The Language of the Gods in the World of Men: Sanskrit, Culture, and Power in Premodern India
by Sheldon Pollock (c) 2006 by the Regents of the University of California
Published by the University of California Press, pp. 684 + xviii. (http://www.ucpress.edu/book.php?isbn=9780520260030)
[Blogging South Asian Publications thanks the University of California Press for permission to quote extensively from The Language of the Gods in the World of Men in this review. Indian readers should note that an Indian edition is available from Permanent Black in Ranikhet.]
reviewed by Rajesh
This is a remarkable study of language in India, making striking comparisons with language developments in Europe. Yet it is about much more than language. The key point of the book is the contrast between the development of a unified language (Latin) in Europe through the imposition of political power, and the total lack of imposed power in the development of Sanskrit as the unified language of politics across all of southern Asia (as far as Indonesia) in the first millennium C.E.
Pollock is careful not to overstate his theses, in fact often points out how little evidence there is to untangle complex questions. As he admits in his closing epilogue, “The language of the gods had a history in the world of men more complicated than any one scholar or book can capture with real adequacy” (568). But this is a paradigm-altering book that deserves wide study, and a lengthy account of the argument will be presented here. Many important points and most documentation for points cannot even be mentioned here. Readers of the book and of this extensive review are welcome to point out what they perceive to be key points that this reader has skipped over.
A statement in the Introduction indicates that Pollock is undertaking an examination outside traditional concerns.
Sanskrit studies, heir to a brilliant and imperious intellectual tradition that had set its own agenda in the important issues of the human sciences, has had grounds to rest content with addressing the questions predefined by this tradition – and the historical expansion of the realm of Sanskrit culture was not one of them. (11)
The Introduction also contains a hit at both standard theorization and one of the standard tropes of the colonial era.
Virtually all the reasons adduced for explaining vernacularization in South Asia as originating in a socioreligious rebellion are dubious. The presumed concomitance between Sanskrit and Brahmanism on the one hand and vernacularity and non-Brahmanism on the other does not hold for much of the period in question [1000-1500 C.E.]. The vision of Sanskrit as a sacred language “jealously preserved by the Brahmans in their schools” may not be the pure illusion of the colonial officer who gave it expression, yet it is undoubtedly something that developed late in this history of the language…. (28)
Part One contains seven chapters on “The Sanskrit Cosmopolis,” the last chapter being a comparison with language in Europe. Chapter one looks at the emergence of Sanskrit, which is largely lost to history and left to conjecture. The first appearance of the name Sanskrit, however, is in the Valmiki Ramayana, where Hanuman thinks to himself that in speaking with Sita he should not use saṃskṛta speech but rather a “human language” (44-45).
Pollock cites Patanjali’s reasons for studying Sanskrit grammar to demonstrate that “the purposes of Sanskrit language analysis were more or less exclusively tied to sacred performance and to the pedagogical practices, both social and discursive, pertaining to the knowledge of the sacred” (47). This is just one point supporting his conclusion about early Sanskrit.
The most plausible assessment of Sanskrit’s social and discursive world for perhaps the first thousand years of its existence on the subcontinent seems to be the following: At least two species of the language family usually called Indo-Aryan were in use as far back as we can see. One of these, Sanskrit, was a formal speech, viewed as correct by the custodians of the language and employed in particular contexts broadly related to vaidika [Vedic] activity; the other was a demotic speech with what are usually called Middle-Indic characteristics. Sanskrit thus had a mutually self-limiting relationship with textualized discourse on liturgy and liturgical knowledges: this discourse was composed exclusively in Sanskrit and Sanskrit culture consisted entirely of this discourse. (49-50)
The relation of Buddhist traditions to Sanskrit language provides a fascinating case study on the use of early Sanskrit. Vinayapitika 2.139 relates an incident where the Buddha forbids his disciples to put his words into Sanskrit and affirms that his words should be taught in the vernacular language of each monk (54). Pali was developed as a sacred language, but it “combined elements of such geographical disparity that it would never have constituted the ‘native’ language of anyone, certainly not the Buddha” (55).
Yet Sanskrit proved impossible to resist. By the third century C.E. “Buddhist Hybrid Sanskrit” had appeared, which Pollock considers “an idiom that seems less a failure to achieve Sanskrit than a continuing reluctance to use it fully” (56). Reasons for the embracing of Sanskrit by Buddhists remain a historical mystery.
What exactly prompted the Buddhists to abandon their hostility to the language after half a millennium – the first instance of giving up resistance to Sanskrit and giving into its power, a process that would be reenacted time and again in Indian history – and finally adopt it for scripture, philosophy, and a wide range of other textual forms, some of which they would help to invent, is a question for which no convincing arguments have yet been offered. (56-57)
Discussion of the Ashokan inscriptions sets up the first great change in the history of Sanskrit usage. Writing seems to have begun with Ashoka around the middle of the third century B.C.E. The documentary texts in the inscriptions were not composed in Sanskrit, but in Middle-Indic dialects (Prakrits). This was only to be expected, as Sanskrit was the language of liturgy and was restricted to that purpose (67). It comes as a surprise, then, when the first Sanskrit inscription appears around 150 C.E. in praise of the Saka ruler, Rudradāman. Pollock suggests that there might be as many as one hundred thousand inscriptions across the Indian subcontinent,
Yet nothing whatever has been discovered to unsettle the certainty that Rudradāman’s text marks a true break in cultural history. For the first time, self-consciously expressive Sanskrit, with all the enormous authority, power, and cultural value garnered by the very fact of its centuries-long monopolization and ritualization, was used in a public space, in bold letters for all to see, for the self-presentation of a living overlord. (69)
If this were an isolated instance it would not be so remarkable, but “from that point onward for a thousand years and more, political poetry would be made only in that language” (73). Sanskrit had shifted from a liturgical to a literary language.
“Literature” is the subject of Pollock’s second chapter, and distinguishes two distinct genres into which Sanskrit came to play a dominant role. One, as just mentioned in the example of Rudradāman, is praśasti or “royal inscriptional eulogy” (75). The other, a broader designation that can even include praśasti, is kāvya (75). “In the Sanskrit critical tradition itself kāvya came to be theorized as a species of discourse fundamentally different from the Veda” (76).
