Classical Telugu Poetry

An Anthology

TRANSLATED, EDITED, AND WITH AN INTRODUCTION BY Velcheru Narayana Rao and David Shulman

ACKNOWLEDGMENTS		xi
NOTE ON PRONUNCIATION		xiii
Introduction		1
1.	Nannaya	55
	Entering the Mahābhārata	56
	Udanka and the Snakes	61
2.	Nanněcoda	67
	On Poetry in Telugu	68
	How to Make God Fall in Love	71
3.	Pālkuriki Somanātha	76
	The Brahmin Widow and the Untouchable God	77
4.	Tikkana	82
	The Slaying of Kīcaka	83
5.	Mañcana	102
	The Brahmin Who Kept His Wife in the Basement	102
	Quick Wit	104
	The Obliging Husband	104
6.	Ěrrāpragada	106
	Vena and Prthu	107
7.	Nācana Somanātha	112
	Naraka and Ūrvaśi	113
8.	śrinātha	118
	A Definition of Poetry	119
	Burning the Three Cities	120
	The Birth of Sukumāra	127
9.	Bamměra Potana	133
	Visnu the Dwarf	135
	The Rescue of Gajendra	143
10.	Annamayya	147
	Songs for the Lord of the Hills	148
11.	Allasāni Pěddana	156
	The Brahmin Meets the Courtesan	158
	Sanskrit and Telugu	164
12.	Krsnadevarāya	166
	The King's Dream	168
	Visnu-citta of VilliputtŪr	170
13.	Nandi Timmanna	178
	Satyabhāma Kicks Krsna	179
14.	DhŪrjati	191
	The Story of Natkīra	192
15.	Těnāli Rāmakrsna	201
	On Becoming a Frog	202
16.	NŪtana-kavi SŪranna	216
	Beauty or Wealth?	216
17.	Pingali SŪranna	222
	Beauty Unadorned	223
18.	Appakavi	230
	On Poetry and Grammar	231
	On Good Books	235
19.	Ksetrayya	239
	Courtesan Songs	240
20.	śatakas	245
	DhŪrjati. Kālahastīśvara-śatakamu	246
	Kañcarla Gopanna [Rāmadāsu]. Dāśarathi śatakamu	247
	Kāsula Purusottamakavi. āndhra-nāyaka-śatakamu	248
21.	Cātu Verses	251
22.	śāhāji	256
	Take My Wife	257
23.	Samukhamu Venkatakrsnappa Nāyaka	279
	The Love of Indra and Ahalyā	279
24.	Muddupalani	293
	How to Read a Book	294
	Rādha Instructs Ila, Krsna's New Bride, in the Arts of Love	294
25.	Tyāgarāja	297
	I Can't See You Smile	298
	Take Me for Your Guard	298
	What Did You Give Them?	299
	Reach Him Through Music	299
	Won't You Remove the Screen?	300
BIBLIOGRAPHY		301
INDEX		305

A TELUGU WORLD

This verse by the sixteenth-century poet RāmarājabhŪsana celebrates a vital and continuous literary tradition, fully formed and mature, in the language of Andhra in southern India. The poet, working at a historic moment of intense creativity in Telugu, points to a canon already in place. Each poet is paronomastically identified with a divinity. First there is Vāg-anuśāsanundu, the Maker of Speech—Brahmā, in the classical Hindu pantheon—who has both created and married the goddess Vāc, Language or Speech. Within the Telugu tradition, however, this is also the title given to the first poet Nannaya (eleventh century), who established the contours of poetry and poetic style. The thousand tongues belong to the serpent Kundalîndrundu-ādiśesa, who holds the world on his thousand hoods; ādiśesa is also the underlying

identity of the great Sanskrit grammarian Patañjali, author of the famous commentary on Pānini's foundational sŪtras. After the creation of speech itself, one needs grammar at the root of language. But the same title applies to the second great Telugu poet, Tikkana, who is said to have performed a sacrifice known as kundali (thus explaining his title here). The moon, Soma, is probably Nācana Somanātha, the author of the Telugu [Uttara-]harivamśamu (fourteenth century). Bhāskara, the sun, is Hulakki Bhāskara, who produced a Telugu Rāmāyana (late thirteenth to early fourteenth centuries). And God himself, the Lord of Prosperity, is śrīnātha, the fourteenth-century poet who revolutionized Telugu taste. Together, and also no doubt accompanied by other, unnamed poets, these figures created and maintained— in the eyes of the poet who sang this verse—an entire universe, rich with life and feeling, fashioned in and by language. And it is to this language, imagined as a goddess, that the poet pays tribute.

Our anthology attempts to represent, in a modest way, the world of Telugu poetry as imagined by poets such as RāmarājabhŪsana. We present selections from each of the major poets over a period of some nine centuries, from the eleventh to the nineteenth, on the verge of modern times (although in some sense the classical tradition has continued in Andhra into the twentieth century).

Perhaps something of the integrity of this literary world and the striking originality of its makers will come through the distance of time and language. In the following pages we offer a synoptic overview of the Telugu literary tradition, pausing to consider certain key figures in detail.

BEGINNINGS

Telugu literature begins with Nannaya, but Telugu language is much more ancient, attested in place names from as early as the second century A.D. Prose inscriptions from the middle of the first millennium show a gradual evolution toward the classical language. Verse and the appearance of a literary style are attested in inscriptions from the late ninth century on (or even earlier: the Turimělla inscription of Vikramâditya I, in the seventh century, is sometimes seen as already marked by a "high" style).

Early references to the language call it āndhra-bhāsāa or Těnugu or Tělugu: the etymology of the latter term has been much debated, with some tenaciously arguing for a Sanskrit folk-etymology from trilinga, the land of the three lingas, and others deriving it from caste or tribal names (Tělěgas, Tělāganya). Most probably the name is related to the Dravidian root těn "south"; thus, Telugu would be the southern language, in contrast to Sanskrit or any of the Prakrits. Telugu is classed as Dravidian and is thus a sister language to Tamil, the oldest attested Dravidian language, with a continuous literary tradition going back at least to the first century A.D. The cultural presence of Tamil radiated northward into Andhra from very early times: Nannaya seems aware of a great tradition of Tamil poetry, and the powerful forces of Tamil religion, with its concomitant institutional features, unquestionably played a major role in the history of Telugu culture. It is also important to acknowledge that Telugu crystallized as a distinct literary tradition after the full maturation of Sanskrit erudition, including the domains of poetic theory, grammar, social ideology, scholastic philosophy, and so on. Unlike Tamil, which absorbed Sanskrit texts and themes in a slow process of osmosis and adaptation over more than a thousand years, Telugu must have swallowed Sanskrit whole, as it were, even before Nannaya. The enlivening presence of Sanskrit is everywhere evident in Andhra civilization, as it is in the Telugu language: every Sanskrit word is potentially a Telugu word as well, and literary texts in Telugu may be lexically Sanskrit or Sanskritized to an enormous degree, perhaps sixty percent or more. Telugu speech is also rich in Sanskrit loans, although the semantics of Sanskrit in Telugu are entirely distinctive. We will return to this theme.

Already, however, we begin to sense the richly composite nature of the Telugu world. One might think of Andhra as one of the great internal frontier zones of South Indian civilization and at the same time, as such, a melting-pot—a domain of intense interaction among rival cultural currents, with their associated social and historical formations. It is not simple to isolate the various currents or to date their appearance in Andhra history, and one must bear in mind that much of the prehistory—before Nannaya—is hardly known. Still, there are some things that can be said in a general and perhaps slightly abstract manner.

The frontier is structured, in part, along geographic lines. Andhra is divided in three: (1) the coastal zone (Andhra proper), largely deltaic, especially to the north, where the Godāvarī and Krsnā Rivers flow into the Bay of Bengal (as elsewhere in South India, the delta is associated with heavy Brahmin settlement and influence); (2) Tělangāna, the dry Deccan plateau, home to peasants, artisans, and warriors; and (3) Rāyalasīma ("the royal domain"), the southern reaches of this plateau, tapering off into the mixed ecological regions of northern Tamil Nadu.

In early medieval times, Rāyalasīma was apparently far more fertile than it is today. But even the fertile delta to the north was largely wilderness beyond the immediate proximity of the great rivers; this situation changed dramatically only in the nineteenth century, when the major anicuts were created, thus opening up vast areas for rice cultivation. In medieval times the wild drylands of the interior, peopled by shepherds, artisans, warriors, and a growing proportion of agriculturists, were bound up linguistically, culturally, and sometimes politically with the hardly less untamed but wetter regions of coast and delta.

Andhra history and culture reflect the constant interplay of these ecologically distinct zones, especially of the delta and the Deccan, with cultural innovation often emerging in the latter to be reshaped and domesticated in the former. Over time, ever more serious attempts at integration were in evidence as states based in one region spilled over into, or attempted to absorb, political units rooted in the other areas. Early Andhra history, just this side of prehistory, reveals a Deccan-based kingdom, that of the Sātavāhanas, represented mostly by inscriptions in Prakrit, with only tenuous linkages to the coast. The early state structures in coastal Andhra (especially to the north, in the region known as Vengi) culminated in the rule of the Eastern Chālukyas, who eventually married into the Chola system in the Tamil south. Under the Chālukya king Rājarājanarendra, Telugu literature as we know it began, with the poet Nannaya. By the thirteenth century, the center of Telugu state-building

had shifted to the Deccan plateau under the Kākatīyas, who brought massive tank irrigation to the dry zone and instituted creative forms of military organization based on personal loyalty to the king or queen. Key patterns of Telugu culture were established during this period and later adopted and creatively reworked by the successor-states, including the Vijayanagara super-state based in Hampi, to the west of historic Andhra.

To what extent do these relatively distinct regional-ecological systems combine in awareness to form a single cultural entity—Andhra, as we think of it today? How old is such an awareness? The great poet Tikkana, in the thirteenth century, is apparently the first to refer to an imagined community named Andhra (andhrâvali),

but the boundaries of this community are unknown. Originally, the term seems to be a purely dynastic family title. The earliest fully formed reference to a geographical entity known as Andhra within the Telugu tradition may well be śrīnātha's in the late fourteenth century: here the temple of Daksârānāma in Konasīma is said to be the center (karnikā) of a lotus that is itself identified as the middle part of the Andhra country (āndhra-bhŪ-bhuvana-madhyamu). This suggests that Andhra extends far beyond the delta, conceived (perhaps metaphorically) as the center of this cultural and geographical universe; deltaic Andhra, for śrīnātha, is the symbolic heart of the culture. There are, however, other mandala-like schemes superimposed on the geographic realities of medieval Andhra. For example, the important temple to śiva-Mallikârjunasvāmi at śrīśailam on the Andhra-Karnataka border to the west is said to have four encompassing gateways: Tripurântakam to the east, Siddhavatam to the south, AlampŪr to the west, and Umā-māheśvaram in PālamŪru (near Accampeta) to the north. In this mapping the center has shifted dramatically to the west, to the point of intersection between Tělangāna and Rāyalasīma. This tendency to reorient and to situate a new center contextually is perfectly characteristic of the medieval Andhra understanding of place. Like so many parts of India, historic Andhra has no clear boundaries. In the early sixteenth century, the conquering emperor Krsnadevarāya came from Vijayanagara to śrīkākam, in Krsna District, where the god is known as āndhramahāvisnu or Těnugu-rāya—perhaps demarcatng yet another center. This same king also went on pilgrimage to Simhâcalam, at the northern edge of Telugu speech, and to Tirupati, at its southern limit, as if consciously tracing the contours of his kingdom.

The frontier inheres in Andhra culture in several powerful ways. If we look first to the northern delta, we strain to see traces of a largely invisible Buddhist proto-Mahāyāna culture flourishing in what is called Konasīma, "the corner" between the two great rivers. We know something of this Buddhist culture from archaeological findings at Nāgârjunakônda and Amarāvati, and from the surviving works of the famous philosopher Nāgârjuna, who may have spoken a language that was a precursor to classical Telugu. Five major temple sites in Andhra—Daksârāma, Bhīmârāma, Somârāma, Ksīrārāma, and Amarârāma—were in all likelihood originally Buddhist shrines, as the name ārāma suggests. Today all five are entirely Hindu, though Buddhist statuary is scattered throughout the temple courtyards. This process of Brahminizing an early Buddhist substratum, so evident in the five shrines, must have been general and formative. It was successful in the sense that Buddhism disappeared entirely from Andhra. And yet the Buddhist presence seems to have left behind an active and creative level of esotericism in praxis and concept, including Yogic, Tantric, alchemical, and "magical" trends that became a diagnostic feature of medieval Telugu culture.

