushyanta should reject the pregnant Shakuntala?); traditional readers never highlight such ambiguity, however much poets may have invited them to do so. On the contrary, literature was understood to present not questions but answers, which were easier to learn through literature than through other communicative forms. One could well say, then, that for Indian aesthetics,

there really is no disputing in matters of taste, not because each reader has his own in accordance with the relativist-skeptical stance of modernity, but because all readers have, ideally, the same. Nothing said so far, however, explains how viewers and readers are able to taste rasa in the first place and to grasp its social-moral logic. Is any special knowledge required? What exactly is the role, if any, of aesthetic theory itself in the education of taste? How, in short, does a rasika, a person able to taste rasa, come to be a rasika? Rasa theory would seem to be an account of everyday aesthetic experience, of how viewers and readers react. And after all, what special training is required for getting lost in a book or film? Perhaps more than we know, since although it may seem to be a natural human capacity, Indian thinkers saw "nature" quite otherwise. A rasika may largely be born, not made, but who is born a rasika? Not many thinkers addressed this question directly. For Vishvanatha (c. 1350), only "certain special people" have the capacity for relishing rasa, those who have a "superabundance of sensitivity" and "possess merit acquired in a former existence," or as he puts it elsewhere, the requisite "predispositions." Those lacking such capacities-here Vishvanatha cites from the lost work of Dharmadatta-"are like the walls and wooden posts and stone floor inside the theater." Predispositions are acquired in one's present existence as well as in former ones, and these are what make the savoring of rasa possible. If we did not hypothesize a causal force of predispositions cultivated in a present life, we would expect even dry-as-dust theologians to savor rasa; if we gave no causal force to those acquired in a past life, we could not explain why some who are keen to savor rasa are incapable of doing so.118 There is no doubt a good answer to the obvious question why the endless cycle of transmigration would not eventually endow all people with all predispositions, but our thinkers do not provide it. There is more to aesthetic sensitivity than simply one's predestination for it, to be sure. Abhinavagupta argues that receptivity comes, at least in part, from a previous study of literature. But that too requires the presence of good karma from past lives, a "heart by nature like a spotless mirror," a mind "no longer subject to the anger, confusion, craving, and so on typical of this phenomenal world." Only those traits enable rasa to manifest itself "with absolute clarity."119 In the end, rasa theory is meant to explain the world of aesthetic response, and not-except incidentally, as knowledge that prestructures interpretation-to teach us to cultivate it.

10. THE INTELLECTUAL HISTORY OF RASA DISCOURSE

The foregoing account is one attempt to reconstruct the historical transformations of rasa thought as plausibly as evidence permits. It must be inaccurate in some particulars, given the limits of my knowledge, but it also must be untrue, in several senses of "untrue." I cannot have told the whole story of rasa, and not only because no one can know historical stories in their entirety. First, there is the question of sources. We are painfully aware of texts that have disappeared-in fact, no discipline of classical India has suffered greater losses than aesthetics.120 In addition, some texts have been preserved in only a very few manuscripts, or only in part, or in such a state that for stretches on end they are close to unreadable, or, conversely, in so many manuscripts that the very idea of a producing coherent, let alone critical, edition seems absurd.121 And this is to say nothing of the difficulties that confront the reader trying to make sense of texts that the tradition

