‘In these days, when we don't have any kingdoms worth the name, texts on statecraft are of no use for ruling the state, and they are useful only for historians of shāstra texts’.
—Veturi Prabhakara Sastri‘This country has seen the conflict between ecclesiastical law and secular law long before Europeans sought to challenge the authority of the Pope. Kautilya's Arthashāstra lays down the foundation of secular law. In India unfortunately ecclesiastical law triumphed over secular law. In my opinion this was the one of the greatest disasters in the country’.
—B.R. AmbedkarIntroduction
Past works on the nature and content of state-building in medieval South India have focused largely on the inscriptional corpus, and a limited set of narrative accounts, in order to support classic formulations of such ideas as the ‘segmentary state’ and ‘ritual kingship’.Footnote 1 In this essay, we return to some of the questions raised by our colleagues and predecessors in the field, but with a view to looking at ideological and ideational issues far more than concrete institutional arrangements. We should note at the outset that the specter of a perpetually receding horizon of universal concepts—those that can be used with equal confidence, say, for the analysis of pre-1800 societies in Europe, Asia and Africa—has taken something of a toll in recent decades. Is it at all legitimate to assume that ‘money’ existed in all or even most of these continents?Footnote 2 What of the ‘economy’ itself, or even ‘society’? Is the notion of ‘art’ applicable everywhere? Can ‘religion’ be found in most societies?Footnote 3 It is well known by now that many postcolonial theorists wish to claim that ‘history’ was certainly not present in any more than a tiny fraction of the societies they study, until European colonial rule apparently created the conditions for its worldwide spread as a hegemonic discourse. In other words, it is claimed often enough now that no fit whatsoever existed between these and other ‘-etic’ categories of the humanities and social sciences (with their uniquely Western origins and genealogy) and the highly varied ‘-emic’ notions that may be found in different locales and times in the world of the past, a claim that has become a source of anxiety for some, a source of indifference for others, and a ground for rejoicing for still others who see a positive virtue in ‘incommensurability’, which they perhaps view as akin to a (necessarily virtuous) claim for species diversity.Footnote 4 Related to this is the recurrence of older formulae on the notoriously difficult subject of translation, both from those historians and from those social scientists who claim—on one extreme—that everything is translatable, and those who are eager to sustain equally extreme claims of ‘malostension’ or ‘radical mistranslation’ as a perpetual condition, rather than a contingent (and even potentially reversible) consequence of specific procedures and circumstances.Footnote 5
It is of interest that even in this welter of relativistic claims, one category that few have sought to challenge in its universal applicability is that of ‘politics’. Why has this been so, we may ask? Perhaps the reasons lie not only in an embarrassment with the charged, and patronizing, largely Marxist category of the ‘pre-political’, but also in the fact that to deny the existence of ‘politics’ would be tantamount to denying the existence somewhere in collective human existence of ‘power’, a move that few if any in the academy today would wish to risk.Footnote 6 To be sure, we could follow Benedict Anderson in relativising power, and argue that the ‘idea of power’ in, say, Java was not the same as that in the West; but this would be quite different to denying its very existence or utility as a concept for analysis.Footnote 7 In the case of India, almost any universal concept that one can mention has recently been challenged in its applicability to the present or past situation of that area, with the notable exception of ‘politics’. Indeed, it is instructive in this regard to turn to an essay produced by a leading relativist amongst Indian social theorists, Ashis Nandy, who would argue that ‘politics’ is practically the only category that one can use as a constant to speak of the past 2000 years in India.Footnote 8 Yet, this argument, first defended by him over three decades ago, came paired with an important caveat. For Nandy wished to argue that politics in twentieth-century India was in fact a split field. If on the one hand there were those who practiced politics in the ‘Western’ mode, drawing upon concepts and notions that were all-too-familiar to western political scientists and theorists, others continued to understand and practice politics through a deeply ‘emic’ set of lenses, which is to say while using concepts that had no familiar equivalents in the western political vocabulary. To understand these concepts, and the working of this other field, Nandy went on to argue, it was necessary to return to a series of texts produced in the Sanskrit language in ancient India, which alone could explicate this deep-rooted and culturally specific vocabulary, involving (usually substantive and untranslatable) terms such as dharma, karma, kāma, artha, sanyāsa and the like.
In making this argument, Nandy was paradoxically drawing above all upon a claim that was first set out in colonial India, namely that the only source of ‘authentically indigenous’ concepts could be found in ancient texts in Sanskrit. To his credit, however, it must be stated that he at least posed the problem of whether a possible field of political thought or political theory might have existed in India before colonial rule. Later writers, even those who were comfortable with the notion that concepts of ‘politics’ could be applied to study moments in the pre-colonial Indian past, have rarely returned to this problem.Footnote 9 Those who have done so have usually drawn upon Persian-language materials, and a learned tradition that has consistently maintained that in Islamic societies at least, the idea of ‘politics’ had long existed under such heads as siyāsat.Footnote 10 This view is lent credence by a genealogical claim, wherein the common Hellenic roots of western and Islamic thinking on the issue can be pointed to; the problem then would arise with that part of India where Arabic and Persian did not ever come to dominate as the languages of intellectual discourse.Footnote 11
This is the heart of the issue that this essay seeks to address. We wish to argue that in reality a quite substantial and varied body of material can be found in South India between the fourteenth and the late-eighteenth centuries that attempts to theorize politics, while doing so neither in Persian nor in Sanskrit, even if it may bear traces of contact with bodies of material in these two ‘classical’ languages.Footnote 12 These materials may be found instead in the Indian vernacular languages, of which we shall focus on a particular body, that in Telugu (though a similar exercise could easily be attempted with materials in Kannada or Marathi).Footnote 13 Secondly, we suggest that most writers who have looked into the matter (and they are a mere handful, as noted above) have usually misidentified the location of such materials, by seeking it solely in the corpus known as dharmashāstra. Thirdly, we will attempt to show how the materials that we are fundamentally concerned with, and which usually term themselves texts on nīti rather than dharma (although there is some overlap in the two usages), changed over the centuries with which we are concerned. Nīti may be glossed here by such terms as ‘pragmatics’, ‘politics’ or ‘statecraft’.Footnote 14 Finally, we shall briefly rehearse an argument on how the status of these materials was transformed in the nineteenth century, when British colonial rule reclassified them in ways that were at odds with their place in the universe of knowledge in India in earlier times.
