20 July 2012

Revisiting Greater Magadha

WHEN JOHANNES BRONKHORST'S BOOK Greater Magadha hit the scene a lot of us were over-awed by the scope and complexity of the argument. I wrote about my first encounter with it in a blog post called Rethinking Indian History (2009). A the time we thought it must be significant, and the new theory did seem to solve some of our problems. It was exhilarating to realise that history was able to be re-written.

However such a book is difficult to assess, and even after several years there has been little critical response to it from the field of Buddhist studies. A few reviews, but nothing of real substance. This may be because in order to place Bronkhorst's claims in context one must have a good grasp of a body of literature and evidence that is unfamiliar to most Buddhologists. If we haven't read Bronkhorst's oeuvre for example we'll struggle to really grasp where he's coming from. We also need to be familiar with writers on Indology such as Michael Witzel and Asko Parpola (neither of whom are read by many Buddhologists). The archaeological and anthropological studies are also critical - and they are scattered and from an entirely different discipline. And this work also relies on familiarity with Vedic literature and on the philological problems of dealing with it. In other words there is not much criticism because not many of us are qualified to read Bronkhorst critically.

At the same time Bronkhorst's book seems seems to have over shadowed Geoffrey Samuel's book The Origins of Yoga and Tantra which came out a year later. In fact Samuel is by far the better author, his book is far more readable and accessible, and about 1/10th of the price! His treatment of the relevant material (mainly based on an unpublished book by Thomas Hopkins) seems more credible, though it still has it's limitations.

At the time I was very enthusiastic about Greater Magadha because it was one of those moments when I realised that everything I thought I knew might be wrong, and what could be more exciting for a scholar? However I've been reflecting on Bronkhorst's book in the light of Geoffrey Samuel's book, and particularly Michael Witzel's equally awe inspiring writing (again far more accessible since he shares pdfs of many of his publications for free!). My conclusion is that Bronkhorst's thesis will not stand the test of time.

Bronkhorst, as my friend Dhīvan said in his recent M. Phil. thesis, is often arguing tendentiously. I like this word. It means that Bronkhorst has a conclusion that he is pursuing and this is reflected in how he treats the evidence. Everything is predicated on Bronkhorst's revised chronology and presented in such a way as to support his conclusions. In my view the evidence is read in the light of the theory, which is the opposite of the scientific method.

Poor Reasoning

There are some examples of faulty work in the Book. For example in the Appendix VI covering Brahmins in the Canon. Here he notes that the Ambaṭṭha Sutta may well refer to Sanskrit ambaṣṭha: i.e. someone born of a brāhmaṇa father and a kṣatriya or vaiṣya mother. Ambaṭṭha turns out to have a Brahmin father and a mother descended from a slave and is therefore low caste. Bronkhorst argues that here ambaṣṭha/ambaṭṭha must refer to the mixed caste of the interlocutor which is plausible. However a slave is not a kṣatriya or vaiṣya so Bronkhorst is stretching the evidence to suit himself. Richard Gombrich using the same kind of argument when arguing that the Buddha must have known about the Puruṣasūkta (ṚV 10.90) because he refers to Brahmins being born from Brahmā's mouth in the Tevijja Sutta. Bronkhorst points out that in the ṚV the Brahmins are born from Puruṣa's mouth, not Brahmā's and concludes that they Pāli authors "did not know" the sūkta (p.213). Bronkhorst seems to have a rather irrational aversion to Gombrich and it shows here in his inconsistent standards in treating the evidence. It also shows in his treatment of the humorous passages of the Pāli suttas which do not get a laugh from him.

Another example is the conclusion that because the Pāli texts are familiar with an idea found in the Dharmasūtras, that the Pāli texts must be late. The Dharmasūtras are much less securely dated than the Pāli, though the consensus seems to be that the written texts are originally post-Asoka. However it is also widely accepted that they codify conventions that are a great deal older, so there is no a priori reason to assume that a detail in isolation is late because it is found in a Dharmasūtra. And the Pāli parallels are all details in isolation.

Bronkhorst is caught out using fallacious reasoning on two separate occasions and this must put us on our guard. These examples are from areas I understand well enough to be sure of my ground. It might be argued that these are relatively minor infractions, but if someone like me can spot these kinds of minor problems, what are the professionals seeing? (and when will they write about them?)