One sees hints of this fascination with esoteric strains of thought in central works of Telugu poetry such as Pěddana's Manucaritramu—the height of the classical tradition—as well as in a range of other textual traditions, such as Gaurana's fifteenth-century summation of the Nātha mythology, Nava-nātha-caritra, one of the earliest and richest accounts of the magically oriented Nāths in any Indian language. And while we find esoteric praxis and ideology in many forms throughout medieval South India, the organic and generative impact of these strands on Telugu religion and literature were perhaps deeper than in any other major south Indian tradition, with the possible exception of Kerala. There was also, almost certainly, an archaic Jaina impact on Telugu culture, of which little is now known; the oldest extant work on metrics, Kavi-janâśrayamu, is by a Jaina author, Malliya Recana.

Look now to the harsh Deccan hinterland, a true frontier in many senses. A long process of settlement privileged the resilient warrior, perhaps epitomized by the Deccani god Vīrabhadra—śiva as hero. We find him at Lepâksi, in Rāyalasīma, at the southern edge of today's Andhra—a black, furious deity.

The cultic history of the Deccan must include the expansion of Vīraśaivism, originally a militant movement of antinomian worshipers of śiva drawn mostly from the so-called "left-hand" castes, that is, those not tied to the land (artisans, merchants, migratory groups, and so on). At śrīśailam, in the midst of the wilderness, one can observe stages of a long process—still encapsulated in the temple ritual—that seems to have taken this shrine through Buddhist, Vīraśaiva, more normative śaiva, and finally Brahminized/Sanskritized phases. The exotic "heroic" mode is, in any case, still apparent throughout this region, and we may look here for the first signs of that characteristic individualism—a surprisingly powerful and self-conscious presentation of self as subject—that turns up with consistency in Telugu poetry from at least the time of śrīnātha onward. We would go so far as to posit this interest in the uniquely individual subject, initially present in unsystematic occurrences in the literature but later exfoliated luxuriantly in Nāyaka-period texts, as a diagnostic feature of the Telugu tradition over many centuries.

To these two prominent thematic drives, each in its own way born of the frontier, that cut through varying strata, periods, and milieux, we may add a third, from the still more deeply internalized boundary zone of language. As the verse quoted at the beginning of this essay suggests, Telugu poets have consistently been drawn to an examination of language in its life-creating, world-generating aspect. Perhaps something of this fascination derives from the experience of living within a linguistic reality that is itself unusually lyrical and fluid, a constant exposure to language itself as musical sound. It is

probably not by chance that Telugu became the predominant vehicle of south Indian classical music. This association of Telugu speakers with music is an old one, clearly attested in Tamil in Cayankôntār's Kalinkattup-paran in the early twelfth century. Certainly, the Telugu tradition has pushed the exploration of problems of language (speech, grammar, meter, words) in relation to story, perception, and creativity to a point of unusually powerful feeling and insight.

FIRST POET: NANNAYA

Great literatures classicize their own texts, selecting certain major works or authors over others; they also tend to produce retrospective narratives to make sense of this selection. The result, in the case of Telugu, is a simple developmental scheme that can be found, in one form or another, in all modern histories of this literature, in Telugu or other languages. In this framing of the tradition, all begins with Nannaya, the First Poet (and First Grammarian, since an ordered, premeditated grammar must, in this perspective, precede both normal linguistic reality and the creation of poetry). Earlier poetic works may be presumed to have existed, but they are lost. Nannaya is said to have initiated the age of purāna-like compositions with his adaptation of the first two and a half books of the Mahābhārata epic into Telugu.

After some four centuries, this vogue in purānic poetry gave way to full-fledged kāvya or prabandha texts—elevated and sustained courtly compositions. The transition to kāvya of this type is usually said to have reached its apogee in the Golden Age of Telugu literature at the court of Ksnadevaraya of Vijayanagara (1509–1529). Following the breakdown of the Vijayanagara state-system in 1565, literature is seen as slowly sinking; with the displacement southward of Telugu political power into the Tamil country under the Nāyaka kings (sixteenth to eighteenth centuries), new forms of poetic production, some of them supposedly "decadent," became prominent in the afterglow of the classical efflorescence. Modern poetry then represents a blinding flash of revolutionary brilliance against the smoldering backdrop of the Nāyaka and post-Nāyaka decline.

Such is the standard format, a still regnant mythology of poetic evolution, useful, perhaps, for rudimentary classification of the poets. It bears almost no relation to the deeper currents of this amazingly rich and intricate

tradition. It seems likely that this schematic vision is itself derived from a seventeenth-century retrospective ordering of previous works in a manner that first produced the idealized image of a Golden Age centered around Ksnadevaraya with his eight great poets, the asta-dig-gajas, homologized to the eight elephants who hold up the cardinal points of space. (In this sense, literary history and traditional history have marched in tandem; seventeenth-century texts first seem to have imagined Ksnadevarāya in the mode of synoptic "great king.") Indeed, one could argue that it was this later moment of integration, self-reflection, synthesis in grammar and linguistic metaphysics, and retrospective narrativization, in the mid-seventeenth-century Deccan, that marks the true peak of originality in the mature medieval tradition, if such a temporal definition has any meaning.

We can attempt to substitute for the standard evolutionary scheme a more subtle template that will take account of the profound shifts in style and expressivity as well as changes in major cultural themes and premises. Certain key, perhaps emblematic, figures help us to orient this picture of the tradition: Nannaya, Tikkana, śrīnātha, Pěddana, and Ksnadevarāya, in the early stages. Each of these poets, by virtue of creative innovation, changed the rules of play and transformed the classical tradition. Here again we must begin with Nannaya, not as grammarian

but as the poet who first produced a Telugu style commensurate with a complex, and entirely Telugu, sensibility. Clearly, he knew that he was doing this—knew that he was innovative in creating a musical and flowing poetic form, dense with expressive possibilities and unique to his mother tongue. Listen to the way he describes himself (in the third person):

Let us restate this achievement in somewhat different terms: what Nannaya invented was a style of poetic narrative in which the story line is clear, pleasing, and uninterrupted, but that at the same time allows the hearer/reader

to reflect on it and to appreciate the subtleties of meaning. Moreover, the texture—which includes such components as lexical choices, the play of meter, and, above all, the way Sanskrit and Telugu are combined—is harmonious, economical, and musical. Nannaya himself suggests the following two hallmarks of his poetry: prasanna-kathā-kalitârtha-yukti, "lively narrative … with its inner purpose"—a feature perceptible only to "poets with penetrating minds"—and aksara-ramyata, "the harmony of sounds," the phonoaesthetic interplay of syllables. All of this requires a particular and characteristic blending of Sanskrit and Dravidian words and a creative use of Sanskrit compounds, in a manner unknown in "pure" Sanskrit but, after Nannaya, paradigmatic for Telugu. The long Sanskrit compounds that appear throughout Nannaya's poetry, in meters often adapted, again creatively, from Sanskrit into Telugu, are organized semantically rather than metrically. They tend to be longer than is common in earlier Sanskrit poems, and they often spill over line endings, since Telugu meter, unlike Sanskrit, allows complex enjambment. Put differently, the Telugu patterns established by Nannaya's work are not limited by meter: one reads a Telugu verse by breaking at syntactic-semantic pauses. As a result, the stanza allows for more complex syntactic structures and tremendous variation in cadence. The metrical skeleton hardly ever shows through the poem. What one hears, or notices, is the play of muscle and flesh that constitutes texture. By contrast, a verse that mechanically reveals its metrical organization, its caesura breaks and line endings, is considered either as a failure or as belonging to another level of the tradition, perhaps purely oral. It is this kind of sophisticated texturing, with its complex flow of subtle words and sentences, that Nannaya pioneered, and it is this that helps to explain the miracle of transmutation so characteristic of Telugu literature from that time onward, whereby whole pieces of Sanskrit phraseology can be lifted from a Sanskrit source and reworked into a borrowed Sanskrit meter, and yet be entirely and amazingly Telugu.

This same process applies to the transformation of genre. Nannaya's Mahābhārata both is and is not a purāna. It follows the inherited story line, usually with remarkable fidelity to the prototype. But it also allows, indeed demands, reflection upon this narrative and an aesthetic savoring of the texture of its telling on the part of the reader, a process mostly unknown to Sanskrit purānas.

Something quite new happened, and it became the starting point of a process that continued for a thousand years of Telugu literary production. Technically, too, there is the pattern of interspersing verse, in varying meters, and rhythmic prose (the campŪ style that became normative). At the same time, there is a unique quality that is wholly Nannaya's and could never even be imitated by his successors: a gentleness in tone and a freshness in depiction of characters who are domesticated, but only to a certain point. His Sanskrit kings remain dignified and slightly remote, though they are also brought closer to the familiar range of experience of an Andhra listener. The vehemence and wildness of the Sanskrit Mahābhārata are softened and partly tamed, even as the inner world of the characters becomes more familiar. In this sense, as in the stylistic domain discussed earlier, the existence of the Sanskrit prototype becomes a relatively abstract presence that hardly impinges upon the dynamic world of the Telugu text. Only the modern misapplication of the notion of "translation" to Telugu literary creation could see Nannaya—and a host of other Telugu poets—as primarily "translators."

Nannaya's adapation of the campŪ style also implies a particularly active, participatory role for the listener. The itihāsa epic frame normally requires the presence of a speaker and a listener; for example, Sañjaya speaks to Dhtarāstra within the story, describing the battle to his blind master, but his words are reported by the SŪta-narrator to the "original" listener, śaunaka, and other sages. The SŪta, however, is merely repeating what Vaiśampāyana recited, on the basis of his teacher Vyāsa's composition, to King Janamejaya at the time of the latter's sacrifice of snakes. These concentric frames are reframed by Nannaya, who sings the same story to his patron, Rājarājanarendra. And we, listening to a paurānika reciter, find ourselves in precisely the same dialogic situation. The innovation lies in the assimilation of this format to what is, in effect, a kāvya: an aesthetic, self-conscious literary work. Sanskrit literary kāvya, for whatever reason, does not share this need to internalize the listener. Part of the great power of Nannaya's campŪ lies precisely in this activation and co-option of the listener—a characteristic feature of the oral storytelling mode—within a reinvented literary genre.

In general, Nannaya's manner of narration skillfully combines an economy of words with a perfect choice of phrases that embody the emotional progression in events.

The story often unfolds with great rapidity that unexpectedly allows room for reflection on the depth of feeling: this is the "lively narrative with inner purpose" of which the poet himself speaks. Sometimes a single verse encompasses a carefully articulated transition in state or a progression in emotion. For example, King Yayāti, riding through the forest, hears a young woman—Devayāni—calling for help from the dry well into which she has been pushed by her rival. The king dutifully extends his hand to help her out:

First, there is the hand itself—strong enough to bear the earth with its surrounding oceans, all part of a single strong compound. On the other end, another hand, raised, ready to be grasped, wet with the delicate drops of her perspiration that make it even more beautiful, kamra. Everything lies in the readiness that reflects an intention: Devayāni wants to marry this king. But Yayāti as yet knows nothing of this, and feels nothing; he pulls her out, divicě, with a neutral, simple verb, utterly without feeling. Why does he do this? Because he is vísruta-kīrti, a man of good name; he is doing his duty—all part of a day's work. And then, suddenly, unexpectedly, in the very last word of the verse, there is feeling: prītiton, "with love." Before he realizes it himself, he is lost, taken with her beauty, and not only the beauty of her outstretched arm, which he has held and pulled, but also that of her whole body, since Devayāni was pushed naked into the well. We are not, however, told this explicitly; it is implicit in the earlier part of the story, which the listener certainly knows. A lesser narrator might have elaborated the point, but Nannaya is content to suggest it, or to remind his audience of it, with a single word that closes the verse by revealing the shift in the king's perception. It is one thing to show an object, another to reveal this object through the feelings of a participant or onlooker within the story.