We should begin perhaps with a rapid and schematic survey of the political history of the region with which we are concerned, namely the southeastern part of peninsular India, in which Telugu had emerged already by 1300 CE as a major literary language. A series of kingdoms can be found here, some of modest size and pretensions, others that can be classified as veritable imperial structures. To summarize, the early-fourteenth century sees the demise of the rule of a fairly substantial regional polity, that of the Kakatiyas of Warrangal, and the emergence of a set of far smaller kingdoms.Footnote 15 After a hiatus, the fifteenth century then sees the emergence of the great empire of Vijayanagara, which dominates the region (as indeed much of peninsular India) until the late-sixteenth century.Footnote 16 The collapse of Vijayanagara power means in turn that the two centuries from 1600 to 1800 are marked by a complex period of contestation, without a single stable and hegemonic polity. The Mughals eventually come to play a substantial role in the region, but indirectly rather than as a centralized political structure.Footnote 17 In short, we can see an alternation, with two cycles of fragmented political formations sandwiching an extended central moment of a century and a half of imperial consolidation that is associated with Vijayanagara.
Although it was famously termed a ‘forgotten empire’ by Robert Sewell in 1900, it is clear that the memory of Vijayanagara remained very alive in South India as late as the beginning of the nineteenth century.Footnote 18 However, the lack of adequate lines of communication between a society that already possessed a centuries' long set of continuous intellectual traditions, and a new political power that had assumed the role of ‘civilizing’ a group of ostensibly uncivilized or partially civilized nations, was never more striking than at this early juncture of colonial Indian history. For the traditionally educated Indian intellectual of the early-nineteenth century whom the East India Company might have consulted, India certainly had a sophisticated discipline termed nīti, beginning from early texts such as the Arthashāstra and continuing until their time. There was a whole range of texts on dharma, beginning with Manu's Dharmashāstra (and dating perhaps from the early centuries CE), and also continuing through the medieval period both in terms of a manuscript tradition and by way of extensive commentaries.Footnote 19 But the British administrators and their native assistants in early colonial South India were primarily looking for ‘moral instruction’.Footnote 20 Of the two concepts in the Indian tradition that come close to the idea of morals—dharma and nīti—dharma was seen as somewhat unsuitable for moral instruction because it was too close to the religious world. Manu's celebrated Dharmashāstra was also deeply embedded in the varna and jāti order, and discussed legal matters relating to marriages, property rights and so on. Law courts needed these texts, to administer justice to Indians according to their indigenous laws. The story of Sir William Jones's efforts in this direction and Henry Thomas Colebrooke's translation of legal digests for use in the British courts is too well known to be repeated here.Footnote 21
At the same time, it was also easy enough to argue that there was a direct line of ascent between the medieval regional language nīti texts and the Arthashāstra of Kautilya, and thus to conclude that the regional language texts were derivative and, if anything, bad copies of an original (however elusive that original was in purely philological terms) and therefore not particularly interesting. Another problem was that since the authors of nīti texts invariably claimed to be poets, literary scholars of late-nineteenth and early-twentieth centuries, influenced by notions deriving from Western literary models, began by rejecting any formal literary merit in their texts and then showed no interest in analysing them seriously for their content. Doubly neglected, the regional language nīti texts were relegated to a sort of intellectual no-man's land. Yet, as noted above, native schools still needed moral instruction, and in the absence of an Indian equivalent of the Ten Commandments, or similar codes of virtue, teachers often turned to nīti texts to fill the need.
The principal focus of this essay is the transformation and development of nīti discourse from classical Sanskrit texts to early modern Telugu texts and their later use in the colonial period. Our interest is to show, first, that these texts demonstrate a lively change with time and context as guides to practical wisdom, and strategies of success; and second, that they are not concerned with religion and are therefore mostly ‘this-worldly’ (laukika) or ‘secular’ in character. A third point that is developed in the analysis is of how the late-nineteenth-century colonial interest in teaching morals in schools gave selective, and one might say, distorted attention, to some nīti texts while ignoring the bulk of the others. The sources of the discussion are mainly from Telugu with a few examples from Sanskrit, and Persian.