itself shied away from commenting on.122 My account can be said to be untrue in a second sense, as diverging from what the agents themselves believed about what I have come to think of as the recoding of the meaning of "manifestation," for example, or, more consequentially, about the extension of aesthetic theory from drama to poetry or the shift from formalism to reception. Disagreement with tradition raises knotty if familiar questions about what makes an interpretation valid and for whom, and about history itself-or rather, about history (the scholar's) against itself (the participants'). In one last, related sense my account cannot quite be true, given that in some cases, we have access to texts that few in the tradition did, texts that scarcely entered into circulation in their primary sphere and hence had minimal historical effects. This is surprisingly the case with two masterpieces, Abhinavagupta's New Dramatic Art and Bhoja's Light on Passion. The contemporary scholar is thereby put in the strange position, the reverse of the first predicament, of knowing more of the intellectual history of this discourse than some of its participants had access to. It may be useful to try more precisely to characterize the foregoing history of rasa from two different directions: from the inside out, so to speak, and from the outside in. From the first perspective we can perceive ideas, presuppositions, or objectives that the actors themselves were aware of, even though we may know little about them now; from the second, we can identify conditions that structured their knowledge that the actors themselves may not have been aware of. Each perspective comprises its own kind of historical effectiveness. One constraint on the discourse of rasa, so far little mentioned but far-reaching and largely acknowledged by the participants, pertains to their philosophical or religious affiliations. Many of our literary theorists also wrote on philosophy or theology123 and thus, as in the Western tradition from Aristotle to Kant to Dewey, a formative if not constitutive relationship held between aesthetics and philosophical worldviews. But disentangling this relationship is no easy task in itself, and it is made harder by the unfortunate thinness in the record just where we need it to be thick: on the value commitments of most of our thinkers. If we knew that Shri Shankuka was in fact a Buddhist, we could better understand his arguments both about the place of inference in the aesthetic process (inference and perception being the only two means of valid knowledge that Buddhists accept) and, somewhat more speculatively, about the sources of his view of "imitation" (which perhaps lay in Yogachara "illusionism").124 That Bhatta Nayaka's allegiance to Mimamsa (far deeper than Anandavardhana's) marked his entire system is crystal clear, but if we understood precisely which brand of Mimamsa this was, Kumarila's, Prabhakara's, or another, we might be able to develop a richer sense of how he thought "actualization" worked, and more particularly, how the "eventful narrative" (arthavāda) embedded in scripture "rouses" the ritual agent to re-create the ritual act-and, homologously, the reader the literary narrative. Bhoja is more direct about his Samkhya inheritance125 (less so about his Shaivism), but it is a laborious task to reconstruct just how far this inflected his aesthetic theory, beyond the obvious role of the theory of the three psychophysical elements in his understanding of the "sense-of-self" that constitutes "passion." The situation does become more perspicuous as the meta-aesthetic discourse shifts from philosophy toward religion, where Abhinavagupta's theory of aesthetic consciousness shares many traits with, though is not necessarily conceptually dependent on, his theory of liberated consciousness; the reverse might be posited of the later Advaita aestheticians Vishvanathadeva and Jagannatha, about whose religious views the information in the second case is sparse, in the

first case entirely absent.126 Rupa and Jiva Gosvamin, by contrast, are known primarily as religious thinkers, and their views help us understand how religion and aesthetics were not just related but fused into a new aesthetic theology-at the same time hinting at what we are missing more generally about the religio-philosophical context of earlier aesthetic thinking. A second dimension of intellectual history from the inside out concerns the status and practice of intellectual history in the tradition itself-how the thinkers themselves sought to grasp the development of their discipline-and the sources they had at their disposal. Abhinavagupta is the first to have brought a chronological sensibility to rasa discourse: the prologue to his own "purified" theory in The New Dramatic Art is clearly meant to represent ideas that succeed one another in time and in value. He was able to reconstruct this order-going back some three centuries (to the time of Dandin, c. 700)-because the actual texts were still available to him. The number and diversity of citations in his own commentary from the early writers on the Treatise on Drama attest to this availability; the same holds, naturally, for his familiarity with the works of his own teachers, Bhatta Tota and Bhatta Induraja.127 By the mid-twelfth century, however, scholars were already departing from a strict chronological approach. In his treatment of the history of rhetoric more generally, Ruyyaka places his discussion of Kuntaka and Bhatta Nayaka before that of Anandavardhana, a temporal displacement one of his commentators explains by noting that although both thinkers came later, they are presented as earlier "since they were following the doctrines of the ancients"; by contrast, Mahima Bhatta is said to have "put forward something of his own invention entirely" and so is treated after Ananda.128 Here chronology is inflected by a kind of axiology (the validity of which we will assess momentarily) that comes to the fore in later discussions of rasa. While many of those discussions adopt some version of Abhinava's account, they are subject to two important limitations. First, almost without exception, no later scholar had access to any of the original texts Abhinava cites, or even to The New Dramatic Art, where many are discussed; they all derive their overview from the exposition in his Eye for Light on Implicature (which admittedly sometimes provides more detail), or, far more frequently, from Mammata's précis of this exposition in his Light on Poetry. Second, and no doubt as a consequence of this documentary deficiency, the very content of the chronology and hence its structure begins to change after the eleventh century. Jagannatha (c. 1650) discusses eleven different interpretations of rasa in an order that is entirely evaluative: it starts with the doctrine he accepts (Abhinava's) and ends with those he almost certainly invented to demonstrate the slow descent into ever greater inadmissibility. (In the same spirit of anachronism, unless it is parody, he makes Bhatta Nayaka speak in the rebarbative style of the "New Logicians" that came to prominence only four centuries after his death).129 The natural conclusion to this development of the discourse-what we might call its pure logicization- is found at its endpoint, in the work of Rajacudamani Dikshita (c. 1650, not excerpted here). His "history" of rasa dispenses entirely with historical sources, becoming an account not of what the positions actually were, but of what conceptually they should have been. It is a rather fine line that divides this practice (and tacit theory) of intellectual history from what we can perceive when we look from the outside in. This perspective offers a view of issues that those who made the history could not or did not-so far as we can tell-perceive themselves. A few examples both minor and major, which, given his prominence, are best provided by the works and practices of Abhinavagupta, suffice to give a sense of the problematic as a whole.