The Sanskrit tradition, since at least the second century C.E., has always affirmed that the first kāvya was Valmiki’s Ramayana (77). The remarkable newness of this type of poetic text, which was also adopted into political inscriptions (praśasti), is discussed at some length. The contrast with European traditions between the oral and the written is highlighted.
Long after writing became an everyday practice in the Sanskrit world, a bias toward the oral persisted; knowledge that is kaṇṭhastha, ‘in the throat,” or memorized, was invariably privileged over knowledge that is granthastha, “in a book.” Moreover, the representation of knowledge (or understanding or awareness) itself as impregnated by language-as-speech – and never language-as-text – radically differentiates the medieval Indian world from Latinate Europe of the same epoch. (82)
Pollock’s discussion of kāvya moves into an analysis of the development of the concept that there were three literary languages, Sanskrit, Prakrit and Apabhramsha (90). Prakrit often referred to multiple languages, but in terms of kāvya it was only Maharashtri. Literary languages were distinct from common languages in one significant way.
The one distinctive feature shared by the three, and only these three, was their availability across region, ethnie, sect, and time. Both according to Indic conceptual schemes and in actual fact, none was tied to any particular place, people, creed, or era….To be qualified for literature, it seems, a language evidently had to be universally available – it had to be, in a word, cosmopolitan. (100-101)
The cosmopolitan nature of Sanskrit, which quickly superseded the other two literary languages, is evident from its astonishing adoption as the language for political discourse across all of southern Asia. This is the central theme of Pollock’s third chapter, which begins with this outline.
The new political culture and cultural politics embodied in the public expression of power in Sanskrit spread across southern Asia with remarkable speed. Just to register this digvijaya, or conquest of the quarters – and the very unusual sort of conquest that it was – is to grasp something of the character and reality of the Sanskrit cosmopolis. Within a mere two centuries, in locales that ranged from Kashmir and Puruṣapura (Peshawar) in the foothills of the western Himalayas eastward to Champa (central Vietnam), Prambanam on the plains of central Java, and even beyond in the further islands of today’s Indonesia, from the Kathmandu Valley in the north to the southernmost reaches of peninsular India and even, periodically, Sri Lanka, there arose a shared, Sanskritic way of speaking about and conceiving of the nature of political power. (115)
The Prakrit that Ashoka introduced was the language of inscriptions for many centuries before this emergence of Sanskrit. The relationship of Prakrits and Sanskrit has long been discussed, and Pollock presents a clear picture based on his careful analysis.
The prominence of Prakrit does not reflect ignorance of Sanskrit, and the supposed concomitance between Prakrit and Buddhism as against Sanskrit and Brahmanism is a chimera. A more parsimonious, and historically more accurate, explanation is that the two idioms coexisted everywhere but had entirely separate discursive spheres from the start. By the early fourth century, as the Vakataka record shows, these spheres had begun to intersect in new ways as Sanskrit emerged from its ritual sequestration to take on unprecedented expressive tasks in public, relegating Prakrit to the mundane documentary. Just this division of labor was to be replicated with respect to the language of Place: Sanskrit would monopolize all ideational and expressive functions in inscriptional and other written discourse while assigning to regional languages the quotidian status and function they had in everyday life. (118)
So Sanskrit was never a language for everyday life. It went from a ritual language to a language of poetical expression of political importance which allowed for and even encouraged local vernacular languages for everyday use. Tamil is a useful case study of this reality.
For the first three centuries of Pallava rule [c. 2nd to 5th centuries C.E.], Tamil, the everyday language of their realm, was denied all political function. When it at last appeared in inscriptions, Tamil was wholly restricted to factual communication and would long remain so. (121)
This happened not only with Tamil but in all of the Indian subcontinent, and then even more mysteriously it spread east in times and ways that are not at all clear.
The changes found in the Pallava’s public discourse were part of a larger process that had commenced two centuries earlier and was virtually complete everywhere by the fifth. All across the subcontinent there came into existence, by a startling, nearly simultaneous set of transformations, a linguistically homogenous and conceptually standardized form of Sanskrit political poetry. Power in India now had a Sanskrit voice. And by a kind of premodern globalization – even Westernization – it would have a Sanskrit voice in much of the world to the east. (122)
Suddenly, from about the fourth century on, inscriptions written in Sanskrit began to appear with increasing frequency in the places now known as Burma, Thailand, Cambodia, Laos, Vietnam, Malaysia, and Indonesia. And though this style of political poetry died out rather quickly in Burma, it would continue far longer elsewhere (until the later thirteenth century in Cambodia, the mid-fifteenth century in Java). Just as murky as the causal grounds of the expansion of the cosmopolitan order of culture-power into Southeast Asia are the vector and dating of its dissemination. (123)
It seems likely that the spread of Sanskrit as the language of political power was from south India, but very little is understood about this process. What is clear, however, is that vernacular languages were not stifled, and there was no proselytizing or coercion involved in the process.
All across mainland and maritime Southeast Asia, people who spoke radically different languages, such as Mon-Khmer and Malayo-Polynesian, and lived in vastly different cultural worlds adopted suddenly, widely, and long-lastingly a new language – along with the new political vision and literary aesthetic that were inseparable from it and unthinkable without it – for the production of what were often defining forms of political culture. In itself this is a remarkable development, but given the manner in which it occurred – without the enforcement of military power, the pressure of an imperial administration or legal apparatus, or the promptings of religious evangelism – it is one without obvious parallel in history, except indeed for South Asia itself. (124-125)
Pollock documents and discusses all of the points being briefly summarized here and this skeletal outline of his presentation does not do justice to his scholarship. Numerous nuances are outlined in the engagement of the Sanskrit world with Javanese civilization, for example. The value Sanskrit conveyed to the kingdoms which adopted it was its ability
to make the real superreal, so to speak, by coding reality in the apparent impossibilities of poetic configuration. Since Sanskrit was used to enhance reality, it was excluded from expressing the everyday wherever in southern Asia the inchoate processes of vernacularization made local languages available. (146)
Thus Sanskrit poetics was all about praśasti (royal inscriptional eulogy). This is a problem for historians who want to find hard facts in inscriptional records, but instead find exaggerated eulogies. Pollock closes his chapter by pointing out that
In actuality, we need to grasp that it is not only the facts themselves that are of interest but also the interest in facts evinced by the historical actors….What people believe is the case and what they want others to believe is the case are as important as what is the case. (161)
These discussions lead into Pollock’s fourth chapter which discusses “Sanskrit Culture as Courtly Practice” or the outworkings of Sanskrit in the various kingdoms across south and southeast Asia. The most surprising of many fascinating details in Pollock’s study is the centrality of Sanskrit grammar to political power games in the first millennium C.E. “No sooner had Sanskrit become the premier vehicle for the expression of royal will, displacing all other codes, than Sanskrit learning itself became an essential component of power” (166).
