03 July 2015

Early Mahāyāna: Everything You Know is Wrong

Revised 5 Jul 2015.

The origin of the Mahāyāna has been a subject of some fascination over the years. In its mature form, Mahāyāna Buddhism could hardly be more different from Mainstream Buddhism and still be thought of as Buddhism. A variety of theories have been proposed for how the Mahāyāna came about. In this article, I will précis some recent articles which revise or review of the origins and development of Mahāyāna Buddhism.

Note that the articles that this essay is based on are all available on the internet via academic.edu (individual links are included in the bibliography). I highly recommend the two articles by Drewes (2010a, 2010b). They are very accessible and thought provoking. Karashima's articles are long and technical and general readers may find them a bit daunting, but they never-the-less also provide important insights (and post date Drewes by some years and thus also provide a contrast to his work).

The Name Mahāyāna

Like some other Buddhist terms coined in Prakrit, it seems Mahāyāna might have been the victim of a wrong Sanskritisation. We already suspect that sūtra 'thread' ought to have been sūkta 'wise saying', both words assimilate to sutta in Pāḷi; while satva† 'being' ought to have been sakta 'committed, intent on', both satta in Pāḷi. Hence, bodhisatva ought to be bodhisakta 'committed to awakening', and it's possible that mahāsakta might have signified 'one whose commitment is great'. Karashima (2015b) provides a comprehensive survey of the word mahāyāna in the various versions of the Lotus Sutra. He argues that the word was probably intended to be the equivalent of the Sanskrit mahājñāna 'great knowledge', but that this was pronounced mahājāna in Prakrit. There was a natural ambiguity with the word mahāyana 'great vehicle', and the Lotus Sutra plays on this to some extent (see Karahima 2015b: 215-217). Later, the ambiguity was resolved in the wrong way and mahāyāna became the standard interpretation instead of mahājñāna. So, the fact that we talk about a great vehicle and not a great knowledge is a quirk of history.
†Buddhist manuscripts virtually always spell this word satva, so that, arguably, the correct form in Buddhist Sanskrit is satva. It has been further over-corrected to sattva by editors to make it conform with Classical Sanskrit.
Karashima argues that the use of the word was transferred from the Lotus to the Aṣṭasāhasrikā Prajñāpāramitā Sūtra (2015a: 115) where the word only occurs in the parts of the text that are considered to have been added towards the end of the composition/compilation process.

The term mahājñāna is probably synonymous with terms like sarvajñā, found frequently in Prajñāpāramitā texts, and to prajñāpāramitā, as well. 

Karashima further conjectures, from his own research and work done independently by Peter Skilling (cited 2015a: 117), that the title sūtra or mahāyāna sūtra is a later affectation. Such texts are frequently referred to as paripṛccha (question), nirdeśa (description), vyākaraṇa (explanation/analysis), or vyūha (arrangement/manifestation), as well as sūtra. The addition of sūtra is frequently superfluous.

Many texts have the word Mahāyāna in their title. These titles first appear around 400 CE. Karashima (2015a) shows, by surveying the Chinese Canon, that this was a change that happened over time. He proposes that, originally, they were known as 'irregular' sūtras, signified by Middle Indic *vedulla = Pali vetulla, Gāndhārī *veulla or *vevulla. This corresponds to Sanskrit vaitulya. This word started appearing in Chinese texts about the 2nd century. However, because of the ambiguity in the Middle Indic forms and a change in the perception of these texts, by the 5th century the Middle-Indic word began to be interpreted as vaipulya, 'extensive, incomparable'. One of the characteristics of the vedulla suttas was that they consisted of a series of questions and answers, characteristic of the paripṛccha texts, but also the Prajñāpāramitā texts. Later again, these vaipulya texts were renamed Mahāyāna texts.
* The asterisk here stands for a term derived from grammatical rules, but not found in any extant text.