The New Dramatic Art, written in Kashmir around 1000 C.E., seems to have vanished from there almost as soon as the ink was dry. The only scholars in the premodern era who evince direct knowledge of the work all lived in Gujarat during the twelfth century.130 What others knew of Abhinava's aesthetic philosophy comes from his earlier commentary on Anandavardhana, The Eye for Light on Implicature, or from Mammata's précis (Mammata himself shows no evidence of having read The New Dramatic Art either, though the question awaits systematic study). What does this lacuna mean for intellectual history? For one thing, since the views on rasa in the two works are not identical, Abhinava's mature theory was essentially unknown to subsequent scholars. Consider his understanding of the state of consciousness that the aesthetic experience represents. At various high points of his exposition, such as his definition of drama, Abhinava announces his name for this state: anuvyavasāya, secondary or reflexive knowledge of a knowledge, but he goes on to carefully gloss this in order to signal its newly charged meaning (it is "on the order of a direct awareness," "consists of the light of the bliss that is one's own pure consciousness," and so on).131 The term-Abhinava's version of Bhatta Nayaka's "experience" through "actualization"-is used in Abhinava's aesthetic sense in no other text on rasa discourse, because no other text knew that sense or the work in which it was contained.132 What later texts do represent as part of Abhinava's aesthetic theory concerns the "manifestation" doctrine of rasa that he developed in his commentary on Anandavardhana. As we saw, Abhinava, confronted with the potential obsolescence of Ananda's treatment of rasa in the wake of Bhatta Nayaka's revolution, transformed an object-oriented linguistic notion into a subject-oriented psychological one-what is now "manifested" is the stable emotion in the heart of the sensitive reader rather than rasa in the text. The core terminology for this modality, "manifestation" (vyañjana and its various cognates), is virtually absent from The New Dramatic Art. The concept clearly had no further role to play for Abhinava, since his new theory is hermeneutical. For all subsequent thinkers, however, "manifestation" became the watchword of Abhinava aesthetics, something possible only if The New Dramatic Art was unknown to them. The history of the reception of Abhinava's conceptual leap in The Eye is even more revealing than this bibliographical lacuna. No later scholar ever comments explicitly on the transvaluation of Anandavardhana's idea of "manifestation" developed by Abhinava when he moved it out of the old thought world of formalism into the new one of reception.133 The fact that he appropriated Bhatta Nayaka's concept of "experience" when reworking "manifestation" likewise went entirely unrecognized by subsequent thinkers.134 All that said, the scholarly practices in evidence here have a long history. When Abhinava ascribes to Bharata himself aspects of Bhatta Nayaka's theory, as well as the entirety of his own radically "purified" aesthetic theory ("It is simply what the sage himself has said and nothing new at all"), he is in quest of an old warrant for a new idea, a conventional move in classical thought and found elsewhere in aesthetics per se.135 To assess Abhinava's rewriting of Ananda as a commentator's misinterpretation of his base text, however, would be to misinterpret the commentarial function in classical India. Commentary could legitimately encompass not just exegesis of the old but also promulgation of the new, no matter how much at variance the two might seem to a present-day scholar. Beyond the transformations of rasa thought, how variously they appear when seen from the