In introducing this point Pollock points out that modern linguistics is indebted to Sanskritic linguistic theory.
Readers who know anything about Indian intellectual history are likely to know that the study of language was more highly developed in South Asia than anywhere else in the premodern world….premodern Indian thinkers were consumed by the desire to understand the mystery of human communication. And they had no peer in their explorations until, in one of the subtler ironies of Western intellectual history, Franz Bopp, William Dwight Whitney, Ferdinand de Saussure, Emile Benveniste, Leonard Bloomfield, and Noam Chomsky, learning both substantively and theoretically from Indian premodernity (being all of them Sanskritists or students of Sanskrit-knowing scholars), developed successively historical, structural, and transformational linguistics and, by these new forms of thought, invented some basic conceptual components of Western modernity itself. (164)
From the time of Panini kings were concerned about Sanskrit grammar, and across southern Asia it became important that a master of Sanskrit and its grammar be close to power. Sanskrit had once been separate from the “languages of men” as it was reserved for ritual usage; now it was separate from vernacular languages as it was reserved for elitist discourse including the praise of kings.
As we saw when considering the aesthetics of praśasti discourse, it was the glory of the king that was the preeminent concern of a literary composition in Sanskrit, not the particularities of the everyday world, for which the languages of Place increasingly came to be used – not in a lawless vernacular, unconstrained by predictable and universal grammatical norms and therefore in constant danger of degeneration – could the fame of the ruler receive permanent and even eternal expression. (182)
Chapter five develops this point and the claims of universality for the permanent and unchanging language of Sanskrit. This is partially illustrated from the Kamasutra manuals of sexuality that discussed in rather absurd detail with accompanying broad generalizations every aspect of sexual encounter, which Pollock understands as certainly not descriptions of empirical reality. Rather, the “sole purpose is to show that nothing is beyond the reach of Sanskrit sāstra [text]; everything everywhere, however intimate, is knowable and has become known” (198).
Chapter six focuses on the Mahabharata epic and its influence across southern Asia, to the end of further understanding of the Sanskrit cosmopolis. Pollock does not agree that the Mahabharata is mainly an oral tradition, yet it certainly is an exceedingly complex text.
Even a cursory analysis of the manuscript data available in the critical edition reveals that the majority of the books were transmitted not orally but in reasonably stable form based on written archetypes….This written transmission was vastly complicated by the fact that the text circulated as part of a living culture and grew and changed as dramatically as any living cultural phenomenon has ever done, producing a text-critical problem as large and intricate as any in world literature. (224)
Hundreds of Sanskrit Mahabharata texts are extant from all regions of India, and a study of these texts confirms the status of Sanskrit as a “universal” language.
The Mahabharata manuscripts are written in a wide variety of regional scripts. (That a truly transregional form of writing, Devanagari, would not come into wide use until the fourteenth or fifteenth century, at the end of the cosmopolitan period, is only another of the wonderful incongruities of the Sanskrit cosmopolis.) Today we tend to think of these scripts as having been illegible outside their regions, but in fact there was far more intercommunication between them, at least among specialists, than can be assumed by the usual method of extrapolating from the conditions of modernity….No regionalization of any consequence, in point of dialect or in the particulars of material or social or even religious life, can be detected in any recension or version. To the degree that regional versions can be said to exist at all – and the idea is not entirely modern but was known to editors in medieval times – they mark distinctions without differences. That is, while the text was undoubtedly expanded or contracted in one recension or another, this was unaccompanied by the least hint of localization. (229)
As the Mahabharata became a basic text across all of southern Asia it carried with it its mythology of sacred places, many of the places being actual in geography yet multiplied beyond their actual location.
Thus across Southeast Asia a thorough-going reconstitution of the cognitive landscape occurred, where not only natural features like mountains and rivers but also regions and kingdoms were identified with names borrowed from the Mahabharata. (234)
Kings all across southern Asia presented themselves as universal sovereigns ruling over territory defined in terms of the universality of Sanskrit and particularly the Mahabharata.
This was a political culture, accordingly, whose multiple universal sovereigns, like multiple Mount Merus, Ganga Rivers, and fields of the Kurus, violated no principle of logic. (246)
This expansion of mythological places runs concurrent with lack of substantive information about actual kings and kingdoms in India. The great Gupta empire dating from c. 320 C.E. is Pollock’s prime example.
But what data from the imperial Guptas themselves are actually available to us? The records issued in the name of the kings and queens of this dynasty, including seals, consist of a grand total of twenty-odd fragmentary documents and hardly more than 250 lines of printed text; from the founder of the lineage himself, Chandragupta, we have nothing; from Samudragupta, only four records. Adding to these all the documents produced by those who directly or indirectly declared their subordination to the Guptas, we can double the number, and if we include two dozen recently published copperplate texts, another three hundred lines. All told perhaps a thousand lines of text (some of which are only partial or simply repeat genealogical information) constitute the entire direct textual basis of our knowledge of the nearly three centuries of Gupta imperial rule – thin gruel indeed, even if these calculations are off by a factor of two or three or more. (250)
With the close of chapter six Pollock’s main case about developments in the use of Sanskrit over the first millennium C.E. comes to an end, with only the comparison with Europe and Latin remaining. Thus he summarizes four main conclusions on why Sanskrit spread across all of southern Asia as it did.