So it seems that history has tricked us once again. As we will see below, the early Mahāyāna texts seem to have been composed in Prakrit and only translated into Sanskrit much later, late enough for the translators to no longer have a clear understanding of the intentions of the original authors, and to be mislead by the received tradition. We have so many examples of this kind of thing that we must admit that Buddhist lineages of passing on teachings were quite unreliable. Buddhist lineages amount to a game that American kids call "Telephone". The idea that because your teacher tells you they taught you what their teacher taught them, is no reason to believe that what you have received reflects an unchanging tradition.

What the Mahāyāna is Not
"Mahāyāna was not a distinct sect. It did not involve the worship of bodhisattvas. It was not developed by lay people. It was not an offshoot of the Mahāsāṃghikas. It was not a single religious movement." Drewes (2010a: 59)

Early theories

In his two articles, Drewes sums up a generation of research into the early Mahāyāna. Mostly, it is the story of wrong turns and false assumptions, many of which have the origins in the 19th century. For example, the idea that Mahāyāna was primarily a lay movement can be traced to an 1865 article by V. P Vasilev. The first actual lay origin for the Mahāyāna was put forward by Jean Przyluski in the 1920s and 1930s. Similarly, the idea that the Mahāyāna involved the rejection of the (so-called) arhat ideal was first put forward by T. W. Rhys Davids in 1881. Linking Mahāyāna to the Mahāsāṃghika sect was a popular 19th century idea, being found in the works of Hendrik Kern, L. A. Waddell, and T. W. Rhys Davids.

Stories of this kind proliferated and became a kind of standard narrative with some variations. The Mahāyāna was a reaction against the narrow mindedness and formalism of the Hīnayāna. Hīnayāna was portrayed as a religion in terminal decline that had preserved texts they didn't really understand (and the 19th century Sangha didn't do much to dispel this view). Mahāyāna was said to have embraced a universal ideal whereas the Hīnayāna was all about self liberation. Mahāyāna was institutionally distinct from existing forms of Buddhism. And so on. In the 1950s, Japanese scholar Akira Hirakawa proposed a new theory, which was that Mahāyāna was a lay movement focused on stūpa worship. (Drewes 2010a: 55)

I would observe that many of these historical narratives owe a great deal to the historical narratives of the schism in the Christian Church that gave birth to the Protestant movement, especially as perceived by Protestant Western Europeans. Protestants identify with the breakaway sect which brings with it a renewal of values and ideals, and (in their own estimation) a greater authenticity. They identify the Mainstream (in this case Roman Catholicism) as intellectually moribund (whereas the Catholic Church has always been more intellectually lively) and morally bankrupt (which was certainly true at the time of the Luther and various other points in history).

The irony is that the heartland of Protestant Buddhism was and is Theravādin Sri Lanka (the story becomes inverted without anyone quite noticing). But there never was this kind of schism in Buddhism! Or at least there is no evidence for it. Mahāyāna activity started small, operated within existing monasteries, and only very gradually over centuries came to dominate Indian Buddhism. Nor, for that matter did Theravāda dominate Sri Lanka until 10th century reforms purged non-Theravāda Buddhism from Sri Lankan monasteries and standardised forms. Many of the features we take to be characteristic of Mahāyāna Buddhism are, in fact, culture norms from Tibet, China or Japan. They bear no direct relationship to Buddhism in India.

The tide began to turn in the 1970s. Greg Schopen's 1975 article introduced a new theory of Mahāyāna origins that directly challenged Hirakawa. Schopen argued that early Mahāyāna groups rejected stūpa worship in favour of what Schopen called "the cult of the book". Drewes (2007) himself critiques this seminal article at length, but it began an American-led re-evaluation of the origins of Mahāyāna.

The decisive moment, however, was Paul Harrison's article "Who Gets to Ride in the Great Vehicle? Self Image and Identity Among the Followers of Early Mahāyāna." (1987). This was based on an examination of the first sūtras to be translated into Chinese. This showed that Mahāyāna was overwhelmingly a monastic movement. The texts show little desire for establishing sectarian identity. Some of them even acknowledge the legitimacy of arhatship. They do not recommend devotion to bodhisattvas. They do show a generally negative attitude towards women. (Drewes 2010a: 55-6). And so the lay origins theory died.