Closer to the heart of the matter is that the very concept of Greater Magadha seems flawed. Yes, there are two cultures on the Ganges plain ca. 1000 BCE and one of them is the Kuru-Pañcāla state. The other one is not Magadha, but the Kosala-Videha complex which is formed from Vedic tribes forced to move east by the rise of the Kurus. Witzel has referred to these tribes as para-Vedic, as they seem to have had customs significantly different than the Kuru Vedic tribes. Videha in particular retains connections with the Kuru Brahmins and the Videhan Kings invite them east. This is what we see in Bṛhadāraṇyaka Upaniṣad when King Janaka invites a number of orthodox/orthoprax Brahmins to a debate. A debate which local boy Yājñavalkya wins. Yājñavalkya represents, even personifies, a major shift that is going on in the Brahmanical world. A shift away from the orthopraxy of the Kurus towards a new form of Brahmanism that forms the basis of what comes after. Not only is BU composed in Videha but so is the White Yajurveda, and the single extant recension of the Ṛgveda (which once existed in a number of different recensions). At this time Magadha hardly features in texts at all. Geographically Magadha is isolated from the major players by being south of the Ganges.

By the time of the Pāli texts Kosala is clearly more welcoming to Brahmins than Magadha which is further east, and crucially (in my opinion), south of the Ganges. The Pāli texts show more Brahmin towns and more land gifts to Brahmins in Kosala than in Magadha. In fact Bimbisāra gave only two grants and his murderous son Ajatasattu is not recorded as giving any. Later we know that the Mauryas were not converted to Brahmanism, but still followed śrāmaṇa religions.

Witzel considers that the evidence of the texts themselves, especially the language involved, show that the early Upaniṣads must predate the early Pāli texts by some centuries. Although he does point out that for both literatures there is a long gap between initial composition and final redaction, and this blurs the boundaries. The early Upaniṣads represent a time before the Second Urbanisation (ca 600-500 BCE), while the Pāli texts represent a time when it is in full swing. A difference not dealt with by Bronkhorst as far as I can see. Magadha as a power, with its fortified capital city Rājagaha, is only associated with the Second Urbanisation.

"Greater Magadha" as a region, then, only has meaning in Bronkhorst's idiosyncratic revised chronology which places all of the Upaniṣads post-Buddha. If anything the region is Greater Kosala in the late Vedic period (ca. 700-500 BCE)! It is true that the idea of ethicised karma does make a first appearance in BU, and I think Bronkhorst is on the right track when he says the Buddhist idea is not a development of Vedic eschatology, or at least not a direct development. What seems to mislead Bronkhorst is the idea that the source of this idea came along only in the 5th century. I believe this is short sighted, and ignores what we know about the history of ideas in India.


A fact which no scholar has yet come to grips with is that Brahmins, as far as they are recorded in the Pāli texts are quite diverse: we have at least ritualist, renunciate and theistic Brahmins, we also have some that are just plain villagers. Upaniṣadic ideas and practices are not found with any clarity in Pāli, they don't stand out, but they can apparently be inferred. We never see the Buddha in conversation with a Brahmin about ātman for instance, or about brahman, or the identity of the two (leading to mokṣa), or about oṃ, or the vedas. Where Brahmins express religious ideas in the Canon they are cosmological or related to a Creator God. The cosmological ideas are likely to have been common knowledge. The central ideas of the Upaniṣads are missing from Buddhist texts. This might be seen to support Bronkhorst's thesis, but I'm not so sure. My guess is that Brahmins maintained a relatively orthoprax exterior and kept the Upaniṣads secret for a long time--the word upaniṣad can mean 'esoteric'.

The theistic Brahmins have yet to receive adequate attention from scholars. Gombrich treats references to Brahmā as a criticism of brahman, but this only works in the specific context of the Tevijja Sutta, and what we see throughout the Canon is no mention of brahman, and many mentions of Brahmā. The theistic tendency has parallels in parts of the Mahābhārata, and may represent a kind of short-lived orthodoxy that is quickly over-written by the cults of Śiva and Viṣṇu which relegate Brahmā to saṃsāra just as the Buddhists did.


Taking Witzel's (1997) suggestion that the Śākayas arrived in North-East India rather late, I have developed this idea in my forthcoming article (draft on academia.edu). All things considered we can probably say that they arrived in the decade or two following 850 BCE. That year (± ~10 years) marks the beginning of a major dry period in India. Witzel notes that other North-Eastern tribes such as the Malla and the Vṛji were known to live in the West (Rajasthan and the Panjab) by early Vedic texts, but are neighbours of the Śākyas in the Pāli texts. At this time Kosala-Videha culturally dominates the Central Ganges region, and the Magadhan city of Rājagṛha is just about to be founded.