Though the features that constitute the power of Sanskrit – its various communicative capacities – and enabled it to do its cosmopolitan and imperial work have been addressed individually throughout this study, it is helpful to sum up the most important ones here. The first is undoubtedly the translocality that marked Sanskrit from a very early period. Claims to universal sovereignty would have been unintelligible if asserted in a deśabhāṣā, a language of Place; those languages were to become the medium of an altogether different – a regional – political conception [in the succeeding millennium]. The second feature, closely related to the first, is Sanskrit’s transethnicity, whereby such claims were not only intelligible but in a sense rational, since they comprised a view from everywhere in general and nowhere in particular. (254)
A third dimension of Sanskrit’s communicative capacity is its expressive power. This is primarily derived, not from its archaic liturgical associations – Buddhists, who numbered among the master poets of Sanskrit for a millennium, would have been entirely indifferent to such things – but rather from its aesthetic resources. These include Sanskrit’s ability to make reality in a way more real by making it more noticeable, more complex, more beautiful thanks to the language’s arsenal of formal and rhetorical attributes – the metrics and tropology that fascinated readers across the cosmopolis as well as the presence of a literary corpus offering successful exemplars of such linguistic alchemy. Neither of these resources was available to the languages of Place, which remained unelaborated and pre-aestheticized codes – precisely what documentary languages of deeds and contracts were supposed to remain – until they were transformed in the vernacular epoch….
Fourth, Sanskrit was endowed, as every language of the empire form must be, with the dignity and stability conferred by grammar. Only in a code constrained by a set of norms and therefore escaping the danger of degeneration could important symbolic goods such as fame find enduring expression. (255)
This amazing story of the development of Sanskrit is yet more intriguing when compared with Latin.
How Latin developed, very slowly and over the course of many centuries, from a local idiom spoken in the lower Tiber Valley into a supraregional language constitutes, as a distinguished student of the subject put it, “one of the surprises of history” [Mason Hammond, Latin: A Historical and Linguistic Handbook, Harvard University Press, 1976, pg. 39]. It may have been surprising that this happened at all, but there is little surprise regarding how it did happen: through an intimate and unambiguous dependence on a military-political project, first republican, later imperial, and then later Christian. Latin traveled where it did as a language of conquest: first, as the language of a conquest state, initially Roman but later Carolingian and Ottonian; second, as the language of a missionizing and eventually conquest church. (260)
The contrast could hardly be greater as Sanskrit spread into Southeast Asia without anyone in the Indian subcontinent knowing or caring that it was happening, and local culture and language were permitted to flourish.
Whatever the status of dharma in the Sanskrit cosmopolitan conceptual order, practical law remained resolutely local (exemplary here, if at the end of the cosmopolitan epoch, is the localization of dharmaśāstra in Thailand as thammasat). The supposedly built-in afflictions of Sanskrit culture – caste, patriarchy, Brahmanical power, and the like – are hard to demonstrate as necessary concomitants of the cosmopolitan package. The Khmers, who, to judge from the spectacular political poetry they composed, were full participants in the Sanskrit cosmopolis, developed nothing on the order of the caste practices found in India and were blithely indifferent to Indian gender inequality. Nor do we find in the Sanskrit cosmopolis anything comparable to the influence exerted by a core culture in a center-periphery world-system relationship that we find in Rome. There was no actual center to the cosmopolis, only a conceptual center – and precisely for this reason it was one that could be and was replicated in many different places. The progressive Sanskrit transculturation eastward was not a matter of much interest to the Indian mainland – neither as object of political ambition nor a source of cultural hubris – if we are to judge from the paucity of references in Sanskrit literary texts or geocultural knowledge systems to the world of Southeast Asia. (270-271)
A final point that this reader will highlight from Pollock’s analysis of the realm of Sanskrit as the dominant language of political expression is another contrast with Rome. Despite the ascription of universality to kings which is central to the role of Sanskrit, there is no evidence that any Asian king was ever deified in the way Roman emperors proclaimed themselves as gods.
To be sure, kings in India were constructed as “consubstantial god-men,” as I once called them, but the logical and political effects of this construction were, I suggest, entirely different from what was found at Rome….Whatever the complexities of such political-theological positions and views, three points can be made with reasonable certainty: First, if the Indian king was widely, perhaps invariably, viewed as a god-man, and if his icon might be displayed in temples…he was never the center of a royal cult and never the object of religious worship. Second, the supreme deity was irrelevant as a source of royal authority. A talismanic presence or apotropaic force? Yes, without doubt….But a granter of heavenly mandate, a justifier of rule, a transcendent real-estate agent awarding parcels of land? Never, not for Samudragupta nor for anyone in South Asia who followed after. Last, and concomitantly, the king’s transcendent god was never the god of a political ethnie. Many royal cities in India indeed had their divine myths of foundation (as late as Vijayanagar, 1340), and virtually every dynasty claimed divine origin. But no one, ruler or people, ever claimed at any time that God had chosen them or given them a land or provided them with guidance or enabled them to conquer other peoples and lands. (278-279)
The rise of Sanskrit from a ritual language of priestly functions to the great political language of southern Asia is rivaled in fascination with the story of its decline and eclipse when, seemingly without explanation, vernacular languages came to the fore and the era of Sanskrit ended. “The Vernacular Millennium” is the subject of part two of Pollock’s study, consisting of three chapters focused on Asia and two making further comparison with Latinate Europe.
Chapter eight looks at the beginnings of the vernacular revolution and the marginalizing of Sanskrit. Many languages across India and southeast Asia are presented as examples of this development, with the whole of chapter nine being a case study in Kannada.
When the regional languages first began to attain written form, starting in the latter half of the first millennium generally speaking, they were used exclusively as documentary idioms; it was only around the turn of the millennium that they came to be transformed into codes for political expression and, more or less simultaneously, for literature. (288)
For the first four to five centuries of their written existence, then, Marathi, Telugu, and Kannada produced not a single text of culture or power that we could identify as expressive, imaginative, workly – or, more accurately put, nothing that anyone in medieval India would have called kāvya or praśasti. The languages were all silent, literarily as well as politically. (Nothing whatever indicates the existence of earlier vernacular praśastis, or explains why, if they had once existed, they should have all been lost when Sanskrit examples have been preserved in abundance from all across these regions.) Such vernacular invisibility over the long term was the rule rather than the exception in the world of cosmopolitan Sanskrit. This fact – and its correlative, that vernacularization was not a necessary process but entirely elective and conditional – is demonstrated by the history of Tulu, the language of coastal Karnataka, which had no literary or even documentary existence until as late as the nineteenth century (and even since then, its written uses have remained highly circumscribed), or that of Konkani, a language of Goa and southern coastal Maharashtra, which was rarely if ever used even for documentary purposes until the modern period. In neither case has evidence of substantial earlier vernacularization been lost; rather, it was never produced in the first place. (290)
Once the vernacular languages came into use for political and literary purposes, however, more than just a change of language occurred. A transition from the centrality of the oral in Sanskrit to a focus on the written in vernaculars is also apparent.