The Breakaway Thesis

The standard story about Mahāyāna was that it began life as a breakaway sect from Mahāsāṃghika. There are several highly contradictory accounts of the schism at the second council which saw the conservative Sthāviras split from the more progressive Mahāsāṃghikas over matters relating to Pratimokṣa rules. Here I find the only bum note in Drewes' analysis. He describes this theory of Mahāsāṃghika origin as having died a quiet death. However, Karashima apparently disagrees because he continues to see, for example, the Aṣṭasāhasrikā Prajñāpāramitā Sūtra as being connected with Mahāsāṃghika (2013). However, the part about Mahāyāna being a breakaway sect is deprecated both for lack of evidence and for positive evidence of continued coexistence for many centuries between the Mainstream and the so-called Breakaways. In fact, what evidence we do have suggests that there was only ever one kind of ordination lineage in India (nikāya), there was no distinct Mahāyāna ordination, and no distinct Mahāyāna institutions. Harrison argued that there was no way to determine any sectarian affiliation of the early Mahāyāna.
† The word nikāya is used in different ways, but here means "a group". The nikāya ordinations are basically the same the Buddhist world over. Theravādins also use the term nikāya to indicate ordination lineages within their school.
Mahāyāna Buddhism appears to have developed slowly. It certainly produced many texts in the early centuries, but little hard evidence. Schopen has noted a single statue of Amitābha, broken off at the ankle, but labelled as such on the pedestal, dating from ca. 153 CE. The oldest epigraphical evidence dates from the 4th or 5th centuries CE.
‡ Epigraphical evidence is from inscriptions typically carved into stone at a Buddhist site. Many record donations of money, often from monks.
Once we strip away all the unsupported conjectures and suppositions about Mahāyāna, little remains. There was no "the Mahāyāna", per se, (and I have tried to avoid the definite article in this essay). We know, from epigraphical, textual and eye-witness reports of Chinese pilgrims that Mahāyāna monks lived alongside Mainstream monks in the same monasteries. These monks were mainly concerned with the production and spread of new Buddhists texts, a major preoccupation of the texts themselves. Over many centuries the emphasis of monasteries changed so that Mahāyāna ideas and values predominated, but the ordination lineages remained the same (as they do in Tibet and parts of China to this day). According to Drewes, there were reactions, but these were by the Mainstream against Mahāyāna, or indeed by one branch of Mahāyāna against another. There was no Mahāyāna reaction against the Mainstream. There was a slow evolution over at least 600-800 years.

Karashima, however, points out that in second stratum of the Lotus Sūtra the dharmabhāṇaka (Dharma preacher) proclaiming the Lotus Sutra were "harshly criticised, slandered for having composed the kāvyas (i.e., the Lotus Sutra itself) and for propagating a heresy" and thus "it is evident that their belief was a very dangerous heresy in the eyes of the Buddhist authorities of that time" (2015a: 115). Similarly, Schopen argues that Mahāyāna authors were defensive with respect to the Mainstream. Discussing the Ratnāvalī, a text attributed to Nāgārjuna:
"Even in the hands of one of its most clever advocates it does not appear as an independent, self-confident movement sweeping all before it as... But rather—and as late as the second or third century—it appears as an embattled movement struggling for acceptance." (Schopen 2005: 7)
And also:
"Sociologists, however, who have studied sectarian groups in a variety of contexts, have shown that [the sort of characterization found in the Ratnāvalī] is typical of small, embattled groups on the fringes or margins of dominant, established parent groups." (Schopen 2005: 9)
This suggests that Drewes is playing down the antagonism between Mahāyāna and the Mainstream. It seems clear that Mahāyāna believers did co-exist with Mainstream Buddhists, but they did not necessarily co-exist without tensions and conflicts.