Some Ur-Texts
No Indian text from ancient times has arguably been as used and misused in the context of the twentieth century as the Arthashāstra of Kautilya.Footnote 22 The first edition of this text, from 1909, was produced in Mysore by R. Shama Sastri from a single manuscript (with a commentary by a certain Bhattasvamin) originating in the Tanjavur region. It had already been preceded by a first translation (in the pages of the Indian Antiquary) from 1905 by the same scholar. The text quickly attracted massive attention, and a number of other manuscripts came to light, mostly in southern India (in Grantha and Malayalam characters), with one of the rare northern Indian manuscripts being from Patan, from a Jain collection. The confident initial assertion that the text's author was ‘the famous Brahman Kautilya, also named Vishnugupta, and known from other sources by the patronymic Chanakya’, and that the text was written at the time of the foundation of the Maurya dynasty, has of course been considerably eroded over the course of the twentieth century. Despite the relative rarity of manuscripts, it is clear that the text was known to the medieval tradition in various forms, and that its author was considered to be one of a series of important ancient authors of nīti texts. The Vijayanagara-period work, Rāyavācakamu, tells us that the king Vira Narasimha Raya in the early-sixteenth century was accustomed to hearing recitations from various texts including Canura's Nīti, with ‘Canura’ being a distortion of Canakya.Footnote 23
The text of the Arthashāstra in its modern critical edition, which was not necessarily the received version in the medieval tradition, is of course quite astonishing in its ambition and coverage.Footnote 24 It is a highly detailed text, and not one that simply contents itself to enunciate vague general principles. The text also quotes earlier authors, often pointing to the difference between its author's own opinions (in the third person, as ‘Kautilya’) and those of others. A striking and oft-remarked aspect of the work is that a great deal of its content is markedly ‘secular’. To be sure, in the initial part, the text invokes Sukra and Brahaspati, and then the Vedas; but thereafter, such location devices or references seem to disappear from the text. The first chapter discusses the overall contents, and Chapter 2 (adhyakshapracārah) then begins by noting that there are normally four vidyas: philosophy; the three Vedas; agriculture, cattle rearing and trade (collectively vārtā) and law-and-order (daṇḍa-nīti). According to Kautilya, there are however those who follow the Brahaspati's line of thinking, believing that there are only three disciplines (vidyas) and the Vedas are really a mere façade. We then get a version of the āshrama system of social ordering followed by a description of material life, with no reference thereafter in this extensive chapter to anything that might be understood as ‘religion’. This is once again the case in later chapters on judicial and legal matters, criminals and how to deal with them, secret matters (yogavrittam), and the manner of dealing with other kings and kingdoms in (the themes of Chapters 6 and 7, respectively, maṇḍalayonih and shāḍguṇyam). The highly circumscribed place of dharma in the text has recently been summed up as follows by Charles Malamoud:
‘The originality of the Arthashāstra is that the science of government, the doctrine of royal conduct, is set out there in a perspective where artha appears in a highly limited form and not, as in the Epics or the Laws of Manu for example, where it is assimilated to the perspective of ‘duty’ (dharma). The question in the Arthashāstra, is not that of knowing how, while obeying his ‘duty of state’, the king contributes to order in the world and in society, or even how he guarantees it, but rather of what he should do to attain his ends: conquer territory and hold on to it. To be sure the two perspectives are not wholly incompatible, and many of the ‘Machiavellian’ precepts of the Arthashāstra also appear in texts that lay out the norms of dharma; and there are even some passages in the Arthashāstra that recall some principles regarding the final ends which are dharmic in nature. But all in all, the Arthashāstra does not justify the means by the ends: the means and the ends appear at the same level, and each means is a provisional end. The treatise sets itself the task of laying out in detail the modalities of royal action and to evaluate them in relation to its sole objective: to succeed’.Footnote 25
Unfortunately, we do not know a great deal about the history of the book's subsequent use until far later. The speculation of the past few decades is that it may date from the fourth-century CE, but it is really quite difficult to make a definitive pronouncement on the matter. Buddhist sources seem to have been quite negatively disposed both to the text—on account of its alleged amorality—and to its author as a personage.Footnote 26 We may note that the Kāmandaka or Nītisāra also comes from broadly the same period, but slightly later, and that its author Kamanda states that he knows the Arthashāstra, specifying that the text's author was Kautilya, also known as Vishnugupta. Kamanda also appears to be the source for the confusing claim that Kautilya was the one who broke the power of the Nandas. In a similar vein, the author of the Mudrārākshasa, the Sanskrit play of Vishakhadatta from about 600 CE, seems to have known and used the Arthashāstra.
Unlike later medieval texts that we will discuss below, the Arthashāstra is not aphoristic in nature. Its literary quality is in fact rather interesting, being written mostly in short prose sentences with some occasional shlokas in the middle, and one or sometimes more than one shloka at the end of each chapter, and yet it is composed in a way that does not lend itself to easy oral transmission in this form. It seems largely meant for readers of a written book, and once more demarcates itself from later texts in the fact that ‘Kautilya’ himself, whoever he is, still poses and is regarded as an authoritative author. We shall have occasion to contrast this with the strategy of later texts, which seek legitimacy from their acceptability rather than invoking and using a notion of authority.
A second text from the early period that merits some mention, and seems to slightly postdate the Arthashāstra, is Kamanda's Nitisāra, briefly noted above. This work is shorter and also far less detailed than that of Kautilya, but follows it largely in terms of tone and general content, being partly advisory and partly authoritative.Footnote 27 Again, this text is written in the form of Sanskrit shlokas, not particularly easy for memorization or oral transmission, but perhaps intended more for reading. This text survived far more clearly into the medieval tradition, appearing in a Telugu version in the later sixteenth century (about 1584) as the Āndhra Kāmandaka, with some additional material that the Sanskrit ‘original’ does not contain.
Nīti and its Opponents During the Medieval Period
A very active interest in creating nīti texts is found in Telugu from the Kakatiya period in Andhra, which is to say the period from about the twelfth to the mid-fourteenth centuries.Footnote 28 The emergence of a powerful dynasty of major rulers from the great centre of Warrangal and the conditions that existed for a general upward mobility among many communities in the Deccan apparently motivated many writers to produce such works in Telugu. Some of the authors of nīti books of this period were themselves kings or their ministers, and many were associated with people of power in some manner or the other. The Sakala-nīti-sammatamu (hereafter SNS), a major anthology of selections from a number of nīti texts in Telugu, is of particular interest to us because it demonstrates the popularity of nīti as a subject in medieval Andhra.Footnote 29 The compiler of this anthology, Madiki Singana, was a poet in his own right. In his preface to the book, he declares that nīti should have equal circulation everywhere like a coin with the stamp of the Sultan (suratāṇi), and appropriately enough he calls his book ‘Nīti acceptable to everyone’.
Singana lived in a period when a number of nīti texts were already popular, perhaps each one in a different subregion or community. In his preface to the anthology, he hence expresses a desire to produce a digest of nīti, and lists the names of books from which he has collected his selections. He thus notes that his compilation is of some 982 selections from 17 distinct nīti texts by known authors (we may note in passing that many of the texts that were available to Singana are now lost), several verses from oral tradition, some verses by unknown authors, and his own verses as well. Among the known authors from whom Singana quotes, some are either kings themselves or ministers closely associated with kings. Rudradeva I (1150–1195), who wrote Nītisāramu, was a king of the Kakatiya dynasty; Sivadevayya (1250–1300), who wrote Purushârthasāramu, was the adviser and minister of Kakatiya king, Ganapatideva, and Baddena, also known as Bhadrabhupala, who wrote a particularly celebrated book called Nīti-shāstra-muktāvaḷi—better known as Baddenīti—is considered by modern Telugu scholarship to have been a king, from the Telugu Cola family.Footnote 30 Not much is known about this last poet-savant who addresses himself in his verses with royal epithets, except that he lived sometime before Singana (who himself flourished in about 1420), and that by the early-fifteenth century, his book had acquired considerable popularity, as is indicated by the short title which Singana uses when he quotes from it. The other nīti writers whom Singana quotes are mostly unknown, with the exception of Appappamantri who wrote a Telugu version of Bhoja's Cārucarya, a book of advice about healthy habits for wealthy people to follow.
