A while ago I discussed how Janet Gyatso had objected to my approach of assuming authorial coherence and single authorship in my dissertation on Śāntideva (and in other works). I said there that “there’s an issue here much bigger than the interpretation of any one thinker: should one even try to find the coherent views of an individual author?” I answered yes and I stand by that. I remain firmly in agreement with Thomas Kuhn’s dictum that [w]hen reading the works of an important thinker” one should look for the apparent absurdities and ask how it could make sense that “a sensible person could have written them”. But I didn’t go back to what I implied was the “smaller” issue – which may not in fact be so small.
In the case of Śāntideva, the historical evidence suggests that his most famous work, the Bodhicaryāvatāra, is composite: there is a version of it discovered relatively recently at Dunhuang which seems to be significantly earlier than, and substantially different from, the version known to Indian and Tibetan Buddhist tradition. It would seem that the text as we know it is the work of at least two composers. And that, in turn, poses a problem for someone wanting to use Kuhn’s approach as he states it: what if this text is not the work of an important thinker or a sensible person, but multiple ones? Are we not then entitled to treat the text as incoherent because of all the different minds that went into it?
In the previous post I left this point behind, treated it as the smaller issue, because I was thinking of it as an issue in “the interpretation of one thinker”, namely Śāntideva. But of course a very similar problem applies to a great many thinkers one might study. In early Chinese philosophy, in particular, the work of nearly all of the “authors” we might identify – Confucius, Laozi, Mencius, Mozi, and so on – consists of a single text, whose title is often simply that author’s name. (The single extant work attributed to Mozi is entitled the Mozi.) But as far as I am aware, it is the consensus of historical research on these texts that many if not all of them are themselves composite, the work of multiple people. So careful scholars today will often speak not of “Zhuangzi” but of “the Zhuangzi“: the text, rather than the person, that is called by the name Zhuangzi.
I do not and have not done that here. When writing about Chinese tradition, or for that matter about Śāntideva, I continue to follow Kuhn’s method and look for coherent authorship, even on these texts that do not seem to be the work of a single writer. How can such an approach be justified?
Well, let me point back to the reasons I identified for looking for coherent authorship in the first place. The point of doing so is not primarily, if at all, a historical one. One can certainly read philosophical texts as a way to get an accurate picture of the past as it was. While such a task can have a philosophical purpose (especially for a Hegelian), it is not in itself a philosophical task. What a philosopher aims to learn from texts of the past is something that could be applicable to the present and future. Not merely reading one’s own presuppositions into the text and seeing them confirmed – in which case one hasn’t really learned anything – but being challenged by them, finding something in them that one didn’t already know or believe. It has been my experience that Kuhn’s stated method is an excellent, if not the best, way to do this. Those things that appear most absurd are often the things that, it later turns out, we did not yet understand about the text. For they often come in the form of unstated assumptions underlying the text’s explicit claims and arguments. And it is difficult to speak of assumptions without speaking of an author who holds them.
The author must then be then a device we postulate for explaining the text. I draw for inspiration here on Alexander Nehamas‘s excellent 1981 article “The postulated author: critical monism as a regulative ideal” (from Critical Inquiry 8(1)). For Nehamas, the author is a character we postulate to explain the text’s features, and not necessarily the text’s historical writer. Rather, the author as character is a hypothesis who “guides interpretation, and is in turn modified in its light.”
So how do we apply this approach, one that involves speaking in terms like “what the author was thinking”, when we are faced with a composite text? Well, in the cases at issue here, we can start with the fact that in the past and possibly the present, a long tradition did indeed view the text as the work of a single author. That’s certainly the case for the Chinese texts; in the case of Śāntideva there is a long tradition (in both India and Tibet) of viewing both the Bodhicaryāvatāra and the lesser-known Śikṣāsamuccaya as the works of a person called Śāntideva. In either case, people for hundreds of years have acted as if there is a single mind behind a given text, and that approach has been productive for them in understanding it. The lessons they have drawn from the text are the reason we’re reading it now in the first place.
In such cases, our hypothetical author would begin with as the single author postulated by tradition. Once we learn about earlier versions of the text, the mind of the author may become less important than the mind of the redactor, the one who assembled the text together and put it in its final attested form. For it is the redactor who really gives us the text as known to tradition.
In the case of ancient and classical texts, we’re often even justified using the author’s traditional name and applying it to the redactor. In Śāntideva’s case, he’s quite explicit in the Śikṣāsamuccaya about being a redactor: most of the text consists of quotations from sūtras whose origins are cited. Even with the Bodhicaryāvatāra, the older Dunhuang recension identifies its author as Akṣayamati, not Śāntideva, so the name “Śāntideva” may well be best applied to the later redactor. And with Chinese tradition, often all we really know about the historical writers is the texts attributed to them. So it may be fair to speak of “what Zhuangzi thinks” and by “Zhuangzi” mean “the redactor of the Zhuangzi“.
Some implications follow from this: most notably that then “the date of” such a text may be best treated as the date of redaction, which may be considerably later than the presumed date of original composition. I haven’t necessarily stuck to this practice, as when I date Chinese philosophers on my MediaKron site, and I probably should be more careful about that. But the point of it is: even with a composite text, we can speak of the text’s author in Kuhn’s sense as “an important thinker”, “an intelligent person” who composed a coherent work, and had a mind underlying the text with assumptions that make it coherent. For philosophers, that is usually the most productive way to approach a composite text.
Now this is not the only productive way for a philosopher to approach a composite text. If historical or philological criticism has been successful enough on a given text, especially, we might be able to excavate multiple important thinkers out of that text. We might be able to find conflicts between different authors who were later put together, and who can tell a different story than the redactor who tries to harmonize them. Then within the text we find multiple coherent worldviews instead of just one. But to do that successfully, our criticism of the text has to be very advanced, so that we can pick apart who composed the text at what stage. I think Biblical criticism has reached or come close to that point by now, able to pick apart the different layers in the text. But I don’t think that the study of ancient Chinese or Indian traditions generally has. Until it does, in my view, we’re better off with the redactor.