The Role of Texts

It seems that in trying to understand Mahāyāna we have placed too much emphasis on the proliferation of texts. Too many assumptions were made about the conditions under which a religious group might produce and transmit new texts. Here, again, we can point to the influence of Protestantism. To our Eurocentric minds, the production of new texts must be preceded by schism and must represent irreconcilable differences. But this assumption does not apply in India, generally. The history of Indian religions is very different from the history of Christianity in Europe, especially as seen through Protestant eyes. Buddhist India was far more pluralistic than Christian Europe; more tolerant of heterodoxy, though polemics do survive; and more likely to syncretise. 

For some time it seemed that Schopen was right to say that "each text placed itself at the centre of its own cult" (cited in Drewes 2010a: 59). However, Drewes calls this into question. Firstly, there was no evidence of distinct Mahāyāna communities. Mahāyāna existed within the Mainstream institutional framework. Some have pointed to the divergent doctrinal views of Mahāyāna texts as evidence requiring distinct cults, but Drewes counters that accepting the authenticity of texts with divergent points of views is not a problem today and there is no evidence that it ever was. Drewes suggests that the different sūtras probably reflect the ideas of different authors rather than distinct communities.

So, Mahāyāna was primarily a literary movement, operating within and alongside mainstream Buddhism. It was unlikely to have been a unified movement.

We now have good evidence in the form of the old Aṣṭasāhasrikā manuscript described by Karashima and Falk that the first Prajñāpāramitā texts were composed and/or compiled in Gandhāra in the local language, Gāndhārī (see Karashima 2013).
    It is fairly certain, however, that writing was in use before the development of Mahāyāna. Drewes notes one Mainstream text with a 2σ range 184-46 BCE and another with 2σ range 206 BCE - 59 CE (2010a: 60). The mid points of these ranges are 110 BCE and 146 BCE respectively. But, on the whole, it seems that Mahāyāna textual practices were not different from Mainstream practices and Drewes notes that the distinction between categories like "oral" and "written" is not hard and fast in India. Even written texts are memorised and studied orally.

    As Drewes points out, contra to a popular theory, there is no evidence that Mahāyāna Sūtras were initially composed in written form (2010a: 60). Karashima (2015a: 113) reinforces the point that the texts were most likely composed orally in Prakrit. He proposes a rough time line:
    1. Oral transmission in Prakrit. 1st Century BCE.
    2. Oral transmission in Prakrit. Written Prakrit in Kharoṣṭhī script. 1st~3rd centuries CE.
    3. Broken Sanskrit mixed with Prakrit. 2nd~3rd centuries CE.
    4. (Buddhist) Sanskrit. Written in Brāhmī script. 3rd/4th centuries onwards.
    Drewes places the translation into Sanskrit about a century later than Karashima, i.e., 4th/5th century.

    To sound a contrary note, I have to point out that propagating a literature is generally a community activity. Written texts require a medium, ink, and implements, all of which suggest an economy in which such things were either produced or could be bought. For oral texts to survive for any length of time, they must be memorised by more than one person at a time. But such communities could have existed as cliques within monasteries.

    Forest Dwelling Bodhisatvas

    Paul Harrison (1992) and Reginald Ray(1994) independently floated the idea that forest dwelling ascetics were a significant influence on the development of Buddhism, generally, and Mahāyāna, specifically. The idea seems to have caught on and many scholars have found textual support for this thesis, not least Jan Nattier in her 1993 book A Few Good Men. Indeed, if the heart of the Mahāyāna was in forest renunciants, then this would explain the lack of inscriptional evidence (though for the same reason it is inconsistent with written texts).

    The forest renunciate is a Romantic figure, or at least a focus for Romantic projections, both for their mode of life and for the location of it in wild nature. They are saintly, dedicated to religious practices, especially self-denial, and is associated in many cultures with sacredness. Research has shown that such personal sacrifices are important in encouraging the faith of ordinary people (see Martyrs Maketh the Religion). Thus, the forest renunciate appeals to the Romantic aspect of modernist world view and self view.