Singana classifies his selections under 47 categories covering a range of topics related to kings as well as commoners—courtiers, physicians, pundits, and of course accountants and scribes (karaṇams). Two things stand out from Singana's anthology. In the first place, it does not invoke an other-worldly authority in any place. The goal is mundane, this-worldly, and the only thing that counts is success in any profession. However, it is not an ‘amoral’ text, as the desire for success is considered acceptable as part of a good human life, and it is implicit that success should be achieved within the framework of ethical conduct. The only concept that might suggest a Hindu ‘world-view’ of some sort is that a certain number of the verses refer to the scheme of the four goals of life, the caturvidha purushârthas (that is dharma, artha, kāma and moksha) of which artha and kāma, profit and pleasure, are the most significant areas upon which nīti texts focus. Even this reference, from the tone of its use, does not seem to be particularly religious in the context. While we do not have access to Sivadevayya's text in its entirety to see if it deals with the other two purushârthas, i.e. dharma and moksha, we know that no other extant nīti text deals with them, and in the use of later texts, for instance, the Sumati-shatakamu, the phrase purushârtha-paruḍu simply means a successful person.
It should also be noted that SNS for its part does not include even a single verse from the thirteenth-century dharmashāstra work, Ketana's Vijñāneshvaramu, a Telugu work based on the Sanskrit mitâkshara commentary of Vijnanesvara to the Yajñavalkyasmriti. This, we suggest, emphasizes the conceptual separation that already operated in these authors' minds of nīti from dharma.Footnote 31 For Ketana's work, we should note, followed in the standard, rather Brahmanic, dharmashāstra tradition of normative texts. Its author was a close relative (probably the nephew) of the celebrated Tikkana, who seems to have instructed him and guided him in writing this text.Footnote 32 Ketana was also the author of two other texts, one a Dashakumāracaritramu, an entertaining book of stories, and a grammar of Telugu, Āndhrabhāshābhūshaṇamu. He, like Tikkana, seems to have been creating an intellectual culture of a conservative and ‘revivalist’ kind, as we see from a close reading of the huge Mahābhārata that Tikkana produced at much the same time in Telugu.
To gain a sense of Ketana's Vijñāneshvaramu it may be useful to turn to the vyavahārakānda section of his text, which—though a relatively short section of the whole—starkly brings out the contrast we wish to develop between dharma and nīti texts. Here is a passage where he sets out his conception of rulership:
So we see here the clear evocation of the idea of pāpa (sin) as the ultimate punishment for incorrect action even in the context of statecraft. At times however, as noted by Malamoud in the classical context, the texts of nīti and dharma do converge, as when certain procedures are discussed (for example, on how to collect evidence in the context of a trial, or some other practical affairs). However, often enough, even the flavour of judicial considerations matters considerably, since texts like that of Ketana imply a strong caste variation in trials and punishments, and even seem directly to echo ideas from the Manu Dharmashāstra. Thus, we have the following example:
Where then, may we ask, do dharma and nīti texts in fact overlap without a great deal of tension? This is on those rare occasions when dharma texts deal with rather concrete commercial matters, such as the passage in Ketana dealing with how to write a promissory note (patra).
In general, however, Ketana's text is everywhere marked by a manner of thinking that reflects the dharmashāstras, and is consequently anxious above all to protect and defend the caste hierarchy as the most important aspect of the functioning of the polity. Nowhere is this clearer than in the passages where he gives ways of testing the four varṇas, to see if they are telling the truth or not.
In contrast to this somewhat soft treatment, Kshatriyas on the other hand are to be tested by fire, Vaishyas by water and Sudras by poison. Thus, for Kshatriyas:
It is hence clear that different tests are to be administered to different castes, a feature that markedly does not appear in nīti texts. The division of property among children finds extensive mention in Ketana, as well as the circumstances in which it goes to other kin.
Thus, the difference between the vyavahārakāṇḍa of a dharma text like that of Ketana, and sections dealing with similar matters in a typical nīti text are rather clear. Divine intervention (daivas) is constantly invoked in the former, the notion of sin (pāpa) is brought in, and punishments are explicitly hierarchized by caste. Even judgement is a ritual, requiring the chanting of mantras. In general, we may note that in this vision of things, punishments suggested with regards to castes lower down in the hierarchy (including scribal groups) are very heavy, and most of the discussions, including even those on how murders should be investigated, wind up having a strongly dhārmic flavour about them. The example below demonstrates this amply:
The role of the king is hence clear enough; he is, in large measure, the guarantor of the caste hierarchy and the protector of upper-caste males, but also the defender of their virtue—even against themselves. The examples below make this perfectly clear, and reinforce our notion that we are dealing with a socially conservative text.
The text does occasionally adopt a mildly humorous—or if one prefers, ‘realistic’—tone, but this is far more exception than the rule. One example of this appears in the same section.
At the same time, Ketana is a strong defender of royal authority, which he sees as requiring defence with an iron hand and the most severe of deterrent punishments. Hence:
All in all, then, this is a text that is remarkable for its censorious tone, and marked desire to regulate the moral life of society, rather than the harmonious combination of its parts in some form of social equilibrium. Virtue, for Ketana, must be produced, and if that production requires pain—whether physical or financial—so be it. Even gossips and malicious speech are seen by him as requiring regulation in some form, and that too by the king.