There is yet another aspect to Nannaya's originality, at the very limit of linguistic expression. Perhaps more than any later Telugu poet, with the possible exception of śrīnātha in his Bhīmeśvara-purānamu, Nannaya produces a "magical" or "mantric" effect. At certain points—for example, in the hymn to the snakes in the Udanka section translated below—he exceeds the bounds of poetry, or of reference.

One long Sanskrit compound gives us the whole massiveness and heaviness of earth, indicated both by the long string of elements (forests, oceans, mountains, rivers, and lakes) and by the repeated ha sounds—also built into the rhyme scheme in the second syllable of each line—as if to demonstrate the breathlessness of the great snake who bears this burden on his thousand heads. But this dense alliteration has only begun: it is resumed by a dangling, unusual adverb: ajasra, "always," another Sanskrit loan that would normally require a Telugu case-ending but which here simply flows into the line, rhyming with the following word, sahasra ("thousand"). The dangling adverb, in the rush of alliterating sound, suggests the uninterrupted process of bearing the earth's burden. Now, at last, there is a small piece of Dravidian, the nonfinite verb dālci, "bearing." The work is thus still incomplete; another burden must still be borne. The snake ādiśesa, along with bearing the earth, is also the bed on which the god Visnu sleeps in the ocean of milk, and the poet makes sure that we feel this additional, indeed infinite, weight of the god by another gush of sibilants and aspirates, spilling over the line-break: du-Sa Hatara-mŪrtikin jala DHi-śāyiki pāyaka śayyayaina ay-ya Hi-pati.… These two burdens, incidentally, are never seen together in iconography or joined in story; Nannaya has fused them, doubling the snake's dreadful task and arousing our admiration for him. The listener, by now bent double himself under this weight, miraculously made present through the language, needs to rest. For the god, at least, a soft bed is available: the repeated cushioning of the soft double semivowels, - yy(a)-, a delicately iconic reproduction of the texture of the snake's body. And this entire description is part of an appeal to the snakes on the part of the young Udanka who, as is customary, preludes his request with flattery or praise. This verbal production of overpowering sounds has the effect of making palpable and present the snake's own experience; what is more, the verse also controls the reality it has created, like a snake-charming mantra. Indeed, Nannaya's verses in this passage are believed to serve this very purpose of providing protection from snakes.

Here, as one sometimes finds in Nannaya, it is the sound that matters most, more even than any translatable meaning. The sounds, even beneath the words, create a world of their own. Perhaps all language oscillates between the poles of denotative reference and existential creativity; Nannaya

is often closer to the latter pole. He tells us, not boastfully, that he always speaks truth (nitya-satya-vacanun, 1.9 below), suggesting a self-discipline that fashioned a purity of tongue. This "true" relation to language enables him, in effect, to transcend language.

TIKKANA: BOUNDARIES OF THE ORAL

Nannaya belongs to the Delta, and the north, and to a transient moment in the history of the Eastern Chālukya state. He was also, like the great majority of Telugu poets throughout the centuries, a Brahmin. Literary historians have tended to class him as a court poet and his poetry as an elitist, courtly production. These terms may, however, be inadequate descriptions of the poet's reality. One hesitates to ascribe or to attempt to explain anything of Nannaya's power and stature by reference to the weak, short-lived political system within which he found a place. As was so often the case in Telugu literary history, this poet dwarfed by far his supposed patron. Indeed, in a deeper sense, the poet may be said to have created his patron and to have invested him with fame. Moreover, it is the poet who supplied the ideological or conceptual frame within which the state functioned. This is, in fact, one of the historic roles of major Telugu poets.

Nannaya's immediate successor, Tikkana, came from much farther south, in NěllŪru, and from a smaller political system, that of Manumasiddhi, loosely connected to the powerful Kākatīya state (thirteenth century). Tikkana was minister to this king, and later traditions credited him with active roles in politics and war.

The localization of Mahābhārata themes and contents in an Andhra cultural frame achieved a new impetus in Tikkana's poetry. And if Nannaya invented Telugu poetry, as the tradition correctly insists, it is Tikkana who imagined this tradition into self-conscious existence, making Nannaya, retrospectively, the first poet. "First" implies that others follow, and Tikkana is first among these, even in his own eyes. But the Telugu world that he imagined also has expanded to include new domains: dharmaśāstra texts, foundational texts of grammar, and the story (kathā) tradition, seen as kāvya, for example. Tikkana's disciple Ketana was responsible for early works in these genres. Under his master's sponsorship, Ketana produced the Vijñāneśvarīyamu, a legal commentary, as well as the Telugu Daśa-kumāra-caritramu, after Dandin, and the āndhra-bhāsā-bhŪsanamu, probably the earliest Telugu grammar. Law, grammar, poetry, and "history" (itihāsa)—all these exist in Tikkana's semantic universe. It was, in a way, a distinct universe, entirely conceived and fleshed out in Telugu by this poet, aptly known as kavi-brahma, the Poet-Creator.

In stylistic terms, Tikkana's own large corpus—fifteen parvans of the epic —stands alone; no later poet was able to follow his example. Syntactically, his poetry embodies idiomatic, almost colloquial, Dravidian patterns. At the same time, long Sanskrit compounds serve him whenever there is an architectural need for elevated emphasis:

Translation fails to reproduce the effect of what is a single compound entirely in somewhat arcane Sanskrit—a kind of drawn-out explosion of language intensified by the resort to a borrowed register, where the Sanskrit phonemes, with their conjunct consonants, aspirates, and harsh plosives, carry the expressive urgency. Draupadi is threatening Kīcaka, and the threat comes across precisely because of this borrowed but internalized phonotactic feature. Dravidian clusters tend to be softer, voiced, and much shorter; Sanskrit allows a buildup of dense linguistic pressure. But the real expressivity of a verse like this depends on the combination of these two levels: the relentlessly intensifying, semantically compacted Sanskrit compound striving breathlessly toward a pause, and the framing and containing Dravidian syntax, which tends to break into discrete, short units:

The verse that began with a single long compound that perfectly represents the fearsome internal qualities of the heroes devolves in its second half into a string of singular, largely indexical, staccato movements: "the five of them—right now—you—easily—getting angry—will get you—your honor—your life—no doubt—somehow or other." There is no place inside this verse for Kīcaka to hide: the gandharva husbands will hunt him to the ends of the earth, as if they had marked him with their pointing finger,

represented by the finger of the speaker, Draupadi, almost visibly waving at Kīcaka as she utters this warning. Restated in terms of breathing—a central factor in any recitation of the verse, no less so than in the case of, let us say, a flutist—the first two lines require a single long breath, held almost beyond endurance, setting up a tension within the mind of the listener. After the compound ends and the reciter can breathe again, the poet forces him into a series of hammerlike short bursts of language, none of them allowing him to stop for a full breath until the long vowel at the end, in the vocative kīcakaaa—the moment of exhaustion. It is as if Draupadi had started out, fully in control, with a violent phonetic attack upon her enemy that finally renders him breathless, knocked out by the punctuating series of indexical blows.

Effects such as these are standard in Tikkana. The listener maintains a necessary syntactic presence, within the verse, more than an imagined narrative presence. In this respect Tikkana remains close to an oral, conversational mode, as if speaking to a listener in front of him. His text speaks rather than sings—he is thus less lyrical and more expressive than Nannaya —and the audible dimension is never lost. These are not verses written on palm leaves, printed on pages, read silently in libraries. They exist in the living space between the poet and his listener, who hears and feels every syllable in the body.

By the same token, Tikkana begins a narrative style in which the speaker within the text is entirely identified with the character. It is Draupadi who speaks the above verse, not the poet. She speaks in accordance with welldefined, individualized, subjective patterns that are wholly hers, and no one else's. This is not to say that the poet is absent; his style still overrides, or colors, the speech of his characters. However, within the frame of Tikkana's poetic language, one can hear distinct voices. This individualized "voicing" marks a quantum leap in Telugu poetry (sometimes described, rather lamely, by modern critics as Tikkana's natakīyata, "dramatic quality"). Strikingly, however, this feature seems to stop with Tikkana. For the same reason, Tikkana's verse does not flow with the same ease as Nannaya's, or as purely oral poetry, which must have been dominant in his time. It is not easy to read a Tikkana verse; the syntactical breaks tend to be unpredictable—as in living speech—retarding the movement of the meter, stopping and starting afresh, laden with indexicality. This is poetry at the very edge of everyday language, with all its freshness, individuality, and continual surprise.

Listen, for example, to Sudesna, Draupadi's mistress, as she urges Draupadi, against her will, to go to Kīcaka's house to fetch some liquor:

Sudesna starts, breaks, starts again, breaks off, resumes; literally, the verse looks something like this: "I wanted it. Badly. My favorite drink. Didn't want to send a lowly servant. Asked you. You're making a fuss. As if it's something improper. You're upset. Really. Is this good friendship, my friend?" One long sentence, built around a series of nonfinites, infinitives, and conditionals that break it into a choppy sequence of part-utterances. The sequence culminates in a final appeal, a rhetorical question, which nicely masks the speaker's true intention. Not by chance, perhaps, the fourth line begins, at the point of greatest poetic tension, with the adverb nikkama:"truly." This is followed by the one finite verb, ummaliñciti(vi), "you're upset," an example of the kind of Dravidian root that is common in Tikkana but that later gradually fell into disuse. The feeling of the verse as a whole is one of actual conversation, with a hidden subtext, in a natural syntax that continually frustrates the flow of the utpala-māla meter.

Perhaps it is this recurring frustration in the internal movement of the verse that produced a revolutionary problem and that encapsulates the tension between Tikkana's innovative power and the ingrained habits of recitation. Someone trained to recite, for example, utpala-māla verses in the oral style comes up against tremendous obstacles when he comes to Tikkana. Oral versification requires dhāra—the unimpeded flow of words—and dhorani, the "mode" in which the flow takes place. Shifting to another meter brings the expectation of another mode, but a similar flow. But in Tikkana neither happens. The reciter begins, stops, tries again, is again checked—very much like in ordinary speech, but utterly unlike the normal sung stanza. The combination of speech and meter, which is required if the poem is to speak, explains this difficulty, apparently keenly felt by the singers. It may also explain Tikkana's isolation within the tradition; his style was never taken up or imitated by later poets.

Indeed, the tradition itself recognized and commented upon this feature. Tikkana is said to have made a pact with his scribe, Gurunātha, from the potters' caste, to the effect that Gurunātha would record, without pausing, the poems Tikkana was improvising on condition that the poet never stop the flow of verse. If Gurunātha were to fail to keep his side of the pact, he would cut off his right hand; if Tikkana failed, he would cut off his tongue. The arrangement worked well until, at a point in the text where the internal narrator Sañjaya was describing the epic battle to Dhrtarāstra, Tikkana became stuck in the middle of a verse, unable to complete it. In despair, he

cried out to his scribe: "emi sěppudun gurunāthā" ("what can I say, Gurunātha?"). The scribe kept writing without pause, as usual, and the poem worked, since the poet's cry completed the verse, precisely according to the meter and meaning. The nasal ending of the verb, sěppudun, requires that kurunātha —"lord of the Kurus," that is, Dhrtarāstra—become gurunātha in written Telugu. Tikkana was reaching for his sword to cut off his tongue when the scribe explained to him that all was well with the verse, because of this simple rule.

This story, disarmingly simple in appearance, actually offers powerful expression to the peculiar boundary zone that Tikkana inhabits in Telugu poetry. It seeks, on the one hand, to rehabilitate him, turning him back into an oral poet, since at this period, real poetry is still perceived as inhering in the oral, flowing qualities of a verse. Oral poetics remained dominant, and "literariness"—with its concomitant freedom—was still not poetically acceptable. A literary poem, in this sense, had to conform to the dominant poetics. On the other hand, the story implies a recognition of the innovation that Tikkana had introduced into the tradition. The ambivalence is striking: the story attempts to mitigate the full force of this innovation by assimilating the poet to the older model. Tikkana, in the story, unconsciously fulfills this role, though his completion of the verse is really a cry of despair, at the moment before he will strike himself speechless, cutting off his own tongue. It is as if more deeply, unconsciously, the poet remains a singer. Gurunātha's origin from the potters reinforces this claim, since the potters, a "left-hand" caste, are closely linked to the singing of texts. The story also shows us the new value attached to writing down a text, fixing it graphically without relinquishing the still-valued singing mode (which depends on audible utterance). In fact, the verse "works" only when sung: in writing, kurunātha becomes gurunātha, the cry of despair to the scribe; in recitation, this remains kurunātha, an address to the Kuru lord. One can see, in this vignette, the whole burden of the transition that Tikkana articulates for this tradition.