…literacy was constitutive of vernacularization as a historical process and, what is more,…literacy often took on a cultural and conceptual importance radically at odds with Sanskrit’s nostalgic valorization of orality. Something of this complex transformation is suggested by the history of the word akṣara, “phoneme” or “syllable,” as it migrated from Sanskrit to Kannada (in its tadbhava, or derived, form, akkara). In the Sanskrit tradition the term had long been associated with the notion that the language is both fundamentally phonocentric as well as eternal and uncreated…, as suggested by its usual etymology: “that which does not decay” (a-kṣara). Akṣara also came to connote the Sound par excellence, the primal Sanskrit utterance oṃ. Thus when Kṛṣṇa in the Bhagavadgita asserts his greatness by declaring that “Among words I am the single akṣara” (10.25), he is identifying himself with this irreducible and eternal core of language. By the tenth century in Karnataka, however, the term had come to predominantly signify written letters, the knowledge of writing, and literacy-based knowledge in general. (307-308)
Kannada is clearly an area of Pollock’s expertise, as is clear in a newspaper article that will be referenced in the concluding section of this review. He opens his ninth chapter, which is a case study in Kannada, by elaborating on the importance of the data from this field of study.
Few local cultures of premodernity anywhere permit us to follow the history and reconstruct the meanings of vernacularization with quite the same precision as is possible for Kannada, the language of what is now the southern state of Karnataka. We can chart the shifts in cosmopolitan and vernacular cultural productions without interruption from about the fifth century on, based on texts that are for the most part securely datable – an almost unparalleled antiquity and chronological transparency. Much of the data is the hard evidence of epigraphs, and their quality is breathtaking. The region must be one of the most densely inscribed pieces of real estate in the world, with more than twenty-five thousand records (poems, charters, genealogies, donations, contracts) issued mainly by the courts of kings and their vassals….It is true that as many as half of the extant epigraphs of Karnataka remain unpublished, and that many early works have been lost. (330)
Pollock gives a summary of the dates involved in the development of Kannada.
One of the facts established by intensive epigraphical searching is that the literization of Kannada began in earnest – becoming something more than casual graffiti and entering into cultural and political history – no earlier than the start of the sixth century. For nearly three-quarters of a millennium before that point, all written materials in the region – inscriptions as well as literary, philosophical, and religious texts – were composed in Sanskrit or Prakrit. All ruling lineages expressed their political will first in Prakrit (beginning with Ashokan edicts in the mid-third century B.C.E.) and then, in the dramatic shift recounted earlier, exclusively in Sanskrit. Only slowly and tentatively did Kannada come into use for the documentary – though not as yet the literary – portions of inscriptions. (331)
One of the remarkable developments that Pollock documents is a gap of centuries between the development of vernacular languages and scripts and the introduction of actual literature in these languages.
Something on the order of four centuries separated the earliest written Kannada documents of the late fifth century from the discourse, literarily self-conscious in both form and content, of the Kannada praśasti text produced in the late ninth. (336)
This gap will be seen again in consideration of other languages. Kannada, however, developed its own grammatical science at a remarkably early date, particularly in the Kavirājamārgam (The Way of the King of Poets) (338) which dates to around the year 875 (332).
The Kavirājamārgam went some way in establishing the groundwork for a systematic reflection on and disciplinary organization of literary Kannada. The philologization attested first in the Mārgam is not only precocious but, with respect to other literary cultures of southern India, both autonomous and uncommon. Kannada grammatical science originated in complete independence from Tamil, the only tradition of comparable antiquity. By the same token, Kannada philology seems to have exerted little influence on its neighbors. Marathi, notwithstanding the Kannada example at its doorstep, was grammaticized first by nineteenth-century missionaries. Even so, the history of south Indian philology, to say nothing of its comparative history, is still very much in its infancy in the West, and many questions concerning the wider conversations among Deccani and peninsular intellectuals that may have taken place during the early centuries of vernacularization remain not only unanswered, but even unasked. (363-364)
For Kannada, as for Sanskrit, grammar is of fundamental importance. Yet an important distinction developed; for Sanskrit, literature is not the reality that legitimizes grammar, rather grammar legitimizes literature (365). Kannada became empirical, founding its principles on established literary usage (376-377). Still, the influence of Sanskrit is dominant in Kannada. Kannada is clearly a different language and its actual relationship to Sanskrit is not without complications.
…one thing is as unambiguous in the history of Kannada lexicography as it is in Kannada literature: while Sanskrit had fully penetrated the language, it nevertheless was always perceived as something other than local. (369)
Not only the structure of Kannada grammar is derived from Sanskrit, so is the entire technical vocabulary for the description of Kannada grammatical phenomena….for those instances where Kannada possesses a grammatical function unavailable in Sanskrit, Sanskrit terminology is invented. (373)
The premodern grammarians of Kannada fully understood that their object of analysis was an order of language different from the one [Sanskrit] whose analytical tools they adopted in order to describe it, and that therefore a tension between the two was inherent. They never maintained that the language was a derivative of Sanskrit, any more than the lexicographers and metricians did. Nor did they consider it a Prakrit (let alone an Apabhramsha), in contrast to some north India regional languages (Gujarati, for example, was conceived of as such by its poets as late as the eighteenth century). No grammarian ever adopted the diagnostic model available from earlier Prakrit grammars for deriving the forms of Kannada by transfer of rules from Sanskrit. (374)
Chapter ten moves beyond the realm of Kannada to outline similar processes and results in many other vernaculars. A brief analysis of Tamil is interesting and controversial.