In a similar vein, ethnic slurs, or insults based on caste, are not to be allowed, in this most ‘politically correct’ of utopias.
In other sections, notably the ācāra-kāṇḍa, many passages seem to bear a close resemblance to Manu's Dharmashāstra, at times literally and at other times in spirit. A great preoccupation of the author, Ketana, is with the mixing of castes and the potentially negative effects of this phenomenon. Further, the gender roles are distinctly asymmetrical in this vision of things, all the more so in the context of intercaste relations. Thus:
Further, unlike what would find in nīti texts, it is understood that the rights of women are far more limited, and that they can be unilaterally disciplined for a number of faults, often merely on the basis of accusation. A last verse from Ketana below demonstrates how thorough-going and consistent a vision he embodies.
To develop the contrast, and the opposed visions that we have been suggesting inhere in the different genres, we should now turn to the nīti tradition of roughly the same period. In the nīti texts that were written during the Kakatiya period, by such writers as Sivadevayya and Rudradeva, the localized nature of the king and his kingship is quite evident. Even though the king they address is portrayed as a strong monarch, he is not an emperor ruling over multiple regions or extensive domains. The advice given relating to the protection of durgas (fortresses), for dealing with spies, and for invading the enemy's territory, the conduct of battle and so on, is not on a scale anywhere suggesting a large empire. Yet, the advice is practical and clearly derived from real experience of the administration of a kingdom. We may take for instance, the following excerpts from Singana:
• A king who does not command, is like a king in a painting, (good only for looks). If a king doesn't punish anyone who defies his command—even if the wrongdoer is his own son—he does not rule long.
• To allow merchants to take as much as they want is to ruin your people.
• If you don't make scales and measures uniform, it means you effectively permit thieves to go scot-free.
• If a king increases taxes, that effectively prevents (foreign) goods from entering his country.
• Wherever a letter might come from, a king should never disregard it. It is only through letters that a king knows everything—from alliances to enmities.
• Not killing a criminal amounts to killing a host of gentle people. All that you need to do in order to kill cows is to spare a tiger.
Some of the quotations in the SNS are clearly influenced by a traditional Sanskrit model of kingship, for example when the king is equated with god, quoting Manu's Dharmashāstra:
• The king is godly, and that is what Manu says, and he should be treated as such, and wise people should not treat him otherwise.
• Even if he is a boy, a king should not be treated like an ordinary mortal. He is god, and that's how he should be treated.
• The king may be bad, but the servant should serve his interests.
• If he [the servant] should leave his master for another to make a better living, the new master will never respect him for his loyalty.
However, in the same anthology we find some advice regarding bad kings. It is interesting to note that this advice comes from writers who perhaps served kings themselves in various capacities such as scribes (karaṇams), or soldiers.
• If anyone has caused you harm, go and complain to the king. But if the king himself harms you, who can you complain to?
• If serving a ruler causes incessant pain to the servant, the servant should leave such a master right away.
• He may be rich, born in a good caste, a strong warrior beyond comparison, but if a king is an ignoramus, his servants will no doubt leave him.
• If a king does not distinguish between the right hand and the left, a precious diamond and a piece of glass, it is humiliating to serve such a king—no matter how great a warrior you are.
• A bad king surrounded by good people turns out to be good. But even a good king is difficult to serve if his advisers are bad.
• A king who enjoys hearing stories of others' faults, who enjoys putting people through trouble, and steals other men's wives, brings calamity to his people.
The authoritative figure of Baddena is generously quoted in the SNS, and has some fascinating instructions to a king in his Nītishāstramuktāvaḷi. Contrary to the later importance karaṇams acquired in managing the affairs of the kingdom as ministers and scribes, Baddena strikes a note of caution against too much dependence on the minister. In his words:
• A king should not direct his people and his servants to his minister for all their needs. The king should be his own minister and treat the minister as an assistant.
The major writers on nīti whom Singana quotes in his SNS are already aware of the whole nīti tradition before them, including the Sanskrit Arthashāstra text. Besides, closer to hand, we find medieval texts from the Deccan, such as the Mānasollāsa of the twelfth-century Calukya king, Someshvara III.Footnote 45 Such works as these can certainly be seen to participate in a culture of political realism, and thus give the lie to those who have argued that pre-colonial politics in India was conceived along purely idealist lines. At the same time, the genre of the ‘Mirror for Princes’ is well known in the Indo-Islamic context, where a number of such texts exist both from the time of the Sultanate of Delhi, under the later Mughals, and from the regional Sultanates such as those of the Deccan.Footnote 46 Such texts, often written in Persian, are themselves at times influenced by Indic models such as the Pañcatantra, known in the Islamic world through its translation as the Kalila wa Dimna. Yet, they also bear the clear imprint of the non-theological perspective on kingship that had emerged in the Islamic lands in the aftermath of the Mongol conquests, when Muslim advisers and wazīrs struggled with the problem of how to advise kāfir rulers and princes on the matter of government, without taking them into murky and controversial theological waters.Footnote 47 The ‘Mirror for Princes’ genre ranges wide, and attempts to do everything from forming the prince's musical tastes, to refining his table manners, but the core of the matter is usually politics, both in the sense of diplomatic relations between states, and relations between a prince and his companions, or between different elements in a courtly setting.Footnote 48
The authors included in the SNS appear to be aware of these different traditions, and even draw upon them quite explicitly.Footnote 49 Yet, in contrast to the typical ‘Mirrors for Princes’, these authors offer a top-down, hands-on vision, partly rooted in pragmatic experience, partly creatively adapting the existing literature of nīti-statecraft. This is no armchair pontificating but a largely practical synthesis reflecting the political, economic and institutional changes of the fifteenth century. Still, highly individualized statements that can be attributed directly to the book's author Singana, do alternate with verses that seem to be lifted from standard nīti-texts about politics and kingship. Nonetheless, we are left with a total impression of a unique concoction of pragmatic wisdom, specific constraints, an inherited normative politics.