This transition has further features related to the wider literary world of Tikkana's time. Tikkana refers to himself in the colophons to his work as ubhaya-kavi-mitra, "a friend to both [kinds of] poets." This somewhat enigmatic phrase has several possible meanings, and implies tension between two schools of poets. It could be a question of Sanskrit in relation to Telugu; or of śaivas in relation to non-śaivas; or Brahmins versus non-Brahmins (this categorization, however, is probably a back-formation that we tend to read into the early medieval period); or of oral/folk poetics in relation to the written and scholarly/literary genres, newly emerging from Nannaya's time. We have seen the originality that Tikkana brought to the reconfiguration of Sanskrit and Telugu, and we have noted the particular prominence Dravidian syntactic, especially idiomatic, patterns have for him. But there is a sense in which Tikkana clearly faced a coherent, contemporaneous counter-tradition—that

of the Vīraśaiva poets, such as Pālkuriki Somanātha, who made the non-Sanskritic dvipada meter their main expressive vehicle. In Somanātha's words, the opposition between Sanskrit and Telugu, or between the "high" campŪ style and the purely Telugu-based dvipada, is starkly articulated. Somanātha, in line with the iconoclastic and anti-Brahmin thrust of Vīraśaiva religion generally, pits himself consciously against Nannaya's literary style and the tradition that emerged from it:

Since beautiful, idiomatic Telugu is more commonly understood than heavy compositions of mixed prose and verse, I have chosen to compose this [work] entirely in the dvipada meter. Let it not be said that these words are nothing but Telugu. Rather look at them as equal to the Vedas. If you wonder how that can be, remember, "If a tŪmu is a [large] standard for measure, so is a sôla." Is it not generally agreed that the stature of a poet derives from his ability to compose great poetry in simple language?"

Dvipada, in the hands of Somanātha, offered tremendous promise, given its enormous range of syntactic variation and the hypnotic power of the chanted string of couplets. Lesser poets, however, were unable to equal Somanātha's virtuoso handling of the restricted couplet form (and even Somanātha himself, in his Panditârādhya caritra, fails to maintain the emotional subtleties and syntactic fascination evident in his Basava-purāna). But dvipada tends, in any case, to limit the active role of the listener (and we insist again that at this period poetry was always read aloud to a group of listeners). CampŪ, with its mixture of verse and prose and its dependence upon the elaboration of the single stanza, always allows space for breaks in recitation, for commentary and reflection. In this sense, campŪ, as a final product in performance, always has two authors: the poet who produced the text and the reciter who sings it, explicates it, and opens it up to new experience. In effect, the reciter re-creates the text continually. Dvipada, despite its powerful narrative capabilities, tends to monotones and lulls the listener into passivity. This is a structural feature of the genre, incapable of resolution even by a brilliant poet; it helps to explain the eventual marginalization of the dvipada mode, as opposed to the almost infinite exfoliation of the campŪ style.

Nonetheless, dvipada maintained a presence in the ongoing evolution of the tradition, at least through the seventeenth century.

In the fourteenth century, Gona Buddhārěddi composed a dvipada Rāmāyana (the so-called Ranganātha Rāmāyana), moving the style into a Vaisnava realm. Tension remained between this style and the more courtly forms of campŪ and, later, kāvya, a tension which reflects structural fault lines within the Telugu literary world. At the same time, the campŪ style, at least in the hands of Tikkana, clearly absorbed creative elements and linguistic features from the counter-tradition, especially from Somanātha. In this sense, Tikkana's innovation includes the mediation between these two worlds of Telugu poetry, as his self-given epithet, ubhaya-kavi-mitra, indicates.

Tikkana faced this cultural challenge directly, attempting a synthesis of the conflicting pulls. Populist and elitist strands—or, if one prefers, regional and pan-Indian/Sanskritic—come together in his poetry, which stylistically and thematically forged a new sensibility. The same move toward synthesis is apparent in his choice of deity, Hari-Hara, a composite form of Visnu and śiva whom he addresses in his invocations.

It is also important to realize the more generalized direction of aesthetic transformation that Tikkana brought to his reworking of the Mahābhārata text. The stylistic and syntactical features we have outlined are perfectly in line with this wider shift, which turns the Sanskrit epic into a Telugu family drama of local south Indian chiefs. Emotions are "Teluguized," situated in wholly familiar frames and integrated with a Telugu psychology. When the young hero Abhimanyu is cruelly killed in battle, the response that Tikkana produces from his audience—through the voices of Dhtrarāstra or Yudhisthira and others—is something like, "How could you do this to my grandson (or nephew)?" Here is Dhtarāstra after he hears of Abhimanyu's death:

And here is Yudhisthira, Abhimanyu's uncle:

If we compare Tikkana's articulation of this moment with that of the Sanskrit original, we immediately notice a striking difference in tone, despite the very close verbal correspondence between the two texts. In Sanskrit, Yudhisthira says:

Yudhisthira, whether speaking Sanskrit or Telugu, has definite tendencies toward self-reproach, guilt, doubt, and disgust with his public persona and role. This is a given. But listen to how differently he tells us this in the two languages. In Telugu, the first-person pronoun and first-person verbal forms dominate the utterance: "I should have known; I am the one who sent him; I could have done otherwise; I was greedy; what can I say"—and, finally, "I am bad." Everything is personal, immediate, and fully internalized, and grief speaks directly along with guilt. Moreover, Yudhisthira takes responsibility here for the rest of the family. It is a family tragedy, focused on a Deccan patriarch-hero, that is being displayed, before it becomes the story of a warrior's death. The quality of emotion is closer to home. In some sense, these emotions are also embedded in the Sanskrit prototype, and to that extent we can say that Tikkana is extremely faithful tothe original. But the Sanskrit Yudhisthira wavers, loses focus, and slips into abstract gnomic statements: "A greedy person has no understanding of his faults" (in contrast to Tikkana's "It was greed that blinded me to my faults"). Before this thought is complete, he is already elsewhere, anticipating Arjuna's furious revenge (and, as the Sanskrit text continues, the pragmatic usefulness of Arjuna's rage in the ongoing war). We lose, as he himself does, the full integrity of his grief, and are left with his confusion. Then, concluding the lament, comes Yudhisthira's usual inner move (missing from Tikkana's passage): he is sick of it all, indifferent to victory, and reluctant to rule; none of it has meaning for him any more. In short, he wants to renounce the world. The personal tragedy is already, all-too-easily, precipitated into an almost generalized cultural predicament.

What this implies, among other things, is that Tikkana's characters achieve a deliberately intended coherence in speech, in the way their actions and movements are described, and in their entire presence within the Telugu text. This is a function of a well-constructed written text, and is distinct

from the reader's usual experience of the major figures in the Sanskrit epic. Their shifting focus is a natural result of their position within an originally oral text, which we have in its recorded redaction. To state this point a little differently: the full dimensions of the characters within the Sanskrit epic depend partly upon the text as we have it, and partly upon the oral milieu outside it that supplemented the verses. That oral dimension is now lost when we come to the Sanskrit Mahābhārata, and the written text is venerated. In contrast, Tikkana offers a fully articulated, nuanced world, in which there is room for the characters to be, or to become, fully themselves. They can also cumulate internal experience and grow.

For a Telugu audience, there is something irreversible in Tikkana's reworking of the epic. While Tikkana adheres closely to the verbal text of the Sanskrit Mahābhārata, he has in effect re-created the text in a Telugu mode, alive with Telugu-speaking heroes. Once one has seen the Mahābhārata through Tikkana's lens, which emphasizes and selects elements of the original, it is almost impossible not to experience these elements even in Sanskrit—as when a hitherto unnoticed pattern is pointed out in a design and then can no longer be ignored. A good "translation," in this perspective, creates a new original. This is not a matter of adapting, imitating, or following. The new original preexists in the "old" original, before the translation, but it needs the translator to reveal it. (Put backward, this means that the original anticipates its own translations—perhaps an inexhaustible series—and depends on these translations for its own total expression.) This also explains why "translation," understood in this manner, can be the beginning of an entirely new literary culture.

To sum up to this point: Tikkana reveals an activist, imaginative drive toward fashioning the universe of Telugu literature and culture. This universe expanded under his tutelage to incorporate new levels and genres, including not only dharmaśāstra, grammar, kāvya, and itihāsa, but also the restless, antinomian śaiva world couched in a highly regional idiom. This incorporation of wider elements actually meant the co-option of more specific modes into what now became the mainstream of Telugu poetry, directly evolving out of Nannaya's original template. From this point onward, the distinctive, volatile, and creative fusion of elements inherent in the literary campŪ institutionalized itself as the expressive arena for most Telugu poets. At the same time, Tikkana's expansion of this mode, along the lines we have discussed, prepared the ground for the extraordinary developments of the fourteenth century.

POEMS AND ARROWS: NANNĚCODA

Before we turn to these developments, we may pause to consider the enigmatic figure of Nanněcoda, of undetermined date, but perhaps belonging

to the interval between Nannaya and Tikkana. Nanněcoda, unlike most Telugu poets, was a non-Brahmin—a king of the Telugu Codas ruling in OrayŪru, in Pākanādu (perhaps near NellŪru). There is a sense in which he offers us a non-Brahmin, or Ksatriya, poetics, distinct from both Nannaya and Tikkana on the one hand, and from the other non-Brahminic, dvipada poets such as Somanātha on the other. Nanněcoda explicitly identifies the conceptual pair mārga—"pan-Indian," "supra-regional"—and deśi, "regional," which later served to express the kind of hierarchy of styles operative within Telugu (along the lines discussed above). But for Nanněcoda, mārga means, simply, Sanskrit; for him, all Telugu was deśi.

Given this definition, we nonetheless hear Nanněcoda demanding a radical Teluguization of both modes in the poetry he regards as best (vastu-kavita):

This statement—and the long poem on the birth of the god Kumāra that it precedes—mark Nanněcoda as effectively the first kāvya-poet in Telugu. And yet he stands alone, a maverick whose textures and style were never followed. Telugu kāvya at its height emerges out of the poetic revolution propelled by śrīnātha. Nanněcoda seems to reflect a consciousness of a strictly regional or local poetics, which also produced full-fledged theoretical statements such as Ketana's āndhra-bhāsā-bhŪsanamu (in the generation after

Tikkana) and Vinnakota Pěddana's Kāvyâlankāra-cŪdāmani. Certain features of this regional poetics are intimated in Nanněcoda's aesthetic program: an elevated but deeply idiomatic style (perhaps referred to by his term jānu těnungu, also used by Pālkuriki Somanātha); a density and compression in poetic statement; a notion of conscious construction (kŪrpu) that, while common to all the classical poets, is conjoined with a belief in the innate or natural (naisargika) propensity of the poet to internalize all forms of knowledge; and a sense of the overpowering, physical effect of a good poem:

It is also striking that Nanněcoda introduces explicit attacks on unsuccessful poets (kukavi-ninda) who fail to meet the standards of this regional aesthetic, and also on bad critics. Elaborate images of what constitutes a good or bad poem are now present in the introduction to a kāvya-work:

Such attacks on the poetic antagonists become standard features of the introductory portions to kāvya works, but each time the antagonist is defined anew, in terms of the shifting and evolving poetic ideal. For Nanněcoda, poetic failure is keyed to the image of what is possible in the best poem—a compacted profusion of expression, an organic quality of liveliness, inside

and out. The connoisseur immediately recognizes these features, while the critic of poor taste is seduced by cheap effects:

Despite this highly articulated vision, Nanněcoda's model remained outside the mainstream of the classical tradition for reasons still not understood.

