22 December 2017

What is a Text Anyway?

I often find myself having to stop and reflect on the process and methods I use to explore texts. There is no ideology of method in Buddhist studies; we simply do whatever we think is best at the time. Indeed, there is very little discussion of "Theory" in the sense that presently dominates other subjects in the humanities. So, as I decipher texts, I have to keep asking myself, "What do I have to assume in order for this approach to allow me to make valid inferences, and are those assumptions themselves valid?" Writing my thoughts down is often the best way to organise them. 

My task at present is to identify instances of the Chinese phrase  三世諸佛 in Prajñāpāramitā texts and compare parallels in the extant Sanskrit sources. The phrase means "all the buddhas of the three times" (Three 三 time 世 all 諸 buddha 佛). My hypothesis is that Sanskrit sources will always have atīta-anāgata-pratyutpannā buddhāḥ "past, future, and present buddhas". This is important because the Heart Sutra has "tryadhvavyavasthitā sarvabuddhāḥ" or "all the buddhas appearing in the three times". So far, I can say with confidence that this way of phrasing it does not occur in any of the Sanskrit editions of earlier Prajñāpāramitā texts. In the jargon it is a hapax legomenon, or a one-off (or perhaps a neologism). 

If my hypothesis is accurate, then this phrase in the Heart Sutra can only be a Sanskrit translation from Chinese. This would prove beyond any doubt that the Heart Sutra was composed in Chinese. To be clear, it is already beyond any reasonable doubt that this is accurate. I think I can make it certain. And this is important because some scholars are on the fence as regards the Chinese origins thesis, and some irrationally reject it. I hope to entice the fence-sitters down to earth and to leave the rejectors no wiggle room. It's part of my elaborate homage to the originator of the Chinese origins thesis, Jan Nattier (3 articles published, 1 submitted for peer-review, and 3 more planned). 

One Text or Many?

However, this task is far less straightforward than it might seem at face value. Take, for example, the first occurrence of the phrase in the Aṣṭasāhasrikā-prajñāpāramitā-sūtra (Aṣṭa; "Perfect Gnosis in 8000 Lines"). It occurs in a speech that Maitreya gives in praise of transference of merit. In the translation produced by Kumārajīva’s group in about 408 CE, (aka  "T227") it begins at T 8.548a17 (Vol 8 of the Taishō Edition of the Chinese Tripiṭaka, page 548, panel a, line 17). 

It is easy enough to locate the same speech in Vaidya’s Sanskrit edition (72) [Page 127 of Conze's translation]. However, Vaidya's text is approximately three times as long as in T227. In the translation by Xuánzàng (T 7.791c29), the speech is almost twice as long again. Although we don’t have a Gāndhārī text of this chapter, we know that, in general, it is considerably less prolix than the Pala Era (8th - 12th Century) manuscripts used by Vaidya for his edition. Also, there are seven Chinese translations in total (though only one other is of any great interest). This raises the question of whether "parallel" is even the right term. 

There is, in effect, no single text of Aṣṭa. Each instance of Aṣṭa is unique. And we need to be very cautious about thinking of the Sanskrit text as “original”. In fact, the extant Sanskrit documents certainly do not constitute an “original” for Kumārajīva’s translation, but date from perhaps 400-500 years later. We no longer have Kumārajīva's source document (thought it probably was a Sanskrit translation). 

The prolixity of Xuánzàng's text may be an artifact of his translation process, which often includes an element of auto-commentary: i.e., the text commenting on itself.. Where we can compare the late 1st Century Gāndhārī manuscript we often see that the Sanskrit has expanded one adjective or verb with up to five or six synonyms. For example, where Gāndhārī might have, "speak"; the Sanskrit texts might have, "speak, teach, instruct, draw-out, reveal, illuminate". I think we can see this as a form of auto-commentary. 

This raises some philosophical questions. If Aṣṭa is in fact many texts, in many languages, and specific to particular practice communities in particular places, then are my methods sound? Because, in common with other scholars, I tend to assume a unitary text with at most minor variations that can easily be eliminated. But the variations here are huge and cannot be ignored. And, if it was officially translated, then it was authoritative to someone.


The basic approach of philology grew out of Bible Studies. As 18th and 19th Century European imperialists looted the world, they came across, amongst other things, very old manuscripts of the Bible that had differences from the received text in Europe. I imagine this must have been unsettling for Theologians at the time. Methods were developed to identify the ur-text and restore the Word of God (phew!). Those methods became the main tools of philology and provide Buddhist Studies with some of our most important tools. We, too, spend time collecting and assessing documents, creating critical editions (restoring the original or ur-text), and doing "higher criticism" based on this "original"

But if what I've been saying about Aṣṭa is true, then the possibility of reconstructing the ur-text is doubtful at best. We know from analysis of phonetic transcriptions of some words and names in the earlier Chinese translations and from the existence of a single badly damaged manuscript--carbon dated to ca. 70 CE--that the oldest Prajñapāramitā texts we know of were in Gāndhārī. In other words in the vernacular language of Gandhāra, a kingdom that roughly equated to the Peshawar region in modern-day Pakistan, though it extended over the Hindu Kush and into what is now Afghanistan, as well (hence the Bamiyan Buddhas). Mahāyāna texts began to be translated into Sanskrit around the 4th century CE. The oldest Sanskrit manuscript of Aṣṭa, by contrast, is from the mid 9th Century (it's held in the Cambridge University Library and I have seen the actual object). In other words it is 800-odd years removed from the origins. Lokakṣema's Chinese translation was produced in 179 CE from a Gāndhārī source text. Might it not have a greater claim to authenticity than a 9th Century copy of a 4th Century Sanskrit translation? 