An Imperial Interlude: Krishnadevaraya
Singana wrote in the fifteenth century, and the immediate textual heritage he had available to him came from the period of the Kakatiyas. These were rulers who had dominated a relatively well-defined regional space in the eastern Deccan, and their preoccupations were very much reflective of that fact. In the case of Singana, we may suspect that the political landscape had fragmented even further, and that the kings he referred to were ruling over domains that would qualify a few centuries later as no more than zamīndārīs. But this was certainly not the case by the latter half of the fifteenth century, when a new, diverse and complex polity had emerged to control much of peninsular India south of the Tunghabhadra river, namely the state that is normally known as Vijayanagara (from the name of its capital city).
Normative texts on kingship, or statecraft, are hard to come by for fifteenth-century Vijayanagara. But we are far better served for the sixteenth century, and the times of the Third (Tuluva) and Fourth (Aravidu) Dynasties that ruled over Vijayanagara. A particular high point in terms of literary production, including that within the nīti genre, is the reign of the Tuluva monarch Krishnadevaraya (d. 1529).Footnote 50 When Krishnadevaraya ascended the throne in 1509 it is clear that a number of crucial problems regarding political management still remained to be resolved.Footnote 51 One major concern in the mind of the king was to make himself generally acceptable, and secure an area that encompassed more than one region, one language and one religion. The king's self-perception given to us eloquently in his major work, Āmuktamālyada, suggests that he sees himself as a Kannada Raya, a Kannada king, while the god to whom he had dedicated his book was a Telugu Raya, a Telugu king. Without anachronistically invoking regional nationalisms and language loyalties in the context of the sixteenth-century Deccan, we can still see local polities conflicting with each other and wary of dominance by someone from the outside.
Another way to formulate the dilemma that this king confronted is in terms of an enduring tension between local and trans-local forces. There is a consistent effort to conceptualize some basis for a trans-local polity that could extricate the state from its constant resubmergence in diffuse local contexts. A striking element in this conceptual effort lies in the king's own dynastic origins in one of the most marginal, and recently conquered localities—the western coastal plain of Tulunad. A kind of upstart, whose own family inheritance dictated that he prove himself outside the family context, finds himself articulating, at times somewhat inchoately, a vision of trans-regional, highly personalized loyalties.
Once a trans-regional state system is conceivable, its ruler runs up against its external boundaries. The manyam forest regions (especially the northern and northeastern frontiers but also implicitly to the south-west in Kodagu, or Coorg, and the Western Ghats) thus figure prominently in the Āmuktamālyada's section on rāja-nīti and require special treatment. External boundaries, however, coexist with the internal wilderness, as we see in a verse about a farmer marking off his field and then slowly making it free of stones and other impediments. But the text is also marked by a consistent suspicion, at times bordering on hostility or even contempt, for peoples like the Boyas and the Bhils, who could be found both at the border regions of the empire (in the north-east) and at the internal frontier. A prose passage within the nīti section thus advises the listener: ‘Allay the fears of the hill-folk, and bring them into your army. Since they are a small people, their loyalty or faithlessness, their enmity or friendship, their favour or disfavour, can all easily be managed’. Another passage, this one in verse, runs as follows:
This then is rāja-nīti for building an empire, composed by a rather introspective, yet by now quite experienced king, who has been on the throne for perhaps a decade. In certain key respects, the author departs from conventional wisdom. For example, he recommends posting Brahmins as commanders of forts, durga, and the fact that this was practical advice is shown by studies of the prosopography of the notables of the empire in that time.Footnote 52
Brahmins, in this view, have certain clear advantages over non-Brahmins, even though this caste is theoretically at least not to be associated with warrior functions (though numerous exceptions, both in the epics and earlier historical instances could be found):
Beyond this, however, lies the Brahmin's relative freedom from local attachments. At the same time, these Brahmins are clearly trained by now in military ways and engaged in worldly activities.
The potential for conflict between kings and ministers, that would be a staple of the histories and treatises produced by the karaṇams, the class from which the ministers themselves came, is also ever-present here, though its resolution is rather more to the king's advantage. The following extended passage makes this clear enough:
Early Modern Variations
The post-Krishnadevaraya period in Vijayanagara changes the context of such writings, in particular once we enter the period of the dominance of the Aravidu family. The growing role of Aravidu (‘Aliya’) Ramaraya's relatives and his extended family spread out in smaller kingdoms all over the Deccan already marks a significant shift in this respect. The nīti of the empire, articulated by Krishnadevaraya, again gives way to the nīti of small kingdoms, most of which survive with the help of kinship relations and support from the extended family. While this also creates the usual family intrigues, rivalries and battles, the new political conditions also give rise to opportunities for upward mobility. The emergence of the Nayakas from the flexible and uncertain political conditions in the post-Krishnadevaraya period is reflected in the nīti texts of this time.Footnote 53
The Āndhra Kāmandakamu by Jakkaraju Venkatakavi was written in 1584, and is of crucial interest to us in this context. Venkatakavi was employed in the court of Kondaraju Venkataraju, himself a small king from the Aravidu family. The personal history of this Venkataraju is interesting, especially because he is reputed to have renovated the Ahobilam temple, when it had been ruined by the Turks (turakalu). Even so, the nīti book Venkataraju has authored does not have any mention of Muslims, either disapproving or approving. What is instead noteworthy for us now is the regional and ‘secular’ (in the sense of non-sectarian) nature of nīti in the Āndhra Kāmandakamu. Even though the author states that his work is a translation of the earlier Sanskrit Kāmandakiya or Nītisāra, the later work in fact includes a number of nīti statements that are not to be found in the original, making it more an early modern nīti text rather than a simple restatement of a classical nīti vision. For instance here is a passage concerning the treatment of relatives and other political allies.
Sons of your maternal uncle and aunt, and your nephews and your maternal uncle himself, sons of your mother's sister's sons—these people are allies by blood (aurasa-mitrulu).
Your sons-in-law, brothers-in-law, your wife's brothers and sisters, are allies by marriage (sambandha-mitrulu). Kings of the lands on the other side of the country with which you share a border are allies from a related foreign land (deshakramāgatulu).