ś RīNāTHA: POETRY AND EMPIRE

By the fourteenth century, then, we have a ripening tradition, already extraordinarily rich in forms and resonance, with a range of available modes. An enormous effort to produce narratives, mostly from the epic and purānas, permeates the literary scene. These narratives are all in campŪ style, imbued with a Sanskritic atmosphere and imparting dignity to both language and theme. Mārana's Mārkandeyapurānamu, Ěrrāpragada's Nrsimhapurānamu and Harivamśamu, and Nācana Somanātha's [Uttara]harivamśamu are but a few of the major works. This vogue in purānas, in the campŪ style, continued all the way to the twentieth century. But the last decades of the fourteenth century and the first decades of the fifteenth also witnessed a great breakthrough in the literary-cultural domain with the appearance of the revolutionary figure of śrīnātha, arguably the most creative poet in the entire history of this literature. Once again, the poet was aware of his own innovation. śrīnātha is the first to attempt to produce in Telugu a version of Sanskrit kāvya, a task previously considered impossible or even taboo. Here is how śrīnātha speaks of his achievement, with reference to the Sanskrit original of śrīharsa, the Naisadhīya-caritra:

This statement of translators' protocol is unique to Telugu literature; it shows us an intense awareness of the new enterprise śrīnātha has undertaken. For the first time, a Sanskrit kāvya has been entirely transmuted into Telugu in a sustained and elevated style appropriate to the original text. What has śrīnātha translated? He lists a series of separate components— sound, intention, feeling, mood, figuration, and propriety—all of which add up to an attempt to reproduce the texture of the Sanskrit. Sound lies at the root of this attempt and generates the real problem, given the obviously divergent phonoaesthetics of the two languages. Here is where the poet's originality is truly tested. Often in this text, the Naisadhamu, śrīnātha appears to reproduce the phraseology of his model, sometimes to the point of lifting, verbatim, most of a verse. For this very reason, the tradition mocks him: when śrīnātha showed his translation to Sanskrit pandits, they laughed at him and said, "Take your Telugu case-endings—du, mu, vu, and lu—and give our Sanskrit text back to us." Even more trenchant an expression of this same view lies in the story that śrīnātha, upon finishing his poem, sought the approval of a young poetic genius called Pillalamarri Pina Vīrabhadrudu, whom śrīnātha found playing in the street with his friends. The boy, called upon to judge śrīnātha's complex kāvya, asked the poet how he had handled one of śrīharsa's arcane phrases: gami-karmī-krta-naika-nīvr;tā [literally, "having put into effect the verb ‘to go' by wandering through many lands"]. śrīnātha immediately recited his corresponding verse, with the "Telugu" phrase gami-karmī-krta-naika-nīvrtudanai (that is, the Sanskrit original rendered verbatim, swallowed whole by the Telugu line with the sole addition of the Telugu first-person and adverbial endings udan-ai). Here are the two verses, in an attempt at English translation; one should bear in mind that the Sanskrit original is deliberately opaque, erudite, and enigmatic, its diction elevated to a point of near-absurdity (which is, in a sense, the point) —while śrīnātha's Telugu rendition, although almost entirely couched in Sanskrit, has a happy grace and ease. The context is the first meeting between Prince Nala and the famous captive goose, who has seen the lovely Damayantī and wants to make Nala fall in love with her:

Like all beautiful Hindu women, Damayantī is so thin at the waist that one can legitimately wonder if her body has a middle part at all. The bird is evidently well-trained in grammar, as we see from the critical compound cited above, "recycled" by śrīnātha in the first of two verses based on the Sanskrit prototype:

Obviously, śrīnātha has completely "re-Sanskritized" the verse, which has only a single Dravidian root (the finite verb, kan tin, "I saw"); all the rest is a combination of elegant and playful invention (in the long composita) on the one hand, and a mellifluous cumulation of familiar metaphors and attributes (the conventional descriptions of Damayanti's beauty) on the other. Within this ludic reshuffling of the given terms of the message, the refractory compound in question now stands out in all its ostentatious obscurity

and in brilliant contrast to the clarity and predictability of the remaining assimilated Sanskrit forms. śrīnātha has both internalized and reframed this phrase in a manner that exposes its peculiar grandiosity for what it is, a plaything for pandits situated somewhere very close to incipient irony.

None of this, however, was apparently capable of impressing the young Vīrabhadrakavi, who is said to have remarked with scorn, on the basis of this example—like the Sanskrit pandits mentioned earlier—that śrīnātha's version of S or ai.

Reading through the Naisadhamu, one can sometimes see why he (speaking for the literary tradition) adopts this sardonic view. As always, the folk-literary-critical narrative makes an incisive comment, although it fails, in this case, to illuminate the subtleties of śrīnātha's Sanskritizing techniques, often deeply transformative of tone despite the surface impression of wholesale, almost mechanical transposition of entire blocks of Sanskrit. If anything, the Telugu Naisadhamu seems to bear out a happy speculation by Seferis:

My fancy sometimes reaches even this absurd limit: if all the poets of the world were permitted to use one word only—the same word—the good poets would still find a way to differ from each other and create with this single word different personal poems (a thought that verges on Zen).

We might also invoke the shade of Borges's hero Pierre Menard, "the author of Don Quixote," whose early twentieth-century version of Cervantes' classic presents us with passages identical to the seventeenth-century original, but entirely different in tone and meaning in their new cultural and temporal context. "The text of Cervantes and that of Menard are verbally identical, but the second is almost infinitely richer. (More ambiguous, his detractors will say; but ambiguity is a richness.)"

The story we have cited highlights both the radically innovative nature of śrīnātha's work and the tradition's resistance to it. A given verse of śrīnātha's Naisadhamu may look and sound remarkably like—indeed, almost identical to—śrīharsa's original verse, but in fact it is always something quite new, and entirely Telugu. This is the true miracle of transubstantiation. However, the tradition also mocks this miracle. In effect, śrīnātha has subverted the older distinction between mārga ("high," "Sanskritic") and

deśi ("regional," "local," "Telugu"). Until this moment, any Telugu work was by definition deśi. But from now on, the boundaries of mārga and deśi are not linguistic but purely hierarchical: a mārga work can exist within Telugu, in distinction from deśi-Telugu forms. It is this new claim to status that the story objects to, and that śrīnātha boasts of when he describes himself as āndhra-bhāsa-naisadhâbja-bhavudu, "the Creator-God of the Naisadha text in Telugu."

But the actual achievement can be more precisely defined. śrīnātha's claim is to have translated Sanskrit kāvya into Telugu kāvya. In fact, however, kāvya is not really a genre, but admits tremendous variation, and śrīnātha works his own particular transformation on the Sanskrit model. Again, we need to look at a specific example. Here is śrīharsa's description of the women of India in relation to the as-yet-unmarried Nala:

The Sanskrit is so densely compacted that the initial effort merely deciphers it, restoring some kind of linear syntax. śrīnātha, however, narrativizes the verse, producing an expanded and well-defined sequence of events leading up to the suggestion of Damayantī's unique beauty:

In the Sanskrit, the word order completely masks the narrative order: "Am I a match for him?"—to see themselves—in the hand—seeing him—by a beautiful woman—held—except for Damayantī—deflated—by which [such beautiful woman]—was the mirror—by a sigh—unclouded?" The reader struggles to reconstruct a more intelligible syntactic pattern, which would follow the natural sequence of events. It is this sequence that śrīnātha has strung out and re-imagined. Moreover, each phase of the sequence is marked by a repetition, in nonfinite form, of the previous finite verb: "They stare.… staring." This kind of repetition constitutes a figure in alankāra texts: mukta-pada-grasta, "picking up the earlier word."

Not only does the figure serve to highlight the continuity and its recurrent, and universal, elements, it also sets off the uniqueness of the poem's striking denouement, in which Damayantī comes to constitute a set of one. Her, and her alone, the mirror does not disappoint. Even she, it seems, has run to check out her beauty, with the same doubt in her mind that afflicts all other women, but in her case, the doubt is ultimately dispelled. This is not merely a technical triumph, the final decision in a beauty contest. What śrīnātha has succeeded in suggesting is that there is something unique in Damayantī's presence, a hint of subjectivity and entirely individual potentiality—almost as if she were herself emerging from the mirror. Of course, in a sense, the same conclusion is implicit in śrīharsa's verse, though it appears so densely and intricately compressed and so deeply hidden by the nonlinear syntax that one can reach it only by a process of logical deduction. It is this logical progression that śrīnātha turns into a lyrical statement, wholly unpacking the Sanskrit puzzle and ultimately projecting the image of a living, irreducily unique woman. Another way to state the difference is to notice how a near-total nominalization in the Sanskrit verse—a series of nouns and their modifiers—turns into a straightforward verbal series that completely dominates the Telugu poem.

While this example may be slightly more vivid than usual in śrīnātha has invented. It still tends more toward lyrical narrative than toward the deliberately nonlinear, timeless, self-reflecting expression of śrīnātha's reworking of the Naisadhīya text, it does illuminate the poetic mode that śrī-harsa's Sanskrit kāvya. We are still some way away from the full-fledged, autonomous kāvya-world of the sixteenth-century Telugu poets—for whom, however, śrīnātha has opened the way.

Naisadhamu is the earliest of śrīnātha's extant works. Already it contains many of the hallmarks of his mature style—a dimension of full-throated orality and musicality, where the words flow with ease without losing their scholarly elevation; a gift for producing, perhaps even improvising, a long Sanskrit compound without scuttling the Telugu syntax; and above all, the alchemical power to do things with language that bring a reality into existence. In a sustained way, these features blended together to produce, elaborate, and establish an entire temple, that of Daksârāma in the Konasima delta śrīnātha's Bhīmeśvara-purān amu became the foundational text for this shrine, which it more or less sang into existence in its medieval (Hindu) form.

Bhīmeśvara-purānamu is a symphony swelling slowly to a pitch of almost unimaginable presence, as the poet calls the god himself down into the shrine. This is a text which cannot be retold or paraphrased, only performed and re-performed, creating its own reality anew each time. It is, from a certain perspective, not unlike a Vedic text, pregnant with mantric power (in the sense of bringing a world into existence, not simply of controlling an already existing reality or coercing a presence). But these same qualities are abundantly evident in many of śrīnātha's individual stanzas, or even in single lines.

He had his own way of stating this aesthetic, at the start of Bhīmeśvarapurānamu:

Three features produce poetry for śrīnātha: a curved line that suggests the potential full (rounded) form; a firmness, almost toughness, replete with some destructive energy that is also somehow erotic and creative; and a fluid, seductive softness. At the center of the series is the goddess, angry, intent upon putting an end to time, her breasts thrilling at this prospect— but, for śrīnātha, they are still breasts, attractive, possibly nurturing, enticing, womanly. The compound that says all this takes up almost half of the verse, connecting to the first two lines in a single syntagma crackling with sharp dentals and retroflexes: kālânta-sphurac-candikā-parusodgādha-payo-dhara-sphuta-tatī-parayanta-kāthinyamun;. Here everything is distinct, sharply outlined, yet still musical, with a suggestion of tactile softness. This is just how a good poem should work—at the edge of temporality, almost transcending time itself; destructive if misunderstood or misappropriated; and rewarding to a devoted listener. Everything depends upon the blend, as the verb sambhaviñcu—a co-occurrence, a coming together—indicates. Or we might picture this as a process actually undergone in the course of listening to this, or any other, good verse: the articulated point at the end of time (kālânta), where disjunction and discontinuity are present with other harsh boundaries, is superseded by the liquid internal state of softness (sarasatva) which is continuous, nonspecific, untimed (cira-kālam)—an unbroken movement, a dance.

The result is a new kind of poetry, a new, dynamic line, resonant with oral energy, breaking the bounds of earlier poetic forms. And what is true of the individual line is also true of this poet's traditional biography, which sets up a new model for poetic identity. In fact, śrīnātha is the first Telugu poet to have a full-fledged, orally elaborated biography, structured around oral verses attributed to him. These verses tell, for example, of his competition with the scholar-poet Dindima Bhattu at Vijayanagara; the contest ends, of course, with śrīnātha's vicorty and the destruction of Dindima's bronze drum. Similar stories take śrīnātha to other courts and other kings. The poet moves throughout Andhra and even beyond its geographical range, from patron to patron, kingdom to kingdom, drawing together this spatial map into a virtual literary empire of which he, śrīnātha, is the emperor (kavisārvabhauma). This literary kingdom will eventually become the necessary prototype for the political structure created by the innovating Vijayanagara kings nearly a century after śrīnātha.