inside out and the outside in, and what the practice of intellectual history means in the two cases, there are conundrums having to do with the overall historical shape of the discourse. The three most obvious are why rasa theory came into being when it did, why it exploded into prominence when it did, and why it came to an end when it did. Theory is related, however obscurely, to practice, and the history of rasa theory roughly maps against the history of the practice of Sanskrit literature-understanding "literature" in the sense accorded to the category in Sanskrit culture itself. In that sense, Sanskrit literature was an invention of the beginning of the Common Era, and the theory of dramatic composition arose relatively soon thereafter. The slow (and fitful) process of incorporating poetry into that theory started not much later than the true efflorescence of poetry (in non-Buddhist circles) around the fourth century.136 What is striking is how quickly rasa became so central to learned discourse in royal courts from Kashmir to southern India. Bhoja, for example, produced two works that engaged the theory of rasa head-on while ruling from a highly visible, even storied, court at Dhara in central India. Why the ruling elite's interest in aesthetics arose when it did (though earlier thinkers like Dandin and Udbhata were also associated with courts) is no easier to answer than the parallel question of why aesthetics in Europe should have emerged first in the early eighteenth century. One can easily coordinate the interests of the court with the cultivation of courtly norms that the aesthetic imagination was meant to reproduce-indeed, perhaps too easily, for coordination all too quickly becomes reduction. It may be true, as I once put it, that good readers-of the sort Bhoja intended his work to form-make good subjects, and, as Terry Eagleton has it, that the aesthetic lies "at the very root of social relations" as the "source of all human bonding."137 But these are bare theoretical bones, and we need more resources, of the sort this sourcebook seeks to provide, if we are to put flesh on them. Far more complicated than the beginnings of rasa discourse or its consolidation as an important cultural-political form is the question of its ending. Space permitting, I would have concluded the Reader with Rajacudamani's Mirror of Poetry because that work shows not only a marked discursive transformation-one entirely dehistorized, as we saw, where early thinkers have become ideal types entirely disconnected from their actual works (most of which had disappeared as much as a millennium earlier)-but also a marked exhaustion. Rajacudamani reproduces the same set of topics in play from the time of Abhinava and adds nothing from his own time and place, the remarkable world of south Indian culture at the height of Nayaka power.138 Later works advance in not one particular our understanding of either the substance of rasa theory or its history.139 Many works purportedly dealing with rasa are actually anthologies of poetry illustrating the rasas, in imitation of Bhanudatta's River of Rasa (c. 1500), but containing nothing of his analytical concerns.140 Clearly, if somewhat perplexingly, the analysis of literary emotion had ceded place to the creation of literary emotion. The remarkable flowering of a new rasa theory among Bengali Vaishnavas was accompanied by a remarkable flowering of new poetry, but while the production of poetry continued, no further theoretical contributions were made after the seventeenth century. The last work I examined from before the colonial caesura, Acyutaray Modak's Essence of Literary Art (Sāhityasāra, c. 1820), is fully representative of the endpoint of the discourse. His interests are altogether other (mostly rhetorical), and when he turns to rasa in the final chapter of his treatise, he does no more than offer a few verses illustrating the erotic.