Kings who seek your protection in time of need are protected allies (rakshita-mitrulu).
A king should take note of these four kinds of allies and nurture their friendship.Footnote 54
We have already noted that the relationship between kings and their ministers had been a matter of concern for both Baddena and Krishnadevaraya, both of whom have some words of caution to the king regarding the choice of his ministers. Venkatakavi goes a step further and describes the corrupt practices that bad ministers could adopt in order to enrich themselves. The verse below gives several kinds of bribes a minister could take:
If the minister comes to a festival, what he gets is called kānuka. What he receives by way of things he appropriates from people is called porabaḍi.
If he gets kickback in cash it is called paṭṭubaḍi. The money he gets privately in return for taking care of their business is called lañcam.
A king should make sure that his minister does not take any of the above, and such a person should work for the king and receive his livelihood only from the king.Footnote 55
Nīti and Karaṇam Culture
The political landscape we have described changes again from the seventeenth century onwards. A new group of people who made writing their profession emerged as a politically and culturally important group. In Andhra, Karnataka and Orissa, these people were often called karaṇams, and they were considered to be the counterparts of the munshīs in northern India.Footnote 56 Often seeing themselves as mantris or ministers of kings, the karaṇams perceive themselves broadly as managers of public affairs. Most members of this group were not connected with major empires or powerful kings, but they nevertheless had an enormous influence in running small kingdoms, zamīndārīs and petty principalities. They were also successful managers of properties, accountants, poets and historians. They prided themselves in their multiple language skills, their ability to read scripts of many languages and above all their skill at calligraphy. They were also at the same time accomplished at writing a highly unintelligible cursive script, which could be read only by other karaṇams. They came mostly from Brahmin castes, and in Andhra they were mostly Niyogi Brahmins—as opposed to the Vaidikis.Footnote 57 The former managed public affairs while the latter specialized in ritual texts and ritual performances, even though both wrote poetry. Karaṇams used the pen for their power and prestige. They were writers in the true sense of the word as we understand it today.
The self-image of the karaṇams is fascinating. They have left behind a large body of writings about themselves, their code of conduct and training, in addition to a number of historical texts. Here is what some of the verses tell us about a karaṇam:
Included in a list of 32 legendary ministers is a certain Rayani Bhaskarudu, who appears most frequently in manuscript sources. Here are a few poems about him from tradition.Footnote 58
We also find a verse concerning a minister inscribed on the front gate of the Gopinathasvami temple in Kondavidu.
We can see that a number of developments led to the growing importance of karaṇams in the affairs of the state. The increased use of Persian as a language of administration, and the presence of multiple languages in which smaller kings had to correspond with their political allies and neighbours, the availability of pen and paper, and the elaborate new accounting responsibilities made the position of scribes far more important in society than what it had been before. Now scribes were employed in jobs of higher status and power than simply serving as persons who could take down dictation or copy manuscripts. Reality is now what was written down, and not as earlier, what was uttered. We can see a corresponding change even in the popular mythology and Hindu iconography. The goddess of language and arts, Sarasvati was now endowed with a book in one of her hands, in addition to a vīṇa, the stringed musical instrument. Yama, the god of death, acquires an assistant Citragupta, who keeps accounts of living beings in separate files, and as in the poem that was quoted earlier, can even become more powerful than Yama himself.
The people who called themselves ministers (mantris) were not always ministers of a ruling king. Mantri was in a sense more an honorific caste title rather than a fixed position or office. Often these ‘ministers’ were themselves independent chiefs of a locality or even a village. However, in keeping with the convention that a king should be a warrior, the minister who has taken independent control of an area, also describes himself in military terms. But by seventeenth century there was a significant shift in the values of peninsular Indian society. Greater importance was given to dāna, charity, rather than vīra, valour in battle. The possibility of acquiring wealth in the form of cash created conditions of upward mobility, that were different from those created by simple military conquest. The emergence of the left-hand caste Balijas as trader–warrior–kings as evidenced in the Nayaka period is a consequence of such conditions of new wealth. This produces a collapsing of two varṇas, Kshatriya and Vaishya, into one. Acquired wealth, rather than status by birth in a family now leads to an entirely new value system where money talks. The Sumati shatakamu records this change rather cynically:
The presence of cash also generates charity. Members of the nobility are now constantly advised to excel in charity. In keeping with the changes in the social values, nīti is no longer regarded as a matter that simply concerns kings and courtiers. It is for everyone, and in particular for anyone who desires status and social recognition. Nīti is now told in the form of stories rather than aphorisms and shāstric statements. Kuciraju Errana's Sakala-nīti-kathā-nidhānam, a book of stories that teaches nīti, indicates an early recognition of this change.Footnote 60 Errana adopts a number of stories from Betāla-pañca-vimshati and other kathā sources, both from Sanskrit and from Telugu. The main thrust of the stories is to teach the individual wise and tactful ways of handling oneself, and thus maximizing one's chances for success.
One book that codifies the conduct of karaṇams is the Sumati shatakamu.Footnote 61 Written by an unknown author probably in the eighteenth century, this book is variously attributed to Baddena and to an even more ancient Bhimana. Perhaps both authorships were ascribed by karaṇams to make the text serve two different purpose. Baddena's authorship serves the interests of the karaṇams in claiming political legitimacy among kings and other aspirants to rule an area, and the Bhimana authorship makes the text speak with a voice of the authority of an ancient, god-like poet to serve the interests of the same community when they desire legitimacy among the people in general. The Sumati shatakamu elevates the role of the minister (karaṇamu) and treats it is more crucial for the maintenance of the order of the kingdom than that of the king himself.
It also gives practical wisdom for ordinary people such as the following:
Despite such practical advice, the Sumati shatakamu is at bottom a cynical (rather than simply an amoral) text, which believes women are not trustworthy, that kings never keep their word, and friends last only as long as you have money. In the hard world it depicts, you have to take care of yourself—no one else helps.
Conclusion
When the British government and its native employees wanted ‘morals’ to be taught in the early-nineteenth century, the Telugu equivalent that their pundit informants could find was nīti. This was based on a rather curious misunderstanding: for even if there are some ethical teachings and moral statements in these texts, they are not exactly the kind of moral code that one would apply to all people. Vennelakanti Subbarao (1784–1839), translator for the Sadr ‘adālat of the Madras Presidency, a Telugu Niyogi Brahmin who rose to the highest post a native could aspire to in the East India Company administration at the time, and who commanded competence in about half a dozen Indian languages in addition to English, was one of the more prominent of the Company's interlocutors already from an early time. When he was appointed member of the Madras School Book Society, he submitted a report in 1820 on the state of teaching in schools, in which he wrote that children in schools were taught neither adequate grammar nor morals. So they came out of their schools with no real ability in using the language and they were not trained to become upright members of their society either. Therefore, he recommended—addressing the need for teaching morals—that ‘tales extracted from different books composed chiefly of morals written in modern languages’ be prescribed for study.Footnote 62
In this context, Ravipati Gurumurti Sastri also put the Pañcatantra stories into Telugu prose and taught them at the College of Fort St. George in Madras. This was soon followed by another translation of Pañcatantra by the very influential Paravastu Cinnaya Suri.Footnote 63 Now the Pañcatantra was not in fact a ‘Book of Morals’; rather, it was statecraft taught by means of animal fables. When the first generation of colonial schoolboys needed a textbook, Puduri Sitarama Sastri, a pundit in Madras wrote a text called Pědda Bāla Siksha (The Big Book of Lessons for Children), which was published in 1847. This work contains a number of items such as basic arithmetic, the names of the weekdays, months, and years according to the traditional lunar calendar, and many items of conventional wisdom, a few stories, and aphorisms modelled after the statements from nīti texts, to teach ‘nīti’ (now translated in an unproblematized way as ‘morals’) to schoolboys. To be sure, in every nīti text, there were occasional statements that looked like teachings of virtue, which were carefully selected and included in school textbooks. Verses from Bhartrahari, which were translated into Telugu by several poets during the medieval period, came in handy. Even the Sumati shatakamu, which, as we have seen, is actually a handbook for karaṇams, yielded some nice and acceptable moral statements.Footnote 64 Because of the simple language in which the Sumati shatakamu was written, it came to be particularly popular in school moral curricula. Soon enough, lines from these verses came to adorn classroom walls and copybooks. Thus, in the end, books on statecraft and worldly wisdom could serve as acceptable substitutes for the Ten Commandments.
Our central purpose in this essay has been to widen the rather narrow conception within which ‘political thought’ has hitherto been studied in an Indian context. We would only caricature very slightly if we were to say that the usual strategies espoused by analysts are two: either they assume that modern politics in India was a pure product of the interaction with colonialism and colonial modernity, or at best, they leap over the intervening centuries to classical India and its materials. In this context, we welcome the development of interest in recent times in the Indo-Persian corpus, and what it might tell us about both institutional arrangements and political thought at the time of the Sultanate and the Mughals. The problem does remain however of that part of India where Persian was not the principal language in which such thought was expressed. The example of the Maratha polity in the seventeenth and eighteenth centuries brings this home, even though the Marathi used by them was heavily inflected by Persian. It is clear from the researches of Hiroyuki Kotani and Narendra Wagle, however, that the eighteenth-century Maratha Deccan continued to witness a struggle between precisely the forces we have set out in this essay, that is between the proponents of nīti on the one hand (who no doubt drew on the Indo-Persian corpus as well) and those who remained fiercely attached to the highly dharmashāstra-oriented vision of social ordering and political functioning. The continued presence of terms such as dosha and prāyascitta in the vocabulary of the Maratha polity possibly testify to the waning influence of the nīti tradition in that system.Footnote 65
A celebrated reflection on the ‘history of concepts’ written some 35 years ago proposed to historians of Europe that they needed to go beyond their preoccupation with social (and political) history to look at both individual concepts, and groups of concepts, to clarify that which underlay the functioning of the political and social systems in the societies they studied. In that context, Reinhart Koselleck wrote:
‘The relationship between the history of concepts (Begriffsgeschichte) and social history (Sozialgeschichte) appears at first sight to be very loose, or at least difficult to determine, because the first of these disciplines primarily uses texts and words whereas the latter only uses texts to deduce facts and movements which are not contained in the texts themselves. It is thus that social history analyses social movements and constitutional structures, the relations between groups, social strata and classes; beyond the complex of events, it tries to come to terms with medium or long-term structures and their changes (. . .). The methods of the history of concepts are very different’.Footnote 66
We would hardly wish to be so immodest as to claim to be introducing the history of concepts (Begriffsgeschichte) into the study of the Indian past. However, in this collaboration between a historian and analyst of literature, and a social and economic historian, we hope to have opened a window into a neglected, and yet highly significant, corpus.
We began this essay with two quotations, one by a scholar and editor of forgotten texts, lamenting the loss of importance suffered by the nīti tradition, the other by one of the most important political figures in twentieth-century India. B.R. Ambedkar was, we are aware, a keen student of the Indian past, and had even studied with R.P. Kangle, an authority on the Arthashāstra. The remark by him that we quote refers precisely to the tension between the nīti- and dharma-oriented traditions that have lain at the heart of this essay. He glossed these respectively as ‘secular law’ and ‘ecclesiastical law’, and there are many—especially among the growing number of ‘anti-secularist’ intellectuals in India—who would immediately object to these translations.Footnote 67 But perhaps Ambedkar was not so wrong after all in his use of the term ‘secular’ (however problematic the word ‘ecclesiastical’ might be). Not as cavalier in his disregard of the Indian past—or dismissive of history—as writers such as T.N. Madan and Ashis Nandy have usually been, it may well be that his view of a struggle between different conceptions of political and social arrangements in pre-colonial India might shed light on the deeper roots and more profound purchase that ‘Indian secularism’ has, than that of a mere transplant from distant climes. To explore that line of inquiry would take us, however, beyond the confines of this essay.