So powerful is the creative presence of śrīnātha that it engenders, within

his lifetime, parodic imitation. A younger contemporary, Vallabharāya, parodies the śrīnātha style in his carnivalesque text, the Krīdâbhirāmamu—a tour of the alleys and bordellos of Kākatīya Orugallu/Warangal. Vallabharāya also quotes whole verses by śrīnātha, comically reframing them. This unique text, mixing something like acute ethnographic observation with radical antinomian imagination, reveals the new ecology of genres that śrīnātha's poetic revolution made possible. In another sense, Vallabharāya may mark the outer limit of that revolution and the exhaustion of its original impulse. No one ever composed poetry like śrīnātha again. At the same time, śrīnātha refashioned the didactic purānic form and style that he in herited; each of the nominally purānic works of his that have survived— Kāśī-khandamu, Hara-vilāsamu, Bhīmeśvara-purānamu, and śiva-ratri-māhātmyamu—is effectively a self-contained and integrated composition, elevated and uniformly sustained in tone, focused in theme and intent. In short, these are proto-kāvyas, considerably removed from the earlier campŪ style.

All of this adds up to a moment of far-reaching transition. The literary horizon has expanded enormously; so has the internal complexity of the individual poetic line and the individual stanza. A powerful integration of oral poetic features with fixed, classical forms has become possible. Institutionalized modes of purānic or Brahminical ritual, linked with the great temples of the delta as well as with śrīśailam in the interior, have become reflected in literary production. Sanskrit has entered into the inner life of the poem in surprisingly innovative ways. Parody and other commentaries have appeared in relation to newly dominant texts. The image of the poet—his powers, his language—has shifted. A direct line leads from this set of features to Pěddana, probably the most "classical" of all Telugu poets, whose work is unimaginable without śrīnātha

—just as the Vijayanagara super-state could hardly be imagined without the role of this Emperor of Poets, kavisārvabhauma.

Pěddana takes the poetic line to new levels of intensity, luminosity, and lucidity. At the same time, the impulsive orality and expressive extravagance of śrīnātha have lost their usefulness for Pěddana, who invents a more controlled, economical, and highly reflective style. His is also one of the strongest voices in what might be considered a thematic shift in Telugu literature: it is not by chance that Pěddana chose for his text the story of Svārocisa Manu, the First Man. Manucaritra, which tells the story, focuses on issues of generativity in the definition and genealogy of humankind.

But similar themes turn up consistently in the works of contemporaneous poets, including that of the great patron-king Krsnadevarāya (1509–1529). His āmukta-mālyada is in a class of its own—a masterpiece of realistic observation and personal expression, stylistically unique. We must now attempt to define something of this far-reaching reframing of the poetic enterprise as we see it in the works associated with the heyday of the Vijayanagara court.

KāVYA: REALITY IN THE MIRROR

The early sixteenth century marks the acme of the last major supra-regional state-system in south India, that of Vijayanagara, with its capital at Hampi in the western Deccan. As we have said, the poet-king Krsnadevarāya (1509–1529) has come—apparently since the mid–seventeenth century—to exemplify this period of political, military, economic, and cultural expansion. His rule also witnessed structural change in Vijayanagara society, with the mobilization of a new regional-based aristocracy directly tied to the royal center at Hampi and to the person of the king.

The literati and courtiers drawn from the ranks of these families were the first audience for a major breakthrough in the literary domain. In the hands of several poets of genius, Telugu kāvya came into its own, assuming a form distinct from well-known Sanskrit kāvya modes. The Vijayanagara poets produced complex, visionary narratives that are self-contained, stylistically sustained, and thematically integrated; each such work also reflects the inimitable voice of its author, who claims credit for his creation.

Like Nannaya long before them, the major sixteenth-century kāvya poets, such as Pěddana, Krsnadevarāya, and the somewhat later Bhattu-mŪrti/RāmarājabhŪsana, knew they were doing something unprecedented. Pěddana, for example, has his patron-king, Krsnadevarāya, refer to him— the outstanding poet at his court—in the following terms:

This verse forms part of the king's invitation to Pěddana to compose a book on the subject of the First Man, Svarocisa Manu. In explaining his invitation, the king (as cited by the poet himself) defines the poet as āndhra-kavitāpitāmaha, "the creator of Telugu poetry," who has a living memory linked to the whole of the earlier tradition. Not only has the poet learned the texts, but he has internalized their meanings in a manner that provides depth for

his creative resources. In this way, this new type of poet combines two features: an innovative capacity (he is catura-vaco-nidhi, "a master of crafted speech") and a connectedness to the great texts whose meanings he reinvents from within his own capacity. This combination underlies the new poetic style and allows the poet to see himself as the creator of poetry itself.

Let us examine one of Pěddana's invocatory verses to the Manucaritramu in order to see something of what the new kāvya means. Here is a prayer to Ganeśa the elephant-headed god, seen as a child sucking at his mother Pārvatī's breast:

There are three superimposed elements of desire and illusion: the young child reaches for his mother's second breast, which isn't there (since Pārvatī is here the left, female half of śiva's androgynous form, Ardhanārīśvara); he finds, instead, the snake-necklace draped around śiva's neck; and he mistakes the snake for something no less tasty to him than the breast, that is, the lotus-stalk so similar in texture to the snake. Is the child frustrated? Not at all. Is his mistake a mistake? Only in a certain technical sense. In "reality," the superimposed or imaginatively recognized object is no less substantial than the one originally sought. This kind of creative illusion, or projection, is characteristic of the mature kāvya world, where the poet's creation can stand on a par with any external universe. The story he tells is thus an autonomous, fully existing, organic creation, produced entirely within language, and language here also contains the material world, residually present as memory, as we saw in the case of the definition of a poet above. Neither of these worlds—that of the poem or story and that outside it—needs to rival the other. True, there is some awareness of "illusion"—the god's mistake, substituting one thing for another in his mind—yet this apparent mistake has the quality of a happy and even generative projection. Compare this

verse, for example, to the common Vedāntic analogy of the rope taken to be a serpent; there, the mistake is to be corrected, just as normal, that is befuddled, perception is to be transcended by the experience of an ultimate, divine truth. But in this kāvya mode, the apparent mistake is really a creative necessity, not meant to be overturned but rather to be extended, played with, and enjoyed.

This entire process becomes a theme throughout Pěddana's great poem, in which the beautiful VarŪthini, who has fallen in love with the Brahmin Pravara and been rejected by him, "mistakes" a love-stricken gandharva male for her lost lover, makes love to him, and thus, through a further series of displacements, becomes the progenitor of the first human being. Everything depends on the creative illusion, which ultimately generates a real world. Pěddana's text is thus sprinkled with verses like the one just discussed, where one thing is lovingly taken for another. The kāvya poets are fascinated with this playful movement, internal to language and consciousness, and its existential and experiential consequences.

If we look at the way the above verse is put together, we see a single continuous sentence with a narrative sequence, in complex hypotactic embeddedness, entirely subordinated as modifiers togajâsya, the "elephant-headed god." Syntactic complexity of this sort constitutes a major step forward in the formation of kāvya. At the same time, the "narrative" is itself pregnant with cultural information that is simply assumed to exist in the mind of the listener: Ganeśa has an elephant's head; śiva is divided in his body into female and male halves; he has snakes around his neck; and so on. The verse has no need to retell the story in the more straightforward narrative syntax familiar from earlier periods. Rather, satiated as it is with the memory of this knowledge, the culture can revisit it only at a higher level of complexity, reintegrating the various elements into a new syntactic pattern where narrative becomes an adjective and the elements making up this narrative can be playfully interchanged. It is as if the language that contains this world had turned in upon itself in order to explore the details and arrangement of its own inner space. Within this space, projection, memory, mirroring, and perceptual disjunction are the main creative features.

We can illustrate this turn with two more "mirror" verses, each emblematic, in its own way, of the new kāvya style and themes. Listen to the way Krsnadevarāya begins his great text, the āmukta-mālyada, invoking his per-sonal deity, Visnu/Venkateśvara at Tirupati:

Once again we have three levels, sequenced and linked. One—although this level is almost entirely unexpressed in the verse—is that of the two figures standing, as it were, in some visible, external world. They are, however, seen not in this freestanding mode but rather as reflected in each other's ornaments, each thus seeing himself or herself on the surface of the other; this is level two. But we then learn that these reflections are really like externalizations of the deeply internalized images each one carries of the other in mind or heart (ātmal'andu). This is the third, perhaps most important level, which is made to coalesce with level two—that of the mirror images. One actually forgets, reading the poem, that in between the inner image and its reflection in the jewels there exists the full-fledged "objectified" presence standing before the poet's eyes (and of course, if we wished, we could regard his inner visualization and linguistic embodiment of the god as yet a fourth, and perhaps fifth, level). Indeed, the objective, outer stance seems intended only to bring about the coalescence of the depth image with its reflection. Put differently, the apparent reality—the mirror vision—is brought into relation with, or indeed merged into, the "real" reality of the inner image expressed outward. The several levels are present, quite explicitly, in the repeated forms of the verb tocu, "to appear, to seem to be, to occur in the mind": first the infinitive/nonfinite topa in the second line, referring to the reflections as seen by the "normal" lay observer (a face and its mirror image); then (paiki) toci in the third line, the inner images coming out; and finally the finite verb tocěn in the fourth, the complete, full appearance of the god and goddess. More precisely, the final tocěn explains the earlier topa; that is, it uncovers the depth and reality of the mirror image, which is no longer, in any sense, a reflection.

This rather nonlinear progression can be restated discursively. The mirror image is not of the outer, concrete object (were this the case, the mirror image would be unreal). What the mirror shows is not a reflection but a reality. The inner images of the two deities have emerged and now inhabit the space of the mirror. There is a notion of fullness and clarity in this "appearance," as intimated by the adverb astokatan, which bridges the gap between the penultimate and ultimate lines—the final emerging into visibility is full, whole, and literally "not diminished." Syntactically, the key to the

entire revelation lies in the marker ana, "as if": the Lord of Venkatam is "as if" there, "as if" he were standing visibly outside in some objectified dimension. In fact, he is less real than the image inside the mirror, which is, the poet insists, the actual inner reality brought into being. And once again, the sequence is couched in the form of a single complex sentence, embedding within it a series of clauses, each linked to one of the three primary levels of perception and existence. At the height of this sentence, when the epiphany is complete, we find the hint of subjunctivity, the "as if" that marks off what is full from what is less than whole. Listening to this sentence, replaying it in the mind, one is, in effect, sucked into the mirror, into a space of conflated contours and ever more deeply embedded levels of perception—an infinite, self-generating, self-bounded space, opened up by language and immensely real.

We could also say that, in kāvya of this kind, language has syntactically cut loose from external reference, although external reality continues to inhere in it in all its potential fullness. In this light we can understand better the extreme tangibility and concreteness of Krsnadevarāya's descriptions of reallife scenes and events. This is realism of a different kind than what we might encounter in Western literatures, where language subserves a supposedly outer reality. Here external reality subserves language, which enhances and intensifies its ontology. We have to emphasize that this view is totally remote from the romantic understanding of a purely internal world of felt essences, a world that supersedes the material one. There is nothing romantic about of Krsnadevarāya's kāvya. Rather, the poem, like the mirror above, holds the real presence, only making it more real than anything outside language.

A very self-conscious variation on this same vision occurs in RāmarājabhŪsana's invocation to the goddess of speech, Sarasvati, at the start of his Vasu-caritramu:

Speech—Vānī, who is Sarasvatī—holds in her hand an aksa-sara, a string of beads identified here as the phonemes that constitute language. But the

nonverbal music of her vina here melts these seemingly solid syllable-beads into a liquid state that is perfectly clear, transparent—thus reflecting the image of the goddess, or of language—and also eternal, beyond time or sequence. At the same time, this ultimate linguistic act brings into being a state of fullness, in which the self is inevitably and integrally present (ātma-pratipādakatvamu). Whose self? Whoever speaks. And since whoever speaks is, at the moment, the goddess of language herself, it is also her self that is fully reflected back to her with each syllable. One could also say that, by this definition, any poetic use of language marks, and creates, the living subject and his or her subjectivity. This is a process in which the inherent musicality of poetry overwhelms the congealed, routinized forms of everyday speech (which lacks the transparency of poetry). And if, in Krsnadevarāya's statement above, we find external reality subsumed within language, here it is the speaking self that is subsumed. Together, these two verses offer a complete picture of language in both its objective and subjective aspects, seen as completely without conflict; both are necessary for the fully creative work of language. The exploration of these aspects, from a point within language, is the outstanding task of the kāvya poet.