The picture does not change even if we widen our lens beyond the sphere of Sanskrit intellectuals. Across the early modern vernacular world, poets were clearly fascinated by rasa; the pen names even Muslim poets adopted, "Raslin," "Raskhan," and the like, attest to this. Yet Sufi masters who wrote in Avadhi, aside from incorporating rasa categories into their romances, had no interest in advancing the theoretical project of classical aesthetics. Hindu intellectuals produced large numbers of studies, but again these were either restatements and anthologies on the model of Bhanudatta's specialized treatise on the typology of leading female characters (nāyikābheda) in his Bouquet of Rasa (Rasamañjarī), or vernacularizations of older classics (like Kulapati Mishra's 1670 Secret of Rasa, Rasrahasya, a version of Mammata's Light on Poetry), which embody nothing of the conceptual ferment that had marked the discourse over the previous millennium.141 Although many works await editing, nothing, published or not, suggests that early modern intellectuals of whatever linguistic orientation or religious persuasion had anything to add to the rasa conversation.142 That the end came on the eve of colonialism is entirely coincidental, and cannot by any means be construed as a consequence. From one angle, rasa discourse in the period 1650-1800 presents a picture of intellectual stasis. At times it was hard to move the conversation forward, since innovation in Sanskrit thought was always threatened with Ockham's Razor. Bhanudatta in 1500 had to defend his invention of the "fantasy" rasa from the charge that it had "no traditional standing," just as centuries earlier Anandavardhana had to defend "implicature."143 The impediment of "scholarly convention"144 did not of course stop either scholar, or Bhoja, who directly attacked it. But Jagannatha in the mid- seventeenth century evinces a newly heightened sense of traditionalism when he refuses to entertain the possibility of any modification of the received aesthetic system-by the addition of the devotional rasa, for example-lest disciplinary chaos ensue.145 All the other questions, of the sort that modern aesthetics learned to ask-about the criteria for identifying something as art, or interpreting it, or evaluating it beyond the traditional system of genre compliance, rhetorical exegesis, and the specification of disqualifying "flaws"-were even further removed from the agenda. Whatever the force of such explanations for the discipline's denouement, another seems considerably less cogent, namely, that rasa theory was simply too inflexible to account for new kinds of poetry that appeared in the early modern period, since there are two unwarranted assumptions in the argument. One is that the new literature was radically incommensurate with the old, but this has yet to be convincingly demonstrated. The other is that rasa theory as such was somehow narrowly tied to that old literature, but this is based on an impoverished understanding of the theory's aim, which, as the materials offered in this Reader show, is to account for the emotional core of literature and why we respond to it the way we do, and which accordingly cannot be tied to any historical moment.146 A related notion, that rasa theory was exclusive to drama and Sanskrit drama declined in the early modern era, is wrong on both counts. 147 Viewed from another angle, and with greater hermeneutical charity, the fact that rasa discourse did come to an end might be taken as marking the attainment of a state of conceptual plenitude. After a millennium and a half of the most searching analysis the world had ever seen-on the basis of a carefully elaborated lexicon, stable categories, and fully shared assumptions of core questions-of the emotional structure of literary artworks, thinkers were perhaps justified in believing that they had carefully weighed every possible alternative and fully understood the

nature of aesthetic response-and that there was nothing left to say.

11. "TASTE" COMPARISON AND THE PORTABILITY OF RASA THEORY

Why should "taste" have become the pre-eminent metaphor for understanding aesthetic response in classical India? Curiously, this is something our authorities never care to argue out on philosophical grounds. They unpack the idea only for its metaphorical implication, and then only with respect to the capacity of the aesthetic object to combine disparate elements into a whole, the way a mixed drink combines its ingredients to produce a single gustatory experience (the asymmetry between the six sensory tastes-sweet, sour, bitter, etc .- and the eight aesthetic tastes holds little interest for them). There may have been the incidental implication that rasa theory imparts "the ability to detect all the ingredients in a composition," as Hume famously defined taste, but this was never directly stated, and the very relationship between knowing rasa theory and improving reading practices went largely unexamined.148 Occasionally the image is extended to the "chewing over" required to get the full sense of a poem, the way sugarcane must be slowly chewed to extract its juice.149 But generally speaking, the metaphor did no further work. In particular, Indian thinkers seemed unconcerned to explain the relationship, obscure on the face of it, between nonrational "taste" and the highly rational social and moral judgments in which the rasa experience of literature is meant to school the reader. Perhaps it was too obvious to them, and only modern Western readers feel the need for such an explanation, living as they do in a world where knowledge has become the preserve of reason alone, with the relationship between taste and moral judgment severed and the "aesthetic concept of morality" lost for good.150 Even more curious than the presence of the metaphor is the fact that a second great tradition of aesthetic analysis, at the start of the modern era, should have independently settled on the same way of expressing the response to art. True enough, for both traditions it would have been obvious that this response occurs initially at the experiential, even physical level, and only subsequently at an intellectual one. "Aisthesis" in the radical Greek sense of the word, as a general term for this object of study-the "feeling" part of art-makes very good sense. But "taste"? Does it really offer a "natural" metaphor for the aesthetic sensibility, and if so, in what sense?151 It may be only their vagueness that makes other metaphorical locutions for aesthetic experience, such as being "touched" or "moved," any less curious. But on the face of it, to say, however figuratively, that taste is the medium of our interaction with art is no less strange than saying it is smell. Unlike smell, taste admits of degrees; as a bodily sense it also has a more direct relationship with the object as well as with the object's pleasure than the "distance" senses such as sight (we like tastes in a more intimate way than we like sights).152 Though natural, it can in principle be improved with training, the sort of training that, in the case of artworks, aesthetics would hypothetically be able to provide (as we have seen, there is uncertainty about this in India). It can also be "acquired." But additionally, taste seems to capture that special phenomenological truth formulated by Dufrenne. Feeling is embedded in the object no less than in the subject, and the viewer experiences feeling because affective quality belongs to the artwork; in the same way, we have the taste of a thing only because the thing itself has taste, as it does not have sight. The long debate over rasa's location can be seen as a search for an understanding already gained by the