Tamil country, as noted several times earlier, presents a more complex and contested history of literary culture than anywhere else in South Asia. Yet here, too, there is little doubt that major transformations of literary culture occurred in the centuries just before and after the beginning of the second millennium. The element of complexity is introduced by a unique, and uniquely obscure, vernacular literary prehistory. This includes a corpus of courtly literature, purportedly from the caṅkam [Sangam], or literary “academy,” of the first centuries C. E. (a dating that would be clearly anomalous in South Asian literary history), that gradually disappeared from the historical record – almost completely so by the mid-second millennium – until rediscovered at the end of the nineteenth century, as well as noncourtly hymns of Śaiva and Vaiṣṇava spiritual masters typically dated to the seventh or eighth centuries that were lost and rediscovered only at the beginning of the second millennium. (383)
Pollack then refers to “the legend of the caṅkam” (383) and indicates his position that only in the eighth or ninth century did Tamil begin to replace Sanskrit (384).
Equally as striking as the production of new literature in Tamil in the last two centuries of the first millennium was the philologization of the language. In addition, commentaries on such culturally central Tamil texts of the Tirukkuṟaḷ were beginning to reread the Tamil literary past through Sanskrit rhetorical categories. Indeed, the critical vocabulary of literary and linguistic analysis came to be deeply colored by Sanskrit, though determining precisely when this occurred is tied up with the Tamil tradition’s thorny problems of dating. (384-385)
Shifting the focus to north Indian languages, there was a much slower vernacularization process and Pollock has a theory which might account for that.
The single most arresting fact about developments in the north is that no grammaticization whatsoever was produced: none of the languages of Place – Assamese, Bangla, Gujarati, the varieties of Madhyadeshiya – had a written grammar until the colonial period. Even in Maharashtra, where Sanskrit studies achieved uncommon brilliance…, the Marathi language itself was entirely ignored. (399-400)
In the last section of chapter ten Pollock lays out his theory on “religion and vernacularization,” which runs against some standard positions.
If the vernacular polity created in southern Asia during the first five centuries of the second millennium remains obscure as a structure for exercising power, there is no doubt that the vernacularization project was initiated (in many cases) and promoted and practiced (in most cases) by those who exercised power. This judgment is based on overwhelming evidence, only a sample of which can be presented here; it is, however, completely at odds with scholarly opinion, which holds that religious consciousness and especially the religious movement now called devotionalism (bhakti) constituted the engine of the vernacular revolution. (423)
Impressive evidence for Pollock’s position is outlined; the early vernacular literatures were taking the place of Sanskrit as languages that supported political power, new cosmopolitan languages with established grammatical structures. Later, a second type of vernacular revolution developed where instead of formal language it was popular speech that was presented in literary form (and in this second development bhakti was indeed central).
A more detailed account of the later history of vernacularization would show that for some parts of India we can speak of two vernacular revolutions: one that was cosmopolitan in its register and divorced from religion, and another that might best be termed regional, both for its anti-Sanskritic, deśῑ idiom and for its close linkages with religious communities that developed distinctively regionalized characters. (432, italics original)
Two examples of this later vernacular revolution are presented. One is the Virasaivas (Lingayats).
…these “militant” or “heroic” Śaivas appeared in the northeastern Karnataka at the end of the twelfth century under the leadership of a loosely knit group of spiritual adepts and political elites….Although many historical questions persist regarding the true character of this project, the transformation in vernacular literary culture it effected is not open to dispute.
Three dimensions of this transformation are especially noteworthy in light of our earlier findings. First, the Vῑraśaivas employed a register for cultural communication that was radically different from the idiom of the cosmopolitan vernacular. This reflected no simple historical evolution of the language – the high idiom would continue to be used for literary composition for centuries to come….Second, they employed a totally new literary form, or better, an antiform – unversified simple prose – that they called vacana, or plain speech. The deśῑ quality of language was thus wedded to a consciously decultured form of composition. To be sure, the anti-aestheticism of the vacana should not be exaggerated. Though it is nonmetrical, any number of recognizable rhetorical and other workly strategies common to kāvya are put to use….Nonetheless, South Asia had not seen anything quite like this, an almost postliterary form of literature. Third,…with rare exception vacanas were never committed to writing by the producers themselves; in fact, few textualizations were made before the fifteenth century. The vacanakāras may thus be said to have attempted at once to retrieve a certain authenticity of orality and to undo a centuries-long cultivation of vernacular literacy. (433-434)
The Virasaivas were of course an anti-Brahman and anti-caste movement that over the centuries was accepted in society on the very terms of caste which they opposed. But Pollock concludes that the linguistic revolution they initiated “contributed to the demise of Kannada cosmopolitan vernacularity” (434); the high “grammatical science” that marked the first vernacular revolution was now laid aside in favor of simpler communication.
The second example provided is from Gujarat.
Gujarati vernacular culture commenced in the late twelfth or early thirteenth century, and for the following two hundred years and more the idiom employed was not just learned but exuberantly erudite. (434)
This formal Gujarati language was shaken by the great bhakti poet Narsi Mehta (Narasiṃha Mahetā, c. 1414-1480).
Although a Nāgara Brahman fully equipped with the Sanskrit culture of that caste group, he renounced both the social and the aesthetic forms in which he had been trained. Legend tells of his very personal relationship to the deity, however unspiritual the matters in which this relationship exhibited itself (the god’s help with letters of credit, dowry, and the marriage of his son). In addition to merchants, Narasiṃha’s socioliterary sphere included the untouchables (Ḍheḍha), who invited the poet to sing among them, to his lasting infamy in the eyes of his fellow Brahmans. Like the Kannada vacanakāras, Narasiṃha seems to have refused to commit his compositions to writing; no evidence is available of any manuscript tradition for almost two centuries after his death. The newness that entered the world of Gujarati culture with Narasiṃha – including the kinds of themes that marked his (sacred) biography, the social worlds in which he chose to move, the idiom of his poetry, and perhaps even some of its forms (if the prabhātiyuṃ, or lyrical morning prayer, was not Narasiṃha’s invention it certainly became indissociably linked to him) – would be fully registered in the consciousness of the literary tradition. This is signaled by the fact that by the mid-nineteenth century Narasiṃha had come to occupy the undisputed position of primal poet (ādikavi) in the vernacular. Narasiṃha’s was a new firstness, a second one created centuries after a very different, cosmopolitan-vernacular literary beginning. (434-435)
So there was a vernacular revolution related to the bhakti movements that developed across India, but this was not the birth of vernacular languages so much as the development of a more simple literary style based on previously-developed highly Sanskritized vernacular languages.