The argument is quite involved and requires the weighing of many separate items of circumstantial evidence, but a case can be made for contact between the Śākyas and the Zoroastrian culture of Iran. What is suggested by this line of argument is that the idea rebirth is found throughout Indian (perhaps it was an indigenous belief) but the introduction of ethicisation follows contact with Zoroastrianism. I try to make the case for this happening in the 9th century BCE, giving it time to infect the early Upaniṣads. However it could have come with Achaemenid influence after Darius claims Gandhāra and Sindh as provinces of Persia ca. 520 BCE.

Revised Chronology

One of the problems with Bronkhorst's argument is that he mixes texts from different eras and is relying on conjectural reconstructions. So he contrasts the Bhagavadgītā which is certainly written in the common era, with other bits of the Mahābhārata (post Asoka, but probably BCE), Pāli texts (ca 4th century BCE) Upaniṣads (7th-5th century BCE), and reconstructions of ideas of early Jain and Ājivika beliefs. This is not comparing apples with apples. A century of ideological development in a milieu which sees a lot of mixing and matching, assimilation and adaptation of each other's ideas and practices, can see major changes. So it seems to me Bronkhorst's method is flawed.

As I said above everything is predicated on Bronkhorst's revised chronology and presented in such a way as to support his conclusions. In other words one has to accept the his new chronology, which places the Upaniṣads after Buddhism rather than before it, and allows Brahmins to absorb ideas, particularly karma and rebirth, from the śrāmaṇa milieu. But the reasoning is circular. The thesis only works if we accept the chronology; while the chronology only fits if we accept the thesis. The same argument applies to the reconstructions of Jain and Ājīvaka religious ideas, especially the latter which are reconstructed mainly from Buddhist texts (a rather unreliable source of information!)


I think Bronkhorst has made a valuable contribution to the historiography of India. He has certainly made many of us rethink our understanding of and approach to the history of India before the Common Era, and this is a valuable service. A major challenge such as this forces us to be more precise in stating our differences of opinion if we have them. There are reasons to be cautious in accepting Bronkhorst's argument. I find I am persuaded by Witzel's account of the evidence as much because he seems to have no particular agenda as anything. Witzel has repeatedly, and at considerable length, played with the pieces of the evidential jigsaw in order to make a coherent picture from them. Samuel has showed that it is possible to read the archaeological evidence as supporting the consensus chronology. Following Witzel I have tried to show that the ideas might have come from a third source, Zoroastrian Iran, and been introduced into śrāmaṇa and Brahmin culture at roughly the same time. (Revising the article for publication is my next job).

Another plus is that Bronkhorst has made it abundantly clear that Buddhism can no longer be studied in isolation, but is a branch of Indology. Ignorance of archaeology and material culture (the gist of Greg Schopen's critique of Buddhist studies as a subject) is no longer acceptable. The Late Vedic literature--the Epics, Early Upaniṣads, Brāhmaṇas, Dharmasūtras, Dharmaśastras and even the Gṛhyasūtras--is starting to look more relevant in understanding early Buddhism. Early Buddhism existed in a context and we have been overlooking, or over-simplifying this context for too long. The downside of this is that an already complex subject appears to become an order of magnitude more complex. And this at a time when we are just beginning to make use of the Chinese parallels to the Pāli Nikāyas and discover the influence of Central Asia in transmitting Buddhist to the East. And this also at a time when Buddhist studies is dying out as an academic subject in the UK.

So far as I am aware no scholar has adopted Bronkhorst's revised chronology. And the whole thesis depends on acceptance of the chronology. It may be that more time is required for scholars to assimilate Bronkhorst's work, and to provide a critique. But in the meantime there are some obvious flaws in it that should make the reader wary of just accepting what he says uncritically.

I want to conclude with a coda on critical discourse. Not so long ago I was speaking to a prominent long time Buddhologist and they remarked that criticising someone else's work in print was coming to be seen as unacceptable. Certainly in the US where tenure depends on a positive reaction to one's work, critical dialogue is dwindling. Journals have apparently refused to publish critical articles.  If the refutation aspect of conjecture and refutation is abandoned, then progress in knowledge inevitably goes awry. Just look at economics!


Bronkhorst, Johannes. 2007. Greater Magadha : Studies in the Culture of Early India. Leiden: Brill.

If you know of other reviews drop me a line.
Related Posts with Thumbnails