    However, as Drewes remarks, "The main problem with the forest hypothesis is that Mahāyāna sūtras, the final court for any theory of early Mahāyāna, provide little support for it." (2010a 61). Of course, the texts do mention forest dwelling, but it is hardly the sine qua non of Buddhist practice in most early Mahāyāna texts. Some texts, e.g., the Aṣṭasāhasrikā, even discourage it! The majority of Mahāyāna texts seem to be concerned with easy practices that enable one to get out of saṃsāra with the minimum of effort.

    Ray turns out to have used a very narrow selection of texts to justify his thesis (and those only in translation). He excludes the large majority of the texts that have been translated, let alone of those which are preserved in Sanskrit, Chinese and Tibetan. But what is worse is that the texts he does cite frequently undermine his thesis on closer examination. For example, he cites the Ratna-guṇasamcayagāthā as advocating forest dwelling, when, in fact, it explicitly discourages it! (Drewes 2010a: 62). Jan Nattier's forest dwelling thesis is also, I hate to say, built on shaky ground. It is based on one text, the Ugraparipṛcchā Sūtra, that is admittedly very early, but "advocates forest dwelling and monasticism inconsistently" (Drewes 2010a: 62). Her other contribution focuses on Akṣobhya's pure land, but here also she seems to overlook the ease with which practitioners are promised entry to Abhirati. "Nattier's general idea that earlier forms of Mahāyāna advocated difficult, jātaka-like practices and that easy means of practice were developed only later has no obvious evidentiary support" (Drewes 2010a: 62).

    Clearly, forest dwelling played a continuing part in Buddhism. It is evidently an important practice for Mainstream Buddhists, but Mahāyāna texts are equivocal about the benefits. They seem to prefer other kinds of practice; often much easier practices.

    The Bodhisatva Ideal.

    The earlier models of Mahāyāna Buddhism had a break-away group who rejected the arhat ideal in favour of the new bodhisatva ideal. We've seen that the break-away thesis is wrong, that the arhat ideal was not rejected in all early Mahāyāna texts. And, in fact, there is no strong evidence that the bodhisatva ideal was particularly influential in Mahāyāna. We also know that the bodhisatva ideal was not missing from the early Buddhist texts. But a number of other characteristics distinguish Mahāyāna texts from Mainstream texts:
    • expanded cosmologies and mythical histories
    • pure lands
    • 'celestial' Buddhas and bodhisatvas
    • descriptions of powerful new religious practices
    • new ideas on the nature of the Buddha
    • a range of new philosophical perspectives
    There's nothing in the actual texts to suggest that the bodhisatva ideal was either the cause of the others, or that it was more prominent than the other characteristics (Drewes 2010b: 66-67). So it seems the focus on the bodhisatva idea is a retrospective emphasis, and the insistence that it was not found in Mainstream Buddhism is a straw man argument, a mistake or a disingenuous piece of misinformation.


    A generation of scholarship has transformed our understanding of the origins and early development of Mahāyāna Buddhism. However, that scholarly understanding has yet to fully permeate the Buddhist community. Certainly when I got involved in Buddhism in the early 1990s the standard narrative was still basically the 19th century one. I believe that it survives largely intact. Part of the reason might be that it strikes at Buddhist self-views and identity building narratives. My understanding of religious belief is that these are the beliefs that are most resistant to counterfactual arguments. Judging by how reluctantly Buddhists have received the knowledge that the Heart Sutra is an apocryphal Chinese text, I suspect that it will be some time before Buddhists catch up with the academy on this, if they ever do.

    A significant drawback in Drewes' articles is that they only relate the etic view, i.e., the views of European and American scholars. They tell us little or nothing of the emic, i.e., of what Buddhists themselves thought about Mahāyāna. I think it likely that etic views were formed in part by the normative stories told by Buddhists themselves. The views held by Western scholars were almost certainly informed by existing Buddhist narratives. It would be useful to know more about the process of view forming amongst these early scholars and the extent to which they were simply repeating what Buddhists themselves believed about their own history.

    It is quite significant if the Buddhist normative stories are at odds with the actual history that we can derive from textual and archaeological studies. My sense is that many of the false stories about the origins of the Mahāyāna are promoted by the sectarian followers of Mahāyāna in order to bolster their own prestige. Buddhists often seem to see themselves as being in a competition to present (and represent) the most "authentic" or most "authoritative" version of Buddhism. Or else they are justifying their own heterodoxy. Many of these historical narratives are dismissive of Mainstream Buddhism, which in light of the actual history now seems bizarre. And the competitive side of Buddhists is still evident in the present.

    This new historical paradigm may well shed new light on old debates about the sectarian affiliations of the some more prominent Mahāyāna śāstra writers, like Nāgārjuna and Vasubandhu. The idea that these writers could have a foot in both camps no longer seems odd. Assigning them to either Mainstream or Mahāyāna might be to misunderstand where their loyalties lay. It is only our perception of sectarian divides that make us struggle to place a figure Nāgārjuna, who both cites āgama texts and uses Mahāyāna ideas, like śūnyatā. As monk, Nāgārjuna can only have been ordained in a Mainstream lineage, because that was the only kind of Buddhist ordination. Perhaps it was entirely natural at the time to have loyalty to a conservative ordination lineage and an innovative textual tradition at the same time.

    Drewes argues against the use of the term Mainstream Buddhism, largely because different scholars have used it in widely varying ways. He suggests non-Mahāyāna, but I disagree. It would be better to have a positive term for what was, after all, the mainstream of Indian Buddhism for a millennium, and to seek a consensus on how it is used. Defining the mainstream in terms of not being part of a minority movement seems perverse. In light of this new picture that has emerged, "Mainstream" seems the best candidate yet as a term to contrast the Mahāyāna tendency in Buddhism over a period of many centuries. We need to be clear that this is a term for discussing the history of Buddhism in India. It does not apply to groups outside India, nor beyond the demise of Indian Buddhism, though, arguably, now that there are millions of new (Ambedkarite) Buddhists in India, it is once again relevant. So the term cannot apply to modern Buddhism movements. Nor is it a fixed term. The mainstream of Buddhism in India was changing all the time; the point that Drewes makes is that the mainstream was gradually taken over by Mahāyāna Buddhism. For an example of this kind of change see my JBE article on how karma changes over time (Attwood 2014).

    I've noted a few points of discussion and disagreement. These ought not to distract from the overview provided by Drewes. Overall, these articles are an important contribution to our understanding of the history of Buddhism. The articles benefit from being well written and organised. Although many scholars contributed to the change in worldview, to see all that work expertly summarised is quite an experience. It brings with it that frisson that the true intellectual feels when they experience a paradigm shift. One's worldview does not simply adjust to the new knowledge, but the new knowledge restructures the world view. The case Drewes makes seems to sit on firm foundations and to completely supersede the legacy view of Mahāyāna. Personally I love it when this happens. Everything we thought we knew was wrong. Fantastic!



    Attwood, Jayarava. (2014) Escaping the Inescapable: Changes in Buddhist Karma. Journal of Buddhist Ethics 21: 503-535.

    Drewes, David.

    Seishi KARASHIMA

    Schopen, Gregory. (2005) Figments and Fragments of Mahayana Buddhism in India. University of Hawai'i Press. http://www.uhpress.hawaii.edu/books/schopen-figments-chap1.pdf

    See also Bhikkhu Bodhi et al., The Bodhisattva Ideal: Essays on the Emergence of Mahāyāna. Buddhist Publication Society, Kandy, 2013. [Reviewed by Dhīvan in the Western Buddhist Review. And seems to make many of the same points from a mostly Theravāda point of view.]
    Related Posts with Thumbnails