metaphor itself-this is just what Abhinavagupta argued153-one not attained in the West until the rise of phenomenological aesthetics. The use of the metaphor in European intellectual history seems to have begun with the Jesuit thinker Balthasar Gracian in the early seventeenth century, for whom taste functioned more as a moral category than an aesthetic one: it is possible to refine the taste as well as the mind; in taste begins the drawing of distinctions and hence social cultivation.154 Taste became central to aesthetics, however, only when aesthetics was first invented as a discipline in the mid-eighteenth century, in a world where the hereditary prerogatives of aristocracy were weakening and judgment itself was becoming the foundation of a new society. Hume's concern, like that of many other eighteenth-century thinkers, was to establish a standard of taste in the face of subjective aesthetic sentiment, which he does by tracing the diversity of taste to a diversity in capacities to register what are, for him, objective qualities of beauty. His emphasis on judgment may set him apart from our Indian theorists; however subjective "taste" may seem to us, there was never any doubt for Indians that a single standard could apply.155 Yet like them Hume holds that the cultivation of art is essentially the cultivation of moral awareness, though the process is very different in the two worlds. For Hume, passion is linked with taste; it is something to be disciplined by taste, which is the true source of happiness; art refines our feelings.156 For Indian thinkers, the relationship of emotion to reason was in general a question of little philosophical interest, but neither was emotion something to be subordinated to or dominated by knowledge, as it was for Plato and most of his successors. And in any case, aesthetic pedagogy unfolds for Indians in a far more explicit manner: the viewer of a play becomes suffused "by the desire to attain the good and to avoid the bad," as Abhinava puts it, and "he actually comes to do the one and to shun the other, given that he has now gained an understanding to this end."157 What is most fundamentally constitutive of the Indian discourse on rasa, namely, the relationship between taste and social propriety ("The one thing that can impair rasa is impropriety," says Anandavardhana. "Composing with customary propriety-that is rasa's deep secret"), is also most occluded, for the sources of social propriety lay far below the level of analysis. Propriety was simply a given.158 This "miscognition" of the social determinants of judgment was largely the case in the West as well, until critique became a component of criticism and taste was identified as a marker of social status. The distinctions social subjects make-between the beautiful and the ugly, or whatever-serve to distinguish themselves ("Taste classifies, and it classifies the classifier").159 "Good taste," "bad taste," "tasteless": these are things for which we can easily provide Sanskrit translations (sarasa, virasa, nīrasa, etc.), but the latter carry, or are permitted to carry, no hint of the social origins that would impugn their naturalness. Whereas the English term "taste" is always applied to the capacity of the social subject, the Sanskrit term is typically tied to the aesthetic object. As for the Weberian account with which we started, not much is left. There was indubitably an autonomous domain of art in India before the coming of European modernity, one fully distinguished from the religious sphere. No Indian Arnold may have ever suggested that poetry could replace religion as a source of salvation, though Bhatta Nayaka came close. But it was only in the sixteenth century that thinkers claimed, or came close to claiming, religion could replace poetry. Before this, literature for Sanskrit thinkers was an affair of this world, and aesthetic theory was a way of making sense of how the world produced rasa, and rasa helped reproduce the world.