Text Theory

Textual theory, like interpretation, is a kind of explanation of phenomena-the text in our hands. Textual and semantic theory are thus intertwined. Textual theory is mainly the attempt to trace the text from the author's hand to ours while interpretive theory can be viewed as part of explaining the author’s writing of it. The "data" is the text in our hand. Textual theorists often portray textual theory as more basic, more certain and more objective (factual) but also logically prior to interpretive theory but most theories are speculative. We have no "laws" of textual transmission, forgery, mistaken copying etc. nor has anyone produced anything like systematic principles of adequacy of a textual theory. We certainly have minimal laws of thought but we do have a several theoretically respectable semantic theories and systematic argument for the principles of "charity" or "humanity" for interpretation.

It is easy to raise doubts about textual theories, especially when we consider the temptations for ancient forgery. After the fall of the book-burning Qin, you could do well by forging "ancient" texts. If 2000 years later we can notice features of a text, then a clever forger could have noticed it. Even if he did not notice it he could get the effect by merely copying and reordering fragments of other texts. Scribal errors, taