Text of a guest lecture by writer Tony Joseph at the 84th annual session of the Indian History Congress at Government Brennen College in Thalassery, Kerala.
Thank you to the Indian History Congress and its office-bearers for inviting me to this annual session for a lecture dedicated to the memory Prof. MGS Narayanan, a historian I greatly admire. Though I never had an opportunity to meet him, his works have lighted the path for me on many occasions. There is no honour I have received that I cherish more.
As I stand before you, I am only too keenly aware of the wide gulf between my own pond of understanding about Indian history and the ocean of knowledge that you carry as academic historians. I am also aware of the perils of tackling a question of Indian history that has been analysed and addressed by the very best minds in the discipline from this very platform and outside over many decades. Therefore, I will not attempt to reinvent the wheel; instead, I will limit myself to using the wheels that have already been invented in a hopefully new and productive manner. Needless to say, if there’s any merit in this paper, all credit should go to those who created the wheels I will be using today. The blame, if any, will rest with me.
Much of what I will be talking about will form part of my next book which will be a kind of sequel to my first book, ‘Early Indians’, and will address the cultural formation of India – as opposed to the genetic formation of India that ‘Early Indians’ focused on. One of the key lessons I have learnt over the last 14 years as a student of history is that one cannot visualize the hen or the egg by parsing the omelette, no matter how hard one tries. Therefore, this paper on the evolution of caste will not refer to the omelette, or caste as it exists today. It will look at caste formation only between the first millennium BCE and the late first millennium CE. For the same reason, it will also not make a hard distinction between varna and jati as modern sociologists often do, because in the ancient texts of the period that we are dealing with, there is no indication that the two were treated as different systems (1). Jatis were part and parcel of the varna-asrama order that was conceptualized in the centuries around the beginning of the Common Era, as seen from the Dharmasutras and later Dharmasastras. This does not mean that jatis and varna had the same origins, but only that they became integrated as part of one system at an early period, with the Manava Dharmasastra even providing a theoretical if improbable reasoning for the existence of jatis: mixing between different varnas in different permutations and combinations.
So to begin at the beginning, as anyone who has studied Caste knows, the earliest mention of a four-tiered hierarchy – brahmana, rajanya, vaisya and shudra - is in the Rigveda, in the Purusha Sukta of Book 10 (X.90). But what is more important than this clear articulation is the academic consensus that this book is among the later parts of the Rigveda and that even within that book, this hymn is a later addition.
This has implications for us in terms of understanding to what extent the Rigvedic period overlaps with that of the earliest conceptualization of a four-tiered caste system, and the answer is: little to none because there is no mention of the shudra at all in the Rigveda apart from that in the Purusha Sukta. What about a three-tiered caste system, then? Even there the answer has to be, very little because there are only rare references to even ‘brahmana’ as a social category. Rajan appears often, but that refers to the king, and the term ‘rajanya’ which is equivalent to ‘ksatriya’ of the later periods is used only once, in the Purusha Sukta. The term ‘vis’ appears more often, but then ‘vis’ is a term that is used to refer to the people in general of the Arya community. Also, that occupations are not necessarily hereditary is suggested by the Rigvedic hymn IX.112.3, for example, which has a composer who says that his father is a ‘healer’ while his mother ‘pushes a pestle’.
On the other hand, a formal contrast is indeed made between priestly and kingly powers in one of the later books of the Rigveda, Mandala 8, hymn 35, verses 16 to 18. To quote Stephanie W. Jamison and Joel P. Brereton in their translation of the Rigveda: “As a number of others have pointed out, these three verses, asking the Aśvins to “quicken” or “animate” various things for us, are appropriate to the three varṇas in turn: verse 16 to the brahmin, 17 to the kṣatriya, and 18 to the vaiśya.”
Verse 16 starts with the word ‘Brahma’ and then requests the Asvins to: “Quicken our sacred formulation and quicken our insights”; verse 17 starts with the word ‘ksatram’ and says: “Quicken our lordly power and quicken our men”; verse 18 does NOT start with ‘vis’, it but it does go on to say: “Quicken our cows and quicken our clans”. The hymns that follow, 36 and 37, also seem to make similar allusions to priestly and kingly powers.
But these are small and ambiguous pickings for anyone trying to find hard evidence for a proper varna-jati system in the Rigveda and, therefore, we can perhaps agree with Jamison and Brereton when they say that the “system seems to be embryonic.”
But that raises the question, how did that embryo come to be? The studies that provide the keenest insights into this question are by Michael Witzel and Thennilapuram Mahadevan. In his paper ‘The Realm of the Kuru: The Origins and Development of the First Indian State’ based on an extensive analysis of the Rigveda, Witzel tackles the question of how the earliest Veda came to be put together. In ‘The Rsi Index of the Vedic Anukramani System and Pravara Lists: Toward a Prehistory of the Brahmans’, Thennilapuram Mahadevan builds on that paper to show how the first caste, that of brahmanas, came about as one consequence. Mahadevan relies on the Vedic lists of the composers of the hymns and their genealogies.
To recount Witzel’s arguments briefly, the tribe of the Kurus, a subtribe of the Bharatas, became predominant after the Battle of the Ten Kings or Dasrajna that is mentioned in the Rigveda. After the battle, the “geographical centre of the Vedic civilization” moved eastwards from Punjab to Kurukshetra, the land lying between rivers Sarasvati and Drishadvati, about 175 kilometres northwest of New Delhi.
As they consolidated their power, the newly-emergent Kurus felt the need for a unified canon, drawing on the sacrificial hymns of their own tribe as well as those of the peoples they had defeated. Many things support this contention. For one, the bulk of the Rigveda, according to Witzel, provides a snapshot view of only five or six generations. For another, it is focused more on the Puru and Bharata tribes though there would have been 50-odd Arya tribes roaming the region. From both these facts, we can surmise that he Rigveda that we have was carefully collected and arranged in a certain manner at some point of time from a larger body of hymns. The fact that the very last hymn of the Rigveda is about unity also supports this argument. This is what the hymn X.191 says: “Come together, speak together; together let your thoughts agree, just as the gods of long ago, coming to an agreement together, reverently approach their sacrificial portion...
Common is your purpose; common your hearts; let your thought be common, so that it will go well for you together.” (2)
Until this happened, around 1000 BCE, writes Witzel: “…the Ṛgvedic hymns were the composition and property of individual priests belonging to a number of “families” or clans. They regarded their poems as personal or as clan property and marked them, either with (parts of) their own individual name, or with certain refrains that indicated their clan origin: a Ṛgvedic version of the copyright notice.”
When a common canon was made out of these collections, writes Mahadevan, the families that were once in charge of their own hymnal traditions did not die out, but became the backbone of a ‘pan-Vedic agency to sing a pan-Vedic corpus. (Remember, the Rigveda collection is entirely oral; there is no writing at this time). They were now a kind of ‘oral agency’ carrying forward a common tradition, each of them equal to the other, but now ‘bound into a biological body’ through new rules regulating marriages among them.
As Mahadevan tells it, each of them, who had a collection of family songs and a First Singer who started their tradition, was now called a Gotra (original meaning: cow pen) which was named after their First Singer. And the new rules were that marriages must not occur within the same gotra (exogamy), but must occur within the 50-odd Gotra groups (endogamy), thus creating ‘One, Out of Many’ (E pluribus unum), or as Mahadevan puts it, “the ‘caste’ of brahmans”.
Mahadevan cites instance after instance from the Rigveda to make his case that until this happened, the ranks of the Vedic poets were open to all, even those who had a non-Arya past. He mentions Rigvedic composers Kavasa Ailusa and Kanva, for example, whom later texts describe as having been ostracized for being ‘abrahmana’ and ‘dasiputra’. But this changed once the canon was fixed. The ‘oral agency’ of the Arya was no longer doing open hiring. it was closed for outsiders, just as no more Rigvedic hymns were being written, because some of the existing ones had been canonized. The brahmanas are thus the first and perhaps the only real ‘varna/caste’ to be formed at a particular time and place, with enormous implications for the future.
The ksatriya or the warrior/ruler caste, at this time and in the future as well, will be mostly decided de facto, rather than de jure: those who manage to get and keep power are ksatriyas. And as long as they wield power, the priests will often find a way to turn them into ‘ksatriyas’. Or as historian D.D. Kosambi wrote in a letter to Pierre Vidal-Naquet on July 4, 1964: “Don’t be misled by the Indian ksatriya caste, which was oftener than not a Brahmanical fiction.” (3) He also describes how several southern kings of tribal origin performed the Hiranyagarbha or ‘Golden Womb’ ceremony to be reborn, in a manner of speaking, with a new caste, usually that of ksatriya. (4)
While the ritual itself may be an invention of the Puranic period as Kosambi suggests, there is no reason to consider the thinking behind it as new. In the Anusasanaparva, or Book of Instructions, of the Mahabharata, the supreme commander of the Kaurava forces and a patriarch of the Kuru dynasty, Bhishma, tells the eldest of the Pandavas, King Yuddhisthira, this: (13.33.19): “Those various men of ksatriya birth – Sakas, Yavanas and Kambojas – have reached the level of shudras because no brahmins are seen among them.” Manusmriti says the same thing using almost the same words in 10.43-44.
That even foreigners such as the Yavanas (Greeks) and the Sinas (Chinese) who have nothing to do with the Brahmanical society are called men of ‘ksatriya birth’ can only mean that it is their de facto power or sovereignty that made them ‘men of ksatriya birth’ in Brahmanical eyes. But despite such birth, they had fallen to the status of the shudras because they have not accepted, or have not had the opportunity to accept, the social and ritual hierarchy that put the brahmins at the top. So when a tribal chieftain undergoes a Hiranyagarbha ceremony thus accepting the Brahmanical hierarchy, he is indeed reborn as a ‘ksatriya’ according to the definitions used in Anusanaparva and Manava Dharmasastra.
Those who do not fall into either of these categories – who are neither part of the ruling families, nor a card-carrying member of the newly-formed priestly varna of brahmins – are considered vis or vaisya or common folk – a residual category of the Arya community that did not need a special definition in any way in the early period.
In the centuries that followed, those who were outside the Arya culture and religion but who served the Arya families as domestic workers or, once agriculture took off, as farm labourers, had to be accommodated within the system and that was done by giving them a low status as ‘shudra’. Those who were outside of all four categories, such as the tribes who lived in the forests or on the margins of settled society, were regarded as outcastes, like the chandala. But this fully fleshed-out varna-jati system will take time to develop, and will be dependent on the speed with which agriculture and, therefore, the need to engage non-family labour, took off.
Many things follow from this. One, the caste system did not arise out of a purity-pollution cline; nowhere in the Rigveda or the other Samhitas is it suggested that there is a hierarchy of purity among the brahmanas, rajanya and the vis. Two, it did not arise out of differences in eating habits; the idea of vegetarianism originated with the sramana traditions in Greater Magadha many centuries later. Three, it had little to do with race or ethnicity to begin with; the three varnas were all considered Arya, though some of the prominent priests could have had an un-Arya past, as we saw. Four, the idea of a varna system did not spread from a pre-Arya or Harappan Civilization because the Rigveda, composed centuries after the Steppe migrations began, has no clear enunciation of the varna-jati system. And five, for the same reason, it was not brought to India by the Steppe pastoralists. It was made in India in the true historical sense. With hindsight, it is possible to recognize that ideas such as purity and pollution and dietary habits were justificatory accruals to the varna system that occurred centuries later.
The immediate trigger for the beginning of the caste system was the perceived need of a newly victorious kingdom to have a unified religion that will encompass all the tribes that it defeated, along with a priesthood that will administer it. It was political contingency that gave rise to the first two castes, that of the priest and then that of the king, and these two are who will guide the evolution of the varna-jati system in line with future political contingencies and economic opportunities as we will see.
The best model, therefore, to fit this evidence is that of a universal ‘Homo Opportunisticus’ -- not an Indian ‘Homo Hierarchicus’ which, according to sociologist Louis Dumont’s analytic model in his book of the same name, represents a cultural predilection for hierarchy. Given the opportunity, political and economic man will create a system of belief and along with it, a consecrated social system, to perpetuate his status and power. The practice of singing hymns at sacrificial rituals and the necessity to continue this tradition in a unified manner at a time when there was no literacy was what created the special initial situation for the first caste to arise in ancient India – and it had consequences that bring to mind the butterfly effect of chaos theory, where small differences in initial conditions can have large consequences.
Brahadranyaka Upanisad, written centuries later, perhaps refers to this special relationship of mutual dependence between the ruling power and priestly power at 1.4.11: “Hence there is nothing higher than the ruling power. Accordingly, at a royal anointing, a brahmin pays homage to a ksatriya by prostrating himself. He extends his honour only to the ruling power.” But then it adds: “Now, the priestly power (brahman) is the womb of the ruling power… it is to priestly power that he (the ksatriya) returns in the end as to his own womb. So one who hurts the latter harms his own womb and becomes so much worse for harming someone better than him.” (5)
When we see the sequence of events this way, perhaps the most crucial characteristic of the caste, or varna-jati system, becomes self-evident: what lies at its heart is the alliance between ruling power and priestly power – that is how it began and that is how it will find itself room to grow.
But as long as the Arya society remained pastoral, the practical consequences of this alliance could not have been large because there would not have been a sufficient surplus of production for anyone to appropriate and, therefore, no opportunity for large class differences to arise. But once the tribes began to settle down to take up agriculture along with cattle-rearing - which happened in the first half of the first millennium CE if the archaeological discoveries of iron agricultural implements are any indication - the situation changed. There was more surplus and this began to be appropriated increasingly through force, rather than through voluntary contributions or bali, as earlier.
But myths, as we know from world history, can sanctify power and thus reduce the need for use of force. Therefore, by the time of the composition of the Brahmana texts, we see a significant effort to elevate the concept of the varna system from the world of mortals to the universe itself. No more will we find a scarcity of references to it. The Jaiminiya Brahmana (1.68-69) is a good example of the way gods, animals, hymns, poetic meters and chants are all mapped into the four-fold varna system. To quote: “Prajapati, in the beginning was this (all)… He desired: ‘May I become many; may I reproduce myself; may I become a multitude.’ He emitted from his head, his mouth, the nine-versed hymn of praise, the gayatri meter, the rathantra chant, Agni among the gods, the brahman among men, the goat among the animals…”
In a similar manner, the text then goes on to say how Prajapati emitted from his arms and from his chest, the tristubh meter, Indra among the gods, the ksatriya among men and the horse among the animals. From Prajapati’s belly came the jagati meter, Visva Devas among the gods, the vaisya among men, and the cow among animals. And from his feet come the anustubh meter, the shudra among men, the sheep among animals, but not a single one among the gods. The Taittiriya Samhita adds in 7.1.1.4-6 that ‘the shudra is unfit for the sacrifice, for he was emitted along with no gods.’ (6)
Similar versions of how the universe began, already categorized into varnas, are repeated in text after text, from the Pancavimsa Brahmana to the Satapatha Brahmana to the Taittiriya Brahmana. The Pancavimsa Brahmana version adds seasons to the list of things that are categorised, with spring being the brahmin among seasons; summer being the ksatriya; and the rainy season the vaisya. There is no season mentioned as being shudra.
In the Satapatha Brahmana, the varna nature of seasons is mapped into how the various varnas should organize their rituals, with the spring being the season for a brahmin to set up his fires, the summer for the ksatriya and the rainy season for the vaisya. There is no season mentioned for the shudra to set up his fires. (Satapatha Brahmana 2.1.3.5).
The division of the whole universe into the four classes is so striking that Prof. Brian K Smith wrote a book on it, titled ‘Classifying the Universe’. In his words, “Varna… was a classificatory system which attempted to encompass within it all of the major sectors of the visible and invisible universe.” The end result of these universal classification schemes, writes Smith, “was that certain humans could present what was an arbitrary social status or status claim as natural and sacred; that is, social hierarchy was presented as inexorably part of the immutable and divinely given order of things.”
Smith argues that this conceptual categorization of the universe in varna terms was an important reason for the staying power of the idea through centuries and quotes art historian E.H. Gombrich to support his argument: “All thinking is sorting, classification.” He also quotes the Hungarian American academic and psychologist Thomas Szasz as saying: “In the animal kingdom, the rule is eat or be eaten; in the human kingdom, define or be defined.”
Side by the side with the universalisation of varnas, there was also a sharper articulation of the ideological core of the system, the brahmana-ksatriya alliance. Satapatha Brahmana (6.4.4.13) which deals with the meaning of sacrificial rituals provides an exemplar. In the quotation that follows, bear in mind that the horse stands for the ksatriya; the goat (the usual sacrificial animal) for the brahmana; and the ass, for the shudra. This is what the text says: “And inasmuch as, in going from here, the horse goes first, therefore the ksatriya, going first, is followed by the three other castes; and inasmuch as, in returning from there, the he-goat goes first, therefore the brahmana, going first, is followed by the three other castes. And inasmuch as the ass does not go first, either in going from here, or in coming back from there, therefore the brahmana and ksatriya never go behind the vaisya and shudra, hence they walk thus in order to avoid a confusion between good and bad. And, moreover, he thus encloses those two castes (the vaisya and shudra) on both sides by the priesthood and the nobility, and makes them submissive.” (Translated by Julius Eggeling.)
In other words, in the way the composers of the Satapatha Brahmana understood it, the ksatriya and the brahmana together carry the burden of keeping the vaisya and the shudra submissive and the rituals both reflect that necessity and ensure their obedience. The intensity of the brahmana-ksatriya alliance is brought out even more powerfully by another passage from the Aitareya Brahmana which was commented upon by Ananda K. Coomaraswamy, in his book, ‘Spiritual Authority and Temporal Power’. Coomaraswamy notes that the formula uttered by the groom to the bride during the marriage function is also uttered by the priest to the king during a royal ritual. The formula begins with: ‘I am That, thou art this: I am Sky thou art Earth’ and Coomaraswamy writes as the first sentence of his first chapter: “It may be said that the whole of Indian political theory is implied and subsumed in the words of the marriage formula… addressed by the brahmin priest, the purohita, to the king, in Aitareya Brahmana VIII.27.” According to Coomaraswamy, the king is unquestionably the ‘feminine’ party in the ‘marriage’ of the ‘Sacerdotium and the Regnum’ or the brahma and the ksatra. The fact that the Aitareya Brahmana visualizes the priest as masculine and the king as feminine is perhaps less important than the fact that it visualizes the priest-ruler relationship as intimate as a marriage.
By the middle of the first millennium BCE, therefore, the core idea of a four-fold division of society, with a strong, explicitly-defined alliance between the ruling power and the priestly power to keep the rest of the society in order was well-established at least in theory. But it is important to ask: well-established where, in which part of the subcontinent? This is important to check for ourselves whether the common assumption that the varna system was all-pervasive early on is correct or not.
As we have seen, the unification of many tribes under the Bharata-Kurus and the gradual emergence of the four-varna hierarchy all happened in what was known as Kuruksetra. This region would later come to be subsumed under the term Aryavarta in the Dharmasutras and other texts in which Aryavarta was defined as being the region between the Himalayas and the Vindhyas, and to the east of where river Saraswati disappears and to the west of the Kalaka forest, which is supposed to have been at the confluence of the Ganga and Yamuna. (7) This was where Brahmanism and along with it, the varna hierarchy took wing.
The Satapatha Brahmana describes an expansion of the Kuru-Pancala varna orthodoxy to areas to the east of Aryvarta. The well-known verse 1:4:1:14 of this text is about King Videgha Mathava and his priest Gotama Rahugana following ‘Agni’ towards the east, from the banks of the river Sarasvati. It makes it clear that the Brahmanization of Videha happened later than that of Kosala, but now that it has been ‘tasted by Agni’, there are many brahmanas there too and the land is very cultivated.
The Kings of Videha carrying the title Janaka appear prominently in later Vedic texts, and one of the most important sages of the Upanisads, Yajnvalkya, was the purohita (priest) in Janaka’s court.
But this situation did not hold for long. Videha ceased to exist as a separate kingdom or entity even before 500 BCE. It became part of the Vajjika League, a confederacy of gana rajyas or oligarchies under the leadership of the Licchavis. Kosala continued to be a monarchy, but we know from the Buddhist texts that its king, Pasenadi, was very close to the Buddha and the Sangha. “The whole of the third Samyutta [of the Samyutta Nikaya], consisting of 25 anecdotes, each with a moral bias, is devoted to the king of Kosala, and there are about an equal number of references to him in other parts of the literature,” writes Narendra K. Wagle in ‘Society at the Time of the Buddha’, 1966, Page 38, adding: “…Kosala, with its capital and king, were favourite topics among Buddhist writers.” Buddha preached 871 suttas in Kosalan territory, 844 of them in Jetavana, located just outside its capital, Savatthi. (8)
What all of these suggest is that outside the heartland of Aryavarta, the Kuru-Pancala Vedic orthopraxy did not gain unchallenged dominance. The quick victory it had in the east, as reflected in the Satapatha Brahmana, was either shallow or short-lived and the conditions were not yet ripe for it to expand phenomenally. What was true of Kosala was even more true of Magadha, where dynasty after dynasty – Haryanka, Shishunaga, Nanda, Maurya – were all major patrons of the Sramana religions of Jainism, Buddhism and Ajivikism.
If the Vedic orthopraxy of Kuru-Panchala found the going tough in the kingdoms of the east in this period, it found it even tougher in the ‘ganas’ or ‘sanghas’ - or oligarchies of clans that saw themselves as khattiya or ksatriya, like the Buddha’s Sakya community. The Vedic orthopraxy that had arisen in the context of a monarchy – that of the Kurus - did not fit so well with the ganas where the Khattiyas were loath to give brahmins the standing they expected and demanded. It is easy to see why.
In the gana oligarchies, the land was the common property of the entire clan and the division between the ksatriya and vis did not exist, as in the Kuru-Pancala kingdoms. The Buddhist text Ekapanna Jataka, for example, says there were 7,707 kings at Vaisali, the capital of the Vrjji gana, while the Mahavagga VIII.1.1. says there were 7,707 palaces in the city. In her book, ‘Social Dimensions of Early Buddhism,’ P 15, Uma Chakravarty says that gana rajyas “reveal a comparatively simple social organization consisting preponderantly of one or more Khattiya clans and their slaves and workmen. Some artisans are mentioned such as the barber and the metal- smith, but apart from these there are hardly any references to other social groups, and even those to brahmanas are not very numerous.”
It should cause no surprise that the need or case for a close ksatriya-brahmana alliance gets weaker when there is no ‘vis’ or vaisya for it to act against. The khattiyas of ganas “have little use for prestations and the rituals are marginal,” says Romila Thapar in ‘From Lineage to State’, 1984. This means the brahmana-ksatriya alliance, the heart of the Vedic orthopraxy, found little purchase in the ganas which were prominent particularly in the foothills of the Himalayas.
What this also means is that the common assumption that by the middle of the first millennium BCE or so, the varna hierarchy was ubiquitous in all of northern India is unsupported. Evidence from both Vedic texts and Sramanic texts suggest that this was not so. For example, in the Ambattha Sutta of the Buddhist text Digha Nikaya, the brahmin student Ambattha is described as being so discomfited by the poor welcome he once received in the Sakyan capital of Kapilavatthu that he nourished a grudge for long, and when he got to meet the Buddha later, he could not contain his anger and called the monks assembled with the Buddha, “shavelings, fake ascetics, primitives and black spawn from the feet our kinsmen”. Shavelings is a reference to the shaved heads of the monks, and the black spawn from the feet of our kinsmen is probably a reference to the Purusha Sukta myth about the origin of shudras.
Ambattha describes the source of his anger this way: “This one time, Master Gotama, I went to Kapilavatthu on some business for my teacher, the brahmin Pokkharasadi. I approached the Sakyans in their town hall. Now at that time several Sakyans and Sakyan princes were sitting on high seats, poking each other with their fingers, giggling and playing together. In fact, they even presumed to giggle at me, and didn’t invite me to a seat. It is neither proper nor appropriate that the Sakyans – primitives that they are – don’t honour, respect, revere, worship or venerate brahmins.”
Then he goes on to say: “Master Gotama, there are these four classes – ksatriyas, brahmins, vaisyas and shudras. Three of these classes – ksatriyas, vaisyas and shudras – in fact succeed only when serving the brahmins. It is neither proper nor appropriate that the Sakyans – primitives that they are – don’t honour, respect, revere, worship, or venerate brahmins.” (9)
The disdain and contempt that Aryavarta had for what can be called the Greater Magadha region to its east – a term coined by the late Prof. Johannes Bronkhorst in his seminal 2007 book ‘Greater Magadha: Studies in the Culture of Early India’ - also comes out from a passage in the Satapatha Brahmana.
Satapatha Brahmana 13.8.15 speaks about the “demonic people of the east”, as quoted by Bronkhorst in ‘Greater Magadha’, P 4. Writes Bronkhorst: “These demonic people from the east, we learn, were in the habit of constructing sepulchral mounds that were round. These round sepulchral mounds are contrasted with those in use among the followers of the Satapatha Brahmana,” which seem to have been four-cornered.
Bronkhorst also quotes the ‘Vedic Index of Names and Subjects’ by Macdonell and Keith as saying this about Magadha and its inhabitants: “Magadha is the name of a people who appear throughout Vedic literature as of little repute. Though the name is not actually found in the Rigveda, it occurs in the Atharvaveda (5.22.14), where fever is wished away to the Gandharis and Majavants, northern peoples, and to the Angas and Magadhas, peoples of the east.”
This cultural contrast between the two regions precedes the second urbanisation and the rise of the Sramanic religions in Greater Magadha. But it becomes far sharper in the second half of the first millennium BCE when wet rice cultivation and use of iron implements lead to dramatic productivity improvements in Greater Magadha, leading to the rise of towns, cities, kingdoms, empires and, consequently, new values and religions that had little to do with sacrificial rituals, hereditary priesthood and the concept of varna.
The leading belief systems of Greater Magadha, especially Jainism and Buddhism, were built on the ideas of karma and rebirth and finding the right way out of the cycle of rebirths. These concepts were foreign to Vedic samhitas and when they finally make an appearance in the Upanisads, they are repeatedly put forward as ideas that were unknown to brahmins earlier, although known to ksatriyas. (10) It is pertinent that both the Buddha and the Mahavira considered themselves ksatriyas or khattiyas, like the ganas they belonged to.
What follows then are centuries of opposition between Aryavarta and Greater Magadha, between the Brahmanic conception of society and the Sramanic conception of it, or quite simply, between the brahmanas and the sramanas. Nothing brings out the intensity of the conflict as much as grammarian Patanjali in his Mahabhasya written around 150 BCE using ‘sramanabrahmanam’ (11) as an example of an oppositional compound or ‘virodha-dvandva’ which combines two elements “between whom there is eternal conflict.” That the sramanas were against animal sacrifices, rejected the authority of the Vedas, advocated renunciation as a way to escape the cycle of rebirths and admitted into their monastic organisations people of all backgrounds without regard for varna, were all reasons for conflict.
The Jaina text Uttaradhyana, for instance, has the tale of monk Harikesiya who was given to great austerities and was a chandala, an untouchable, before becoming a monk. When he was seeking alms, he came to the enclosure of a Brahmanical sacrifice and the brahmans, recognizing an untouchable at the holy place, reviled the sage saying, “Better this food and drink should rot, than that we should give it to you.” This Jain text describes the brahmin sacrificers as “killers of animals, rigid with the pride of birth.” (12)
These instances can be multiplied manifold, but it is enough to say that the conflict between the brahmanas and sramanas reached its peak at the time of emperor Asoka Maurya, a self-declared upasaka or lay follower of Buddhism, whose reign also saw the Sramanic religions, especially Buddhism, making huge inroads into the heartland of Brahmanism, Aryavarta.
The most serious challenge Asoka posed was about who has the authority to shape the moral behaviour and standards of a society. In Brahmanism, that right belonged to the brahmanas, the only ones authorised to teach and interpret the Vedas, the source of all knowledge and wisdom. But Asoka upset that apple cart vigorously.
The best way to frame the radical nature of what Asoka did is to highlight the most famous of his inscriptions, Major Rock Edict 13, where he expresses his remorse over the bloodshed in the Kalinga war. In this one rock edict alone, the word ‘dharma’ is used 12 times. If you put all the 33 Ashokan inscriptions together, the word is used 111 times, writes Olivelle in ‘The Semantic History of Dharma in the Middle and Late Vedic Periods’, 2004. By contrast, Dharma is found only 67 times in the entire Rigveda of 10,600 verses; 13 times in the Atharva Veda; and only in 22 passages in the four Yajurveda Samhitas put together. In other words, this vast Brahmanical literature contains the word Dharma no more than 102 times, less than its use in the brief inscriptions of Ashoka.
The scale of the surprise goes up a few octaves when you start counting the number of times ‘dharma’ is used in the Upanisads, the texts that explore philosophical and religious questions and the fundamental nature of reality and where one would, therefore, expect dharma to play a central role. “That, however, is not the case,” writes Olivelle, adding: “In the four early prose Upanisads - Brhadaranyaka, Chandogya, Taittiriya, and Aitareya (in which the term does not occur at all) - the term occurs in just nine passages.”
But soon after Asoka, who called his inscriptions ‘dhammalipi’ or dharma texts, the word blossoms into something extraordinarily pervasive, powerful and, ultimately, fundamental to all of Indian culture as it would come to be seen later.
Thereby hangs a tale, of a sustained semantic battle over the use and meaning of ‘dharma’ between two different traditions, that of the brahmanas and sramanas. Before Asoka, the Buddhists had also used the word to describe the teachings of the Buddha, even terming the first sermon after his enlightenment as ‘Dhammacakkapavattana Sutta’ or the ‘Discourse That Set The Wheel of Dharma in Motion’.
But after Asoka, the term ‘Dharma’ began to appear very frequently, with one Brahmanical text after another trying to recapture and redefine the word, first through Dharmasutras - or the texts that lay down the rules of conduct for members of the Brahmanical society - and then through the epics and innumerable other didactic and literary texts.
All the Brahmanical texts, starting with the Dharmasutras, uphold the validity of the Vedas and the four-fold hierarchical division of society into brahmanas, ksatriyas, vaisyas and shudras. Nothing brings alive the contrast more than the following two quotations, the first from Asoka’s Pillar Edict II: “Dharma is excellent! But what is dharma? It is the absence of the causes of sin, abundance of good acts, pity, giving, truthfulness and purity.” No mention of varna in there.
Apastamba Dharmasutra, probably the earliest Dharmasutra written in the 3rd century BCE, seems to be showing its irritation over every Gautama, Mahavira or Maurya trying to interpret Dharma in their own way, in verse 1.20.5-9: “Let him not become vexed or easily deceived by the pronouncements of hypocrites, crooks, infidels and fools. The Dharma and Adharma do not go around saying, ‘Here we are!’… An activity that Aryas praise is Dharma, what they deplore is Adharma...’ The Sanskrit word translated here as ‘infidel’ is ‘nastika’ which usually refers to those who do not believe in the Vedas, like the sramanas. The butt of Apastamba’s disdain, therefore, is clear.
Says Patrick Olivelle in ‘Asoka: Portrait of a Philosopher King’: “We have seen that the very reason Apastamba thought about writing a book on dharma may have been Ashoka and his popularization of this cornerstone of his moral philosophy. Pre-Ashokan Brahmanical tradition had not focused much on this concept. Now Brahmanical intellectuals were scrambling to salvage dharma from ‘hypocrites, crooks, infidels and fools’.”
What would have been poignant to Brahmanism at this time was the fact that these ideas inimical to its world view and existence itself were no longer playing out in the distant region of Magadha that Brahmanism had deep disdain for, but in Aryavarta itself. So when it came up with its own series of texts using ‘dharma’ in a way that was out of sync with the way Asoka, the Buddhists or the Jains used it, it reiterated the varna system with great vigour.
Apastama Dharmasutra, for example, begins its section on social classes with a blunt, matter-of-fact statement: “There are four classes: brahmin, ksatriya, vaisya and shudra. Among these, each preceding class is superior by birth to each subsequent. Those who are not shudras and are not guilty of evil deeds may undergo initiation, undertake vedic study, and set up sacred fires; and their rites bear fruit. Shudras are to serve other classes; the higher the class they serve, the greater their prosperity.” (13) The hierarchy of varnas undergirds all that the Dharma texts have to say.
If the Brahmana texts patterned the idea of a varna system into the universe itself, the Dharma texts patterned the varna ideology into every nook and corner of Arya custom and belief from the cradle to the pyre, while visualizing and presenting the result as a well-ordered and desirable form of society and government founded on a fully worked-out scheme of hereditary hierarchy.
It is unnecessary to list the discriminations that the Dharma texts instituted since these are well-known, but one verse is telling. Manava Dharmasastra or Manusmriti says in 2.31-32 that the first part of a brahmaṇa’s name shall denote auspiciousness; a kṣatriya’s, power; a vaiśya’s, wealth; but a shudra’s shall express contempt. This is perhaps the “ascending scale of reverence and descending scale of contempt” that B.R. Ambedkar would later refer to, the contempt being greater the lower on the scale you go.
So you could say what started out as a tentative formation of a single caste to administer a newly-created common canon in the form of the Vedas around the beginning of the first millennium BCE had, a thousand years later, around the beginning of the Common Era, been fully fleshed out, theoretically, theologically and legally, into a full varna system with well-defined, hereditary roles for all to play. The last piece of the conceptual varna scheme was put in place by Manava Dharmasastra composed around the middle of the second century CE: it brought in the theory of asramas, or stages of life, within the ambit of the varna system, which would lead to the formulation of the vaunted system of varnasrama dharma that kings all across the subcontinent will commit themselves to upholding in their inscriptions, from around the middle of the first millennium CE.
But the mere fact that the varna system of hierarchy and discrimination had been fully worked out by around the beginning of the Common Era does not mean it had been put into practice around the subcontinent by then. These texts have to be seen as agenda-setting templates for the future, when there will be enough willing ‘ksatriya’ rulers to apply the principles of varnasrama dharma and that was still many centuries in the future. In fact, the period when the varnasrama system was being fully articulated was also the period when Brahmanism was facing its greatest challenge ever, in its own heartland.
The period roughly between 200 BCE and 300 when the Indo-Greeks, Sakas and Kusanas were successively in control of large parts of north and northwestern India would also see the strongest period of growth for the Sramanic religions, especially Buddhism. That is not to say that these rulers only patronized Buddhism; many of them were, in fact, patrons of deities and heroes who would later loom large in the Brahmanical pantheon, including Siva, Krsna, Karttikeya and Balarama, to such an extent that they put them on their coins.
The foreign rulers were eclectic in their religious tastes and would pick and choose from the wide array of divinities they had available before them, from Hellenic, Iranian/west Asian and Indian traditions. In that sense, one could say rulers in this period provided a broadly level playing field for different religious or cultic persuasions to play out.
The result was that by the middle of the second century CE, around the time when the Manava Dharmasastra was being written, Buddhism was making its presence felt all the way from Khotan in northwestern China to Anuradhapura in Sri Lanka and from Merv in Turkmenistan to Luoyang in central China – an expansion that began around the time of Asoka and reached new heights under Kaniska, both of whom are seen by Buddhist traditions as their great patrons, along with the Indo-Greek king Menander.
Buddhist stupas and viharas (monasteries) sprung up all across the subcontinent and beyond, bearing testimony to the greatest expansion of a religious movement seen anywhere in the world until then. To quote Gregory Schopen in his paper, ‘A Well-Sanitized Shroud: Asceticism and Institutional Values in the Middle Period of Buddhist Monasticism’, ‘Buddhist communities came to be fully monasticized, permanently housed, landed, propertied, and—to judge by almost any standard—very wealthy,’ in the period between the Maurya and Gupta empires..
There might have been a particular wind at the back of Sramanic religions during this period and before - all the way from the time of the Magadhan empires to the end of the Kusana period – namely, the growing importance of trade, urbanism and the rise of rich merchants as a new dynamic force in society. Trade would have started gaining prominence when the new, powerful dynasties of Magadha came up, such as that of the Haryanka, with kings Bimbisara and Ajatasattu expanding their territory and, therefore, a common market, by conquering or merging other kingdoms with theirs.
Buddhist texts are replete with mentions of rich merchants and moneylenders and impressive caravans with hundreds of carts carrying goods long-distance. Digha Nikaya mentions a caravan of wagons managed by a thousand men, and another caravan of a thousand carts going from the eastern region into the western region. (14) Brahmanical texts of the same period do not pay a similar attention to trade or cities, for that matter.
In fact, for probably the same reasons that trade and urbanism in general and cities in particular were giving the Sramanic religions an advantage, Brahmanical texts were insistent in their dislike of them. “The Vedic civilization did not like them,” writes Bronkhost. He quotes the Apastamba Dharmasutra as saying one “should also avoid visiting cities,” and the Gautama Dharma Sutra as saying, ‘according to some, Vedic recitation is always suspended in a town’. (15)
This period of intense dynamism and expansion of Buddhism, especially under the Indo-Greeks, Sakas and Kusanas, finds reflection in the Brahmanical texts in the form of dire forebodings about the Kaliyuga. In Book Three of the Mahabharata, sage Markandeya talks about how things will be at the end of the age, and it is remarkable for its allusions to the Sramana religions – their ‘reliquaries’ and their predilection for logic and argumentation, for instance - and also for showing us how the authors of the Mahabharata felt about the period of the Sramana resurgence: “The brahmanas perform the tasks of shudras. Shudras become the earners of wealth and resort to the dharma of ksatriyas. This is what occurs when the yuga decays. When Kali Yuga arrives, brahmanas abstain from sacrifices and studying. They discard the offerings of oblations to ancestors… Brahmanas no longer meditate and shudras become addicted to meditation.” (As translated by Bibek Debroy)
What that means is from the period of the Brahmana texts where the four-class hierarchy became commonplace, to the period of the Dharma texts that clearly articulated the privileges and deprivations of the four castes, the varna system was always severely contested, at first outside Aryavarta and then, within Aryavarta itself. Even in the centuries around the beginning of the Common Era, the caste system was yet to find its larger, subcontinental footing.
But this situation will soon change, and what will ring in the change is a shift in the relative importance of trade and agriculture in the society of the period. Between 235 CE and 284 CE, the Roman Empire went through what is called the ‘Crisis of the Third Century’ that almost brought it down. In broadly the same period, the Kusana empire also started disintegrating, due to internal fragmentation, external invasions and, of course, trade disruption. A period of great prosperity supported by internal and external trade between India and the rest of the world that had begun during the period of the Nandas and Mauryas and continued during the period of the Indo-Greeks, Sakas and Kusanas, began sputtering after the time of Vasudeva I, the last powerful ruler of a unified Kusana empire (circa 190 to 230 CE).
But the next source of prosperity was already becoming evident: the deepening and widening of agriculture across the subcontinent. This could only happen if millions of people were drawn or assimilated into the fast-growing agricultural system as labourers, especially as agricultural settlements moved into the lands of forest-dwelling or semi-nomadic groups, causing their displacement. And this was a humongous task that the dozens of new kingdoms that came into being as agriculture started in different parts of the subcontinent, took on eagerly, from around the middle of the first millennium CE.
In this task, many of them found the social framework based on a hereditary hierarchy that had been articulated by the brahmins of the Aryavarta in the previous period useful. The Sramanic religions did not have a comparable offering. To implement the new system or hierarchy that ordered society into castes and the outcastes - and to lend it legitimacy – many kings, including those who professed the Buddhist or Jaina faith, often invited brahmanas from Aryavarta to come and settle in their kingdoms on lands granted to them. But such migration was not always necessary, as we will see. Also, there was no uniformity in the way the varna-jati system was customized for different regions. As Suvira Jaiswal put it in ‘Caste: Origin, Function and Dimensions of Change’: “Caste ideology provided an integrating mechanism that did not require uniformity of replication of the four-fold varna structure.”
The interconnection between agricultural expansion along the river valleys, state formation, caste hierarchy and the role that land grants and temples played in all this have been studied extensively by some of the most luminous minds of our time, including Professors R S Sharma, Brajadulal Upadhyay, Suvira Jaiswal, Rajan Gurukkal, Kesavan Veluthat, Upinder Singh, Hermann Kulke, R. Champakalaksmi and many others, so I will not get into that, except to make three points.
The first one is about the humongous amount of intellectual effort that went into building the kind of theoretical ballast that the system needed, over and above the concept of varnasrama dharma itself. Before the temples of the new agrarian-temple-state formation could arise, the Vedic religion built around sacrificial rituals and aniconic gods had to become one centred on bhakti (devotion), pilgrimages and image worship instead; the Brahmanical pantheon needed new divinities who had a local presence on terra firma like Siva, Skanda, Krsna and Balarama unlike the Vedic gods of previous centuries who were mobile and placeless and would arrive wherever the sacrifices are held by the Vedic pastoralists (16); the epics and other literary texts had to be composed so that, one, the lower classes who were denied access to the Vedas could be informed of the norms of varnasrama dharma and two, future ksatriya kings could be shown what is expected of them; the Puranas had to be written so that the integration of local cults into the Brahmanical pantheon could be solidified, new gods could find a place beside the old and sacred spots could be made part of a network of pilgrimage centres…the list is near-endless. The striking point to note is that much of this happened in the centuries around the beginning of the Common Era, when Brahmanism was facing its severest challenge from the Sramanic religions, especially Buddhism. It responded and recast itself in a manner rarely seen.
The second point I want to make is about the Sramanic religions which were the main source of resistance to Brahmanism and varnasrama dharma. It is often pointed out that ideas of karma and ahimsa that were Sramanic in origin helped strengthen the caste system. Karma because it allowed texts like the Chandogya Upanisad (5.10.7-8) to connect being born in a low caste to things done in a previous birth, and ahimsa because it gave a reason for people to look down on those whose work involved causing injury to living beings.
Though there is a point to these arguments, these are also misleading about the role that Sramanic religions played in the institution of the caste system. None of the Sramanic religions had primary texts that created and defended the varna system or exhorted kings to uphold the varnasrama dharma or taught everyone to do their varna duty. No Sramanic religion saw the system as anything other than a human creation either. Over time, however, as the caste system solidified across the subcontinent and established itself as the default social order –just as capitalism is today in much of the world, or slavery was in the American South during the 19th century – they had to make certain accommodations with it, with Jainism doing more of this and Buddhism far less.