The new vernacularism, then – noncosmopolitan, regional, deśi in outlook – combined a different, local way of poetry making with a different, local way of spiritual being. (436)
Chapter eleven looks at the vernacularization of Europe, a very interesting study but outside the concerns of this blog. Chapter twelve, which closes the second section, gives comparisons and contrasts between European and South Asian vernacularization. One of the significant differences is that the romantic attachment to ethnicity and mother tongue speech that marked Europe was missing in South Asia.
One startling and unanticipated contrast between South Asia and Europe pertains to the affective conditions for vernacularization. It was earlier noted that nowhere in South Asia before colonialism did the emotive and naturalizing trope “mother tongue” find expression. Nowhere do we hear a discourse of friendship or love toward the vernacular; there is nothing comparable to what Dante called the “natural love of one’s own speech”… (473)
It is true that only a fraction of the tens of thousands of texts composed by premodern Indian writers in dozens of languages have been read by modern scholars. But nowhere in the texts they have read is it possible to point to a discourse that links language, identity, and polity; in other words, nowhere does ethnicity – which for purposes of this discussion we may define as the political salience of kin group sentiment – find even faint expression. It is equally impossible to locate evidence in South Asia for the linkage of blood and tongue so common in medieval Europe, or for cultures as associations restricted by so-called primordial ties. (475)
…whatever the reason, narratives of ethnicity and histories of ethnic origins of the sort that obsessed late-medieval Europe did not exist in any form in South Asia before the modern period. (476)
Another marked difference was that in Europe the decline in understanding of Latin forced the vernacularization of literature, and it was initially religious literature where the transition occurred. But in South Asia religion was not part of the process, and there was no indication that failure to understand Sanskrit played any role in the process.
In the South Asian case there is no reason to believe that Sanskrit communicative competence diminished to any appreciable extent anywhere, even including Southeast Asia (with the exception of Cambodia after the Angkor polity). This was certainly the case for south India during the entire period considered here (800-1500), and it seems to have been largely true for north India as well. Undoubtedly Sanskrit literary culture in the north was challenged by various new or newly intensified sociopolitical developments from about the twelfth to the fifteenth century, such as civil chaos in Kashmir, or disruptions in traditional patronage networks in places like Kānyakubja and Vārāṇasῑ during the consolidation of the [Islamic] Sultanate. But no evidence points to a widespread decline in Sanskrit proficiency, and in fact the efflorescence of Sanskrit science and scholarship in the succeeding period suggests just the opposite. Italy in the fourteenth and fifteenth centuries, on the other hand, may present a typical case where the loss of Latin learning among a section of the nobility and bourgeoisie prompted both the translation of Latin texts into the vernacular and the creation of a body of vernacular literature. The pioneering vernacular intellectuals in southern Asia, however, did not choose Kannada, Gujarati, or Javanese because they could not write in Sanskrit – and certainly not because they were at last paying heed to some archaic religious injunction to use the vernacular for proselytization. They had other, less instrumental purposes in mind, and less familiar but evidently meaningful values such as vernacular distinction and aesthetic difference. (478)
There is no question that the vernacular millennium in South Asia witnessed a dramatic regionalization of religious cultures, but a top-down determination of this process was not even a possibility on any Indian conceptual map. (480)
What really led to the decline of Sanskrit is simply unknown. During the period of vernacularization there were of course many social, political and economic changes that impacted both South Asia and Europe. Determining the influence of these complex factors on the vernacularization process is almost impossible, but it is clear that Europe and South Asia responded very differently.
The precise weight of the contribution of globalizing Islam cannot easily be calculated because it was only one component of other, larger-scale forces at work in the same period – above all, the new world network of trade that enriched both South Asia and Europe and helped make possible the burgeoning development of agricultural regions and urbanization. All these Eurasia-wide factors may well have conditioned, in ways that await adequate clarification, the development of vernacular literary cultures. But the conceptual and political variations found in the outcomes of the vernacularization process – like the variations in the imperial political and cultural formations that preceded them and upon whose foundations they were built – show that people in western Europe and southern Asia reacted very differently to these forces. (494)
Part three consists of two chapters and an epilogue on the theory and practice of culture and power. Chapter thirteen presents a critique of current theorizing about language and culture and power. Pollock is strongly opposed to any suggestion that there is a natural evolution involved in linguistic and cultural change. Clifford Geertz is one of the theorists challenged by Pollock.
The data Geertz took as empirical evidence corroborating his theory of language in society, making it valid everywhere and at all times (India being the best test case), had in fact been produced by Western modernity. The epistemic paradox here has been demonstrated often in the history of Orientalism, where colonial theory discovered in the archaic East only what colonial power confected in the first place. Viewed from a slightly different angle, linguism may be seen as an element of modernity creating the very past modernity claims to overcome. The supposed irrational attachment to language of nonmodern minds, rather than impeding the development of the modern state, may lie at the very heart of modernity’s project. (507)
The point noted earlier about ethnicity not being relevant to premodern India is again stated (511). Sanskritization as a natural and necessary process also comes in for critique. Western theory is seen as an obstacle to actual understanding of India and its history.
Sanskritization is presented either as completely random or as a ceaseless process, without beginning or end, everywhere available to explain transculturation, as if Sanskrit culture were a higher form of life toward which lower forms inevitably aspire. (514)
The notion of legitimation, along with Sanskritization, ethnicity, linguism, and cultural naturalism, are not the only obstacles that modern Western theory places in the way of understanding premodern India. Two other analytical frameworks are even more obstructive: the one that frames India as the civilization it always was and the other that frames it as the nation it never could be. (524)
Those last two points make up the closing chapter, with subheadings of “civilizationalism, or indigenism with too little history” and “nationalism, or indigenism with too much history.” The first is an attack on the idea of civilizations as static and definable entities.
What possible “conception of the world as a whole” could be said to characterize “Indian civilization,” which has witnessed struggles over conceptions of the world of the most incommensurable and irreconcilable sort for three millennia? (534)
…it is erroneous to think of a civilization, or a “great tradition,” as a unitary entity of whatever size. Indeed, a stable singularity called “Indian culture,” so often conjured up by Southeast Asian indigenists, never existed. (535)
From whatever vantage point we look, if we are prepared to look historically, civilizations reveal themselves to be processes and not things. (538)
The Western concept of nation is then also critiqued, particularly as applied to South Asian history. Particularly the novel and the epic are defined in relation to the Western nation-state in ways that simply do not fit with South Asian realities.