Recently, one of the more inspirational professors of Buddhist Studies, Jonathan Silk, has raised many related issues. He has argued that a critical edition is, in fact, a new composition. The idea that a critical text represents the ur-text is simply a fantasy. The reconstructed text may not even be in the correct language. This is the case with the Heart Sutra. It was composed in Chinese, using a quote from Kumārajīva's translation of the Perfect Gnosis in 25000 Lines. Because it contains a quote, even the Chinese text is not a true ur-text because we can go several steps further back - right back to the Gāndhārī text of Aṣṭa. Other parts of the Heart Sutra may have been composed in the late 7th Century, but they show distinct influence from the idiom of Kumārajīva. 

Lest we think that the prajñāpāramitā literature is a special case, let me assure you that it applies to all Buddhist texts, including the Pāḷi texts. There is every reason to believe that the texts were not composed in Pāḷi but are translations. When we look at the Chinese translations of counterparts of Pāḷi texts they show the same trend, albeit to a lesser extent, as the Mahāyāna texts - later translations are more elaborate and edited for consistency (see, for example, my rough translations of the Chinese Spiral Path texts in the Madhyama-āgama). Consider also that, as yet, there is no proper critical edition of the Pāḷi Canon. The editions we have were from a very small number of documents. Alex Wynn is heading a project to produce such an edition, but the host monastery in Thailand, Wat Dhammakāya, is embroiled in multiple scandals, so will it have credibility? 

Mahāyāna texts were written down soon after being composed, or were perhaps even composed as written, as opposed to oral, texts. And, yet, they were dynamic, open to change, and seem to have grown over a period of some 6 or 7 centuries. Writing did not cause them to be fixed. Mahāyāna texts were anthologized (as in Śātideva's Śikṣasamuccaya) but they were not Canonised. Even the Pāḷi Canon underwent some post-canonical editing. 

Unlike the differences in the Bible, which were effectively an accumulation of mistakes and misreadings, Buddhist texts were actively changed and edited. And not always competently. Part of my work on the Heart Sutra will be repairing the damage done by inept ancient editors.

History of Ideas

For me, this impression is reinforced by the history of Buddhist ideas. We all know that Buddhist doctrines changed over time. Buddhists experimented with many varieties of worldview over time and now have such a broad range of views that any attempt to conceptually unite them necessarily fails. Buddhism has not been a single religion for about 2000 years now. We've evolved into distinct (incompatible) species. Of course we know that there were innovations, and some of us have adopted one or other innovation as our standard view, but we seldom get real sense of why there are variations. What drove Buddhists to innovate?

The obvious answer is dissatisfaction with the doctrines obtained from the earlier texts (aka The Pāḷi Canon). In other words, we have to ask: If the Pāli texts are so authoritative, why did all Buddhists (including Theravāda Buddhists) keep inventing new doctrines and changing the old ones? Often to the point of completely abandoning the earlier material (especially in Tibet and Japan).

While I was researching the history of the ideas of karma and rebirth, I identified a couple of probable reasons for innovation that I still have not seen in any books. My prime example is that karma requires consequences to manifest long after actions, while dependent arising denies that this is a possibility - "when this ceases, that ceases." Once the action has ceased, the possibility of a future consequence should entirely vanish.

Now, Buddhists had four choices: leave it incompatible; modify karma; modify dependent arising; or modify both. Initially they opted for modifying dependent arising in various ways (though all that survived was two variations on the kṣanavāda, or doctrine of momentariness, and the dvayasatyavāda, or doctrine of two truths.). However, later, when the karma doctrine no longer fit their needs, they also modified karma (On this see my 2014 article in the Journal of Buddhist Ethics). 

Over time, Buddhists changed things. They changed what seem like central doctrines and they changed seemingly sacred texts. Often they went down blind alleys. Modern Buddhists tend to have some knowledge that things changed, since most of us have read books from different traditions and experienced the confusion of terms and ideas. On the other hand, very few of us understand the dynamics that produced these changes. The "why" question is often left blank. A notable except is Ronald M. Davidson's Indian Esoteric Buddhism, which describes the socio-economic and geopolitical forces that may have contributed to the rise of Tantric Buddhism. 


I don't expect anyone will rush out and start learning multiple traditional languages (though Pāḷi is not so hard, and we have an excellent Pāḷi teacher in Cambridge). But I hope that we can move towards a more sophisticated view of texts. I'll finish with some key points

  • Buddhist texts have complex histories and change considerably over time.
  • With very few exceptions all the texts we have are translations.
  • Sanskrit is very seldom an "original" language. 
  • Chinese is far more important than we have so far grasped. 
  • Buddhist texts were always unsatisfactory to Buddhists.
  • This unsatisfactoriness was a major driver of doctrinal innovations. 
  • On-going doctrinal innovation is a feature of the history of ideas in Buddhism.
  • Innovation causes legitimation anxiety for Buddhists.
  • Each generation has adapted the Dharma to their needs, while claiming to be reinstating what the Buddha taught.
  • Pāḷi texts (or their analogues in Gāndhārī and Chinese) are perhaps the most important single source of arguments and disputes in the long history of Buddhism. 
  • There is, and can be, no Buddhist "Bible" nor indeed anything that we might call "Basic Buddhism". 

All the best for the solstice. 


Related Posts with Thumbnails