That the Bhakti movement in Tamil Nadu in the sixth-to-ninth centuries which did much to kickstart the temple-centred agricultural economy of the region saw the Jains and Buddhists as opponents and expressed extreme hostility to them even involving violence shows that even in middle-to-late first millennium CE, the religions that started out in Greater Magadha were still standing in the way of the belief system that started out from Aryavarta. Nothing brings out the continuing embers of this opposition more than the following quote from the Buddhist logician of Nalanda, Dharmakirti, who lived in the 7th century. In the first chapter of his well-known work Pramanavartika, he writes: these five are the marks of the crass stupidity of witless men.” (17)
The last of the five barbs – about the practice of mortification – is aimed at the Jains, but the remaining four are at aimed at the form of Brahmanism that was prevalent during his time and included in them is the phrase, ‘arrogant division into castes’, suggesting that more than a millennium after the Buddha, the Greater Magadhan discomfort with the Aryavartan varna-jati system was still continuing.
The last point I would like to make, though not a new one, is that it was not always necessary for brahmins to be present for Brahmanism to spread. The mechanism of a graded hierarchy with a descending scale of contempt and a close alliance between the ruling power and priestly power was so self-evidently well-suited to the needs of expanding agriculture that it found a hundred different ways to replicate itself in new areas being brought into agriculture. In her book, ‘Caste: Origin, Function and Dimensions of Change,’ Jaiswal mentions the case of the Rayalaseema area of southwestern Andhra Pradesh which had been populated by hunters and gatherers. It acquired strategic importance in the struggle between the Chalukyas and Pallavas in the 8th-9th centuries and was thus exposed to the influence of advanced cultures. To quote Jaiswal: “There gradually unfolded transition from tribe to state, leading to the rise of tribal chieftaincies and tribal communities were transformed into a stratified society with the growth of a political nobility exercising control over land. This new class patterned itself on the Brahmanical model which helped legitimise their separation from the commonality of tribesmen and demanded their allegiance.”
The attraction of the caste system to those in power is explained well by Prof. Irfan Habib in ‘Interpreting Indian History’, P 19, where he states that the caste system made agricultural labour cheap and reduced the cost of artisanal products and services by restricting their mobility. In turn, this allowed rulers to extract more of the surplus from the peasantry – irrespective of even whether the rulers were inside or outside the varna system. Then he made a remarkable point about Muslim rulers or writers never questioning the caste system and instead focusing on its alleged polytheism and idol worship, despite the caste system not being part of Islamic law.
What applied to the Mughals, or the Marathas, also applied to Buddhist, Jaina or even shudra rulers of the previous centuries whose inscriptions affirm that they uphold the Varnasrama dharma which had become the only game in town by the late first millennium CE. Jaiswal, as usual, puts it succinctly when she says, in ‘Caste: Origin, Function and Dimensions of Change: “An exploitative system which has the capacity to enrol the best of whatever origin in its own service is far more pernicious and long lasting than the one that is closed and static.”
The core of the varna-jati system had two parts that will continue functioning: one, the alliance between the ruling power and priestly power which is what started the whole process and two, what Ambedkar described as “an ascending scale of reverence and a descending scale of contempt.” (18) Even while adjusting itself to political exigencies and economic contingencies and battling opposition over many centuries, it will hold on to these two core principles that will come to shape many social attitudes. This is perhaps what led sociologist Dumont to think up the concept of Homo Hierarchicus in 1966, but without paying sufficient attention to the long resistance against it and the way it was contingently shaped by economics and politics. Thirty-five years later, anthropologist Nicholas Dirks would publish another, last-mile snapshot of caste in his ‘Castes of Mind’, but without paying sufficient attention to the step-by-step evolution and geographical expansion of caste from the first millennium BCE onwards. What we know now is that caste was historically contingent in its origins and socially contested throughout most of its history and eventually gave rise, in the twentieth century, to Ambedkar’s call to ‘annihilate’ it.
1. See Suvira Jaiswal, ‘Caste: Origin, Function and Dimensions of Change’, 1998, P 42-43
2. As translated by Stephanie W Jamison and Joel P Brereton, ‘Rigveda: The Earliest Religious Poetry of India’
3. Quoted in Romila Thapar, ‘Interpreting Early India’ 1992, Page 106; Alf Hitebeitel, ‘Empire, Invasion and India’s National Epics’, 1998, Page 408
4. See D.D. Kosambi, ‘The Culture and Civilization of Ancient India in Historical Outline’, 1964, P 208-210
5. As translated by Patrick Olivelle, ‘Upanisads’, 1996
6. As translated by Brian K Smith, ‘Classifying the Universe: The Ancient Indian Varna System and the Origins of Caste’, Pages 68, 69
7. See Patanjali’s Mahabhasya 1 P 475 1.3 (on P. 2.4.10), as cited in Johannes Bronkhorst’s Greater Magadha, Page 1. Also Baudhayana Dharmasutra 1.2.9 and Vasistha Dharmasutra 1.8-12
8. See Narendra K. Wagle, ‘Society at the Time of the Buddha’, 1966, Page 38
9. Adapted from the translation by Bhikkhu Sujato
10. See Chandogya Upanisad 5.3.7; Brahadaranyaka Upanisad 6.2.8
11. See Mahabhasya (I p. 476 l. 9; on P. 2.4.12 vt. 2), cited in Johannes Bronkhorst’s ‘Greater Magadha,’ Page 84
12. As translated by Padmanabh S. Jaini in ‘The Jaina Path of Spiritual Purification,’ 1979, Page 74
13. As translated by Patrick Olivelle in ‘Dharmasutra: The Law Codes of Ancient India’, 1999
14. See Narendra K Wagle in ‘Society at the Time of the Buddha’, 1966, Pages 15, 31
15. Johannes Bronkhorst, ‘Greater Magadha: Studies in the Culture of Early India’, 2007, Page 251
16. See Knut A Jacobsen’s ‘Pilgrimage in the Hindu Tradition: Salvific Space’, 2013, Page 44
17. Translation by Padmanabh S Jaini, in ‘Sramanas: Their Conflict with Brahmanical Society’
18. See Dr. Babasaheb Amebedkar’s ‘Annihilation of Caste,’ 1944. The phrase used in ‘Dr. Babasaheb Ambedkar Writings and Speeches,’ 1987, Volume 3, Page 105 is slightly different: ’an ascending scale of hatred and a descending scale of contempt’.