āMUKTA-MāLYADA: REALISM AND BEYOND

The new kāvya aesthetic, fully developed at the royal court in Vijayanagara in the first half of the sixteenth century, allowed for striking individual variation sustained through long, internally coherent works, though common to all of them is the fascination with the autonomous and encompassing powers of the word and with the internalization of this creative potential in the living human subject. An emergent subjectivity of a new type—in certain ways recognizably linked to earlier developments in śrīnātha and others, and to a cultural substratum of resilient individualism at work in the Tělangāna cult of heroism

—speaks to us in a chorus of mature voices from this period. One strand is evident in the courtly kāvya of Krsnadevarāya, who presents us with remarkable, minutely realistic descriptions of everyday life:

It is the height of the hot season: booths are set up for wayfarers, to give them water. The young women who pour water into the outstretched hands of these thirsty travelers are standing slightly above them. As they revive, the men forget their former distress and switch modes, and the women note this as a reliable signal for them to stop pouring. It all happens in the eyes, a rich and silent communication against the backdrop of the trickling water. The effect is akin to a compressed drama, in four lines amazingly full of movement: the men come shaking and shivering, stretching out their hands, with nothing but thirst on their minds; they drink and gradually recover, their heads initially lowered into their cupped hands; as thirst recedes, the heads are raised and the eyes look up, engaged in what a new frame has to offer—breasts, armpits, faces. Still, they have to pretend to go on drinking, so they look down again, or back and forth, though the game is transparent to the girls, who signal to one another with their eyes that it is time to stop. Still, the play continues with their smile, as if the poem had done away with closure. The poet's eye moves with the internal movement he is depicting, both physical and mental, as if a series of close-up frames were quickly strung together in sequence (syntactically, as usual in high kāvya, there is a single complex sentence with its series of dependent clauses). A tremendous activity stretches the verse to its limits, or beyond them, extending outside the poem. And this sense of bubbling movement also comes through in the phonological configuration of the lines: the early part of the verse is dominated by repeated dentals (t, d), highlighting the demand of the tongue, thirsty for water, striking dryly against the teeth; the middle section is rich in playful gutturals (k, g), indicating, perhaps, the pretense of drinking, once the initial thirst has been satisfied; and the poem ends with a powerful series of liquid bilabials, lips smacking in delight, or even pursed for a kiss. Moreover, the double nasals in the middle of the last line—sannalan nagiri—actually force the lips apart, revealing the teeth in a smile that imitates the girls' culminating smiles; there is no way to recite the verse without reaching this point. The reader/reciter is compelled to undergo the same process experienced by

the actors in the poem. This final moment also offers us, at last, the proper subject of the verse—prapāprapālikal, the "water girls"—as if the poet had finished his vignette and needed only to give it a title. So characteristically convoluted and dense is the texture that he could also be said to have signed it.

This verse is, of course, part of a longer passage devoted to the hot season (such seasonal depictions being a staple of kāvya), all this a prelude to the king's visit to his courtesan's house. Thus the erotic tinge foregrounds the narrative event that is to follow. More generally, however, it is important to note that verses like the above, with their pointed observations drawn from experiences of real life, are embedded in a text which is also erudite, philosophical, and at times dreamily imaginative, as if drawing on experience from another world. These seemingly disparate domains— meticulous realism, doctrinal erudition and precision, and baroque imagination—merge into a coherent whole organized by the narrative frame. There is much more that could be said about that frame and its thematic drives; something of the force implicit in the choice of story can, perhaps, be sensed in the dream-narrative that the poet relates at the beginning of the book, when his personal god requests him to compose the poem.

Everywhere, there is a startling profusion of detail: the Brahmin Visnucitta goes off to a scholarly dispute in Madurai, and his wife packs him provisions for the way. We hear exactly, in mouthwatering detail, what kind of food he prefers, how it is prepared ("tamarind and spiced jaggery in equal portions for sambar," cěrakadamu sābāl' Ūnpa jělagu sambārampu jintapandu, 2.97), and which servant should carry what. This same wealth of minute description is then applied to the actual philosophical debate. The same drive toward precise observation and articulation is operative throughout. There is no distinction whatsoever in level; the kitchen merits the same attention as the scholars' chambers or the royal court, and the poet-king is equally at home —alive, observing, internalizing, recording, and making connections—in all of these milieux.

His voice is absolutely distinctive, that of an integrated and self-conscious subject. He sees the world differently than any earlier poet, with an extraordinary sweep and magnanimity of vision expressed in a syntax that sometimes appears tortuous, replete with oddly compacted, sometimes phonologically harsh compounds. Rarely does a verse simply flow smoothly, in the more conventional matter. He also speaks of himself in this same highly energized, intensified, utterly unconventional mode; an example is in the colophon verses at the end of each canto:

One long, high-voltage Sanskrit compound contains a story: when Krsnadevarāya besieged Udayagiri Fort, Prahareśvara Pātra, the Gajapati king's uncle, held it successfully for some days. Krsnadevarāya became impatient and took a vow not to bathe that day until he had stamped on his opponent's head. Terrified, Prahareśvara Pātra sent his crown in lieu of his head, and Krsnadevarāya stepped on it and then bathed. The same self-confident, brazen excess runs through each of his verses and the book as a whole.

TEMPLE POETS: POTANA, ANNAMAYYA, DHŪ RJATI

Kāvya, of the kind we have been exploring, belongs to the royal courts, or to a courtly mode. It presupposes a highly educated, elite audience of connoisseurs and a sustained network of patronage (or a merging of poet and patron-king, as in the case of Ksnadevarāya, just discussed). Patronage of this sort also reveals the latent structure of power relations between these figures: although it looks as if the poet were dependent upon his patron, in effect it is the poet who creates the latter as king.

In stark contrast to this pattern and to the poetic works it produced stands the parallel and contemporaneous tradition of what we have called "temple poets," whose only patrons are the gods they worship. Such poets conventionally look with contempt at the court poets and their presumed sycophantic compulsions. The temple poet proudly refuses to dedicate his book to anyone but the god himself. Prototypical in this respect is śrīnātha's contemporary Potana, the author of the Telugu Bhāgavatamu, who says explicitly:

Potana's text, though couched in the campŪ idiom with its conventional range of meters and prose, is entirely permeated by the tone of devotional surrender. His textures are soft, fluid, and relatively simple, though lexically often erudite.

Often one senses that this work of passionate devotion, like so many of the earlier Tamil bhakti texts, is meant to evoke, or actually to conjure up fully and realistically—that is, to create—the latent or hidden presence of the god. A line leads directly from this foundational text of Telugu Vaisnava devotion to the rich developments at the great Tirupati temple in the fifteenth century, where Tāllapāka Annamacarya (or Annamayya) produced his corpus of songs for Venkateśvara-Visnu.

Here is an example of one of Annamayya's poems:

Annamayya uses a different register than the kāvya poets—a register developed for singing, in immediate and intimate tones. There is less aesthetic distancing than in the courtly works: Annamayya's songs are meant as direct communications to the god he worships, Venkateśvara, for whom he is said to have composed a song a day for many decades. These songs can be celebrative, playful, erotic, meditative, and contemplative, sometimes all at once. There is an old classification (attested already in the generation after Annamayya) of all the surviving poems as either adhyātmika ("metaphysical") or śrngāra ("erotic"). However, this broad division can be deceptive: the above poem, for example, is classed as metaphysical. The poems, known as padams, were engraved on copper plates in the Tirupati temple in one of the most expensive publishing ventures of all times. Tradition says there were over 30,000 such poems, though the surviving corpus is roughly half that figure.

These poems were apparently lost for some centuries and rediscovered only at the beginning of this century, in a locked room of the temple, although there is some mystery about this story.

The padam genre in Telugu began with Annamayya, in the temple setting, and was continued in the works of his sons and grandsons there; later poets, such as Ksetrayya in the seventeenth century and Sārangapāni in the eighteenth, produced padams for courtesan-singers outside the temples. Each such poem was set to a rāga and meant to be sung to the accompaniment of a stringed instrument. The sheer inventiveness of the poets is evident in the amazing spectrum of themes and imagined situations (often lovevignettes of every possible variety). This is a poetry of what we might call "moods," in the sense that each poem calls up a wholly unique and irreducible emotional coloring, resistant to any typologizing, and each poem stands alone, a single experienced moment, unconnected formally to any of the others.

From out of this mood-pregnant moment, a whole theology can be suggested. "Anyone would become like him": there is a hint, perhaps, of the goal of assimilation to the god, sāyujya, here somewhat ironically grounded in erotic obsession. And at times the god seems to turn his face away, to become inaccessible; the poet offers a rationale for this movement. What might border on ridicule, on the surface, actually hides a profound philosophical statement about god. He suffers from kāmâturatvamu, an addiction

to desire; this supposedly underlies his need to have four hands and thousands of forms. Yet just below the surface, we see, without attempting to deeroticize the verse, the god's wish to save his many beloveds, and the infinite variety of experienced personae with which he encounters them. He is truly different each time he makes love. The human truth of this vision, which speaks to the uniqueness of each subjective moment, emerges fully in relation to the divine subject, who is also the subject of the poem—and who is also, as the final crescendo intimates, the inner subject of each of us.

Although the poet is describing his god, the tone is that of coming into contact with a familiar person, intimately known. But this person has unusual attributes: for example, the long fingernails—actually claws—with which Visnu, as the Man-Lion, disemboweled the demon Hiranyakaśipu. These fingernails have here become part of a general depiction of the god's beauty. There is a consistent and subtle progression within the verses, each syntactically completed by the opening pallavi refrain, from the image of the human lover to the slowly crystallizing identification of him with the god of the Tirupati mountain—that is, with an entirely different existential plane—and then, at the culminating moment, with the inner ground of all being. This progression is carried along in a transparently light, even humorous style that manages to articulate the simultaneous distance and intimacy that the devotee feels vis-à-vis the god.

Annamayya was the pioneer of this style, but his direct descendants continued his work, eventually producing an immense Tāllapāka corpus, in various genres, including a hagiography of Annamayya by his grandson Cinnanna. The latter work, Annamâcārya-caritramu, makes the opposition between court poets and temple poets entirely clear: the Vijayanagara king Sāluva Narasimha is said to have asked Annamayya to compose a song for him analogous to one he heard the poet recite to the god;

. this request—natural enough in a period in which the king and deity were, in fact, merging into one within the new political culture we call Nāyaka (from the sixteenth century, the time of Cinnanna, on)—is said to have been met with violent scorn by the poet. The king imprisoned Annamayya, but the god freed him from his chains. This story may anachronistically render the typological opposition too starkly; it remains true, however, that Annamayya's songs reflect a highly sensitized subjectivity, which may require the space of the autonomous relationship between poet and temple deity rather than the burdens of relationship between the poet and another human being such as his patron.

Not all temple poets achieved this extremely personal idiom; not all of them used the space and freedom. The other major example, this time from a S context, is DhŪrjati, probably of the sixteenth century. The literary tradition insists that DhŪrjati underwent a transformation from one category to the other. Beginning as a court-poet with Krsnadevarāya, he became disgusted with this life, left the court, and came to reside at the temple of Kālahasti, near Tirupati in southern Andhra. There he composed two works: the Kālahasti-māhātmyamu, a kāvya-work on the foundation and local tradition of the temple, and the Kālahastiśvara-śatakamu, a century of highly reflective poems formally addressed to the god, śiva, at Kālahasti. It is in this latter work, in the productive śataka genre,

that we hear the profound individuality of this poet's voice:

Often the rhetorical address to the god is hardly more than a device allowing the poet to explore his own inner landscape—especially the darker reaches of this inner world. In this sense, the communication is really within the poet, between parts of the self. The individual stanzas are crafted in the standard meters; but whenever a depth of inner feeling becomes so intense, language becomes lyrical in the extreme, heartrending in effect. DhŪrjati is unusual precisely because of this set of features, but we find similar trends in the so-called prose-poems (vacanamulu) composed by other poets at other temples, such as Krsnamâcārya and Pěda Tirumalâcārya (at Simhacalam and Tirupati, respectively).