Whether [Georg] Lukács’ contrast [in The Theory of the Novel, 1920] – between the epic as the literary form of an integrated civilization and the modern novel as that of a world where meaning is no longer immanent in life – holds even for European literary history may be questioned. But it certainly makes no sense whatever for South Asia. If the Sanskrit epic can be said to be about any one thing, it is about the contested nature of social and political values. The Mahābhārata offers perhaps the most sustained study in world literature of the undecidability of conflicting moral claims – what the text itself calls the “subtlety” of the moral order – and this was something often noted by premodern Indian readers. For an influential ninth-century thinker, what the epic chiefly addresses is the collapse of social value: “[The Mahābhārata’s] purpose as a whole is the production of despair with social life” [Dhvanyāloka of Anandavardhana]. This is an interpretation of epic not as social fullness but as social abyss, of power not as perfected but as unperfectable since, as Vyāsa says, it is “slave to no man.” (554)
The astounding complexity of developing a theory about the development of culture in South Asia is seen in the estimate of the Indira Gandhi National Centre for the Arts that over thirty million manuscripts are still extant across India from the pre-printing press era, which Pollock compares to the paltry thirty thousand extant manuscripts for all of Greek literature (558). It is much easier to demonstrate the problems with current Western theory about India than it is to develop new adequate paradigms. The only option is
…forcing oneself to set aside the conceptual objects and interpretative apparatus constituted by Western modernity and nationalism without having much of anything else to put in their place. (563)
In his epilogue Pollock brings his insights into relation with the current reality of English as a global language.
Few things seem as natural as the multiplicity of vernacular languages used for making sense of life through texts – that is, for making literature. And few things seem as unnatural as their gradual disappearance in the present, especially from the pressure exerted by globalizing English. (567)
But this perspective is challenged by facts, since “the vernacular cultural orders that seem to be threatened everywhere were themselves created over time” (567), and those “new vernacular cultures themselves replaced a range of much older practices that affiliated their users to a global space rather than to a local place” (567).
This striking history should help us think
…about possible future choices in the face of what often seems to be the desperate alternatives available: a national vernacularity dressed in the frayed period costume of violent revanchism and bent on preserving difference at all costs, and a clear-cutting, strip-mining unipolar globalism bent at all costs on obliterating it. (568)
“It is no easy thing to think outside modern categories” (569), but modern categories based on European history do not fit non-European realities. There are indeed striking similarities in the rise of Latin and Sanskrit to “universal” languages.
The world of Sanskrit that came into being a little before the Common Era and the world of Latin that arose almost simultaneously were remarkably similar. After centuries of both discursive and geographical limitation, the two languages embarked on an extraordinary career of expressive elaboration and spatial dissemination. Their near-concurrent development as written codes for what in both worlds was conceptualized as this-worldly (laukika, saeculare) communication occurred after centuries of restriction to liturgical, magical, and generally supramundane (and largely oral) textuality. In addition, both quickly achieved unprecedented diffusion across what in both worlds was seen as virtually a global space. As poets from Gujarat to Java were writing Sanskrit, so poets from Spain to Mesopotamia were writing Latin. This universality pertained to substance as well as to extension. Both worlds evinced a similar style of cultural discipline, discernible in the cultivation of language and in the mastery of the canon of literature and systematic thought. Sanskrit and Latin alike were, in a very literal sense, written to be readable across both space and time – and read indeed they were. And they produced a sense of belonging that affiliated readers to each other across vast space and time. (571)
Yet the differences are as striking as the similarities.
In respect to power they differed as radically as the historical experiences that produced them….Latin traveled where it did as the language of a conquest state, and wherever it traveled – Iberia, North Africa, the Near East – it obliterated the languages it found. The Sanskrit cosmopolis was also created by movement, of course, though not the movement of conquerors. The coercion, co-optation, juridical control, even persuasion of the imperium romanum were nowhere in evidence in the Sanskrit cosmopolis; those who participated in Sanskrit culture chose to do so, and could choose to do so. Far from proscribing local script vernaculars, Sanskrit mediated their creation everywhere it traveled and often at the very moment it arrived. (571)
Centralization was the Roman way, whereas Sanskritic culture produced multiple centers, multiple “Romes” rather than one. (“Sanskrit cosmopolitanism duplicated locations everywhere; it was a world of all centers and no circumferences, with golden Mount Merus and purifying river Gaṅgās appearing ubiquitously,” 572.)
Just as there had been two different cosmopolitanisms, two different ways of being in the great space of culture-power – a compulsory cosmopolitanism and a voluntary one – so there were two different vernacularities, two different ways of being in the small place of culture-power – a vernacularity of necessity and one of accommodation. (574)
This brings one to the immediate situation of culture-power and globalization.
…the choice between the global and the local, whether in the production of culture or in the organization of power, may find some kind of resolution in the blunt refusal to choose between the alternatives, a refusal that can be performed in practice whatever the difficulty in articulating it in theory. (579)
And, the closing words of the book,
To know that some people in the past could be universal and particular in their practices of culture and power without making their particularity ineluctable or their universalism compulsory is to know that better cosmopolitan and vernacular practices are at least conceivable, and perhaps even – in a way those people themselves may never have fully achieved – eventually reconcilable. (580)
The breadth and depth of Pollock’s work has set a standard for scholars in many fields to build on, refine, and correct. Internet searches can locate some of the controversies stirred by his work (first published in 2006), although many discussions are locked in academic journals that are not easily accessible in India.
Prospects for adequate exploration of all the questions Pollock leaves unanswered do not seem very positive. In a striking opinion piece in one of India’s leading English newspapers, The Hindu, Pollock wrote in 2008 about his concern for the lack of historical literary scholarship in modern India. Anyone who appreciated this extensive review will appreciate reading his perspective here: http://www.hindu.com/2008/11/27/stories/2008112753100900.htm. And such people also need to get and read for themselves Pollock’s amazing analysis of the language of the gods in the world of men.