SEVENTEENTH CENTURY: GRAMMAR REVISITED

Mature, elegant, manifold: the tradition that produced the great kāvyas on the one hand, and the playful and precise mood-poems of Annamayya and DhŪrjati on the other, had expanded its range and the sheer quantity of literary activity to an unprecedented degree throughout the sixteenth century, well past the collapse of the Tuluva dynasty and the loss of the imperial capital of Vijayanagara. Indeed, the new style and forms produced at the courtly center rapidly spread throughout the Telugu-speaking region, including, especially, small-scale courts in remote parts of Tělangāna and further east, toward the coast. These dusty towns and villages, such as Nandyāla (home to Pingali Süranna) and Kāměpalli (in Palnādu, associated with Appakavi),

gave birth to powerful new currents of literary and philosophical creativity. In some sense, as we have said, the classical tradition culminates in these works.

By the mid–seventeenth century, a new synthesis was clearly called for. It was at this point that the retrospective orientation toward Krsnadevarāya was fashioned, possibly by a new elite of largely left-hand castes that had come into its own in the political sphere. A literary mythology focused its vision on the court of this synoptic "great king," even as a rich system of oral commentary on the classical texts came into being.

This systemic elaboration and reorganization of the tradition seem to reflect a perceived disturbance in the literary world, exemplified primarily in works from the far south, in the Tamil country, where Telugu Balija families had established local Nāyaka states (in Senji, Tanjavur, Madurai, and elsewhere) in the course of the sixteenth century. Telugu poets at these Nāyaka courts experimented with new genres such as yaksagāna, the courtly dance-drama, or the abhyudayamu, celebrative "biographies," now brought into the literary mainstream. Moreover, the themes that dominate this literature had shifted: in particular, there was a fascination with stories of social, moral, or sexual violation (Indra's seduction of Ahalyā, Candra's love affair with his guru's wife Tārā, and so on). Hitherto clearly distinct social or cultural categories were now mixed: thus the opposition between queen and courtesan was blurred (and courtesan-poetesses became a common feature of the courts); merchants and warriors fused in a world ruled by mobile, self-made men rooted not in inherited ownership of land but in the possession of acquired liquid wealth; and most striking of all, the king was now conflated with the god and worshiped in this mode by his poets and courtesans, who had become his devotees. The poet no longer created the king by his poetry but served him in the devotional mode; and poetry as such was no longer tied to the scholarly establishment, with its Brahminical ideology, but was in the hands of non-Brahmin poets, the kings and their sons, and women. A massive reintrpretation of the social order is implicit in the new literary works.

Partly in response to these developments, but also following on the new awareness of language and its power that emerges from the kāvya world, poet-scholars such as Appakavi took it upon themselves to reestablish the linguistic and literary value systems of the classical order. We can see in Appakavi's great text a mythology of grammar itself, projected backward onto the First Poet, Nannaya. According to Appakavi (reporting on what the god Visnu tells him in a dream), Nannaya's "original" grammar, the āndhra-śabda-cintāmani, was destroyed by his rival Bhimana; fortunately, Nannaya's student, Sārangadhara (the son of King Rājarājanarendra), had memorized the whole book in childhood. This same Sārangadhara was mutilated by his father for allegedly having coveted his stepmother, the beloved young wife of the king. But Sārangadhara managed to give Nannaya's grammar to Bālasarasvati, from Matanga Hill (at Vijayanagara), and a Brahmin from this same site eventually delivered the book to Appakavi.

This story produces a chain of transmission authorized by Nannaya's prestige as the first poet, and therefore also the first grammarian; there is a perceived necessity for these two roles to stand together at the start of the tradition, now nicely tidied up in the retrospective mode. In effect, however, Appakavi has himself synthesized the floating materials of grammar in a new authoritative system.

At the same time, the incorporation of Sārangadhara into the chain of transmission seems designed to address the themes of violation proceeding from the far south, where Sārangadhara's story is the subject of one of the major kāvya texts (Sārangadhara-caritramu of CemakŪra Venkata-kavi).

In line with this redefinition of the role of grsammar, Appakavi reveals a deep awareness of language as creatively reworked by the poet:

In this perspective, Appakavi forms part of a much wider reconceptualization of linguistic themes. This reformulation is not limited to Telugu alone; in Sanskrit grammar from roughly the same period, we find works such as Bhattoji Diksita's Siddhântata Kaumudi, which entirely transform the internal organization of the classical Pāninian system—in large part, away from the generative direction of the latter and toward a more prescriptive and linear pattern. Outside the realm of grammar proper, and reaching toward a radical linguistic metaphysics perhaps unique in the Indian tradition, we find the complicated narrative kāvya by Pingali SŪrana, KalāpŪrnodayamu, from the mid-seventeenth century in Tělangāna. This amazing work demonstrates "practically," one might say—in the course of telling its convoluted story—that speech is narrative, and narrative is reality, already implicit in the mere existence of subtle linguistic forms. SŪrana seems to take up and concretize philosophical positions, or intuitions, similar to those articulated by the great Sanskrit grammarian-philosopher Bhartrhari in his Vākyapadiya. His fascination with making things through language also extends to a playful work of sustained paronomasia (ślesa, the Rāghava-pāndaviyamu, which tells simultaneously the stories of the Rāmayana and the Mahābhārata.

Appakavi and SŪrana epitomize the widespread cultural interest in topics of language in this period that we find in the many oral verses (cātus) on metapoetic and metalinguistic topics. This orally circulating system of literary production and literary criticism also produced a specific image of the poet as the omniscient creator whose utterances can never go wrong, and who can thus make and unmake things—indeed, an entire world—with his words. We thus find at this moment of powerful reconceptualization and reorganization within the tradition—in effect, a time of stabilizing a classical canon and anchoring it in a shared universe of critical values and perceptions—several highly intertextual modes of articulating the uniqueness of language, especially Telugu language: Appakavi's confident and authoritative model of grammar, SŪrana's inventive narrative of language in its lifecreating and consciousness-creating aspects, and the cātu metalinguistic commentary on the literary corpus and on the poet's potentially transformative role. Central to all these modes is the assertion of an essential link between poetry and grammar. In the vision of the seventeenth-century authors, the purpose of grammar is not to tell people in general how to use language (or to describe such usage empirically, as in the early Pāninian school) but to help poets use language. For it is poetic usage, with its particular expressive and active, indeed magical, powers, that is language at its most real.

THE END OF CLASSICAL FORMS

This anthology closes with the eighteenth and early nineteenth centuries (Muddupalani and Tyāgarāja), but the genres and forms of the high-medieval

period remained active and creative through the whole of the nineteenth century and even into the early decades of the twentieth. The rich world of nineteenth-century Telugu poetry is still largely unexplored. Purānic composition, kāvya, śatakas, padams (and the allied, shorter jāvalis), metaphysical texts, a new Christian literature in Telugu —all constitute a wide-ranging literary ecology, continuous with the past but also responding to rapidly changing conditions. Certain forms even reached a new level of creative intensity under the impact, or threat, of the emerging print-culture—for example, the domain of oral improvisation and pyrotechnic memory displays (śatâvadhāna), as exemplified by the so-called Tirupati Kavulu (Divākarla Tirupati śāstri and Cěllapilla Venkata śāstri). By the end of the century, borrowed European forms—the novel, romantic poetry (the socalled bhāva-kavitvamu), modern plays, and short stories—began to appear, as the literary elite came under the influence of modern English education, thereby also distancing themselves from traditional learning.

The classical tradition had a powerful final flourish in the work of one of the greatest of Telugu poets, Visvanatha Satyanarayana (1895–1976). No one had a deeper understanding of the cultural transformation that Andhra was undergoing through colonial modernization. And like his great predecessors Krsnadevarāya and Pěddana, Visvanātha combined, in penetrating and lyrical expression, the three components of vast erudition, meticulous realistic observation, and the ability to create an entirely imagined world.

He composed in both modern and traditional forms, in complex modes; his novels and novellas incorporate traditional structures, and his classical genres include a modern sensibility. Among his outstanding works are a novel of Tolstoyan scope, Veyi padagalu ("A Thousand Hoods"), and a verse Rāmayana in six volumes, the śrimad-rāmāyana-kalpa-vrksamu. The latter work is of a magnitude and complexity on a par with the finest works of the classical authors. Despite this achievement, Visvanatha remains the most misunderstood of all Telugu writers, since the literary establishment has moved into a modern critical mode largely incapable of addressing his organic and innovative genius.

To conclude this introduction, we would like to give one short example of Visvanatha's style, a taste of the unique richness that appeared at the moment of the tradition's passing. Here is a verse from the first book of his śrimad-rāmāyana-kalpa-vrksamu, from the passage describing Ahalyā's liaison with Indra, king of the gods.

Indra has fallen in love with this woman, the wife of the sage Gautama; after a thousand years of longing, he approaches her one evening and is told by his beloved that the time is not right: "Has the cock crowed? (kodi kusěně)," she asks. Indra takes the hint and spends the whole night tossing and turning in agony, waiting for the cock to crow at dawn:

Probably no translation can begin to reproduce the power of Visvanatha's unprecedented Sanskrit compounds, bursting the seams of the Telugu syntax of the poem. These ancient words, yoked together in a relentless rush towards unthinkable climax, are forced to present wholly modern meanings, a consciousness that is exploded and fragmented by desire and memory. Nowhere in the medieval culture does Indra appear with this degree of tormented subjectivity, his mind torn to shreds by fantasy, pointedly described and effectively embodied in the shattering combination of words: manonugā-viśada-tanŪ-samârdra-viniveśana-darśana-dhī-jharī-sahasra-śakalitâksi-golunaku. Each of the units of this compound somehow stands alone, despite the supposition of syntactical unity; it is like a series of blows, each in a different place, mimicking the overwhelming visual attacks upon the nerve endings of the eyeballs: "Images of her limpid body—yielding, wet, / flowing, enfolding—came in torrents, / shattering his eyeballs into a thousand pieces / from within." This compound follows upon an earlier, softer one, in which Ahalya's provocative sentence has begun to echo in Indra's mind. At first, these words and their associated images are beautiful, but as they echo unendingly, obsessively, in his mind, they turn it into a chaotic marketplace; the next stage is the destructive torrent. Compounds like these, combining unusual depictions with totally unfamiliar forms (such as vipanī-krta, the mind "turned into a marketplace," or the nominals strung helplessly

together in the third line: tanŪ-samârdra-viniveśana-darśana-dhī …), puzzled the traditional pandits in Andhra, who dismissed Visvanatha in despair. The modernists, on the other hand, were unable to decipher verses like these even on the overt lexical level, which they found arcane; they therefore treated this poet as an anachronism. And yet Visvanatha's compelling presence, however misunderstood, could not be overlooked: violent storms swirled around his works for most of his life and even after his death. In view of the powerful modernity of this poet, we have not included him in this anthology.

As one reads through the following collection, it is good to bear in mind that the classical tradition was never monolithic; even when new genres or modes became prominent, revolutionizing the previous literary ecology, they never dominated to the exclusion of other, previous forms. Thus we see Nannaya's purānic mode continuing almost into the twentieth century, though kāvya-prabandha becomes central to literary experience from śrīnātha onwards. So-called "minor" or "marginal" genres, like śataka, continued to generate new works alongside highly visible, more elaborate or prestigious genres. To no small extent, this statement applies even today, after more than a century of modernist poetry shaped by the influence of an international "print culture." The Telugu tradition is one of enormous heterogeneity, with highly original impulses continuously embodying themselves in unexpected forms, often from the most remote and localized corners of the cultural universe. Even our tentative attempts at periodization and the description of a developmental sequence may hide deeper continuities. A similar statement can be made about categorical distinctions: court poets may not have needed anything approaching our image of a major royal court, and temple poets did not always belong to a major temple. Still, the Telugu literary sources consistently reveal the workings of an autonomous literary universe characterized by its own dynamics and by a fascination with certain recurrent themes. We have tried in the following pages to allow this tradition to sing in English in its many modes and forms and, insofar as English can allow this, its diverse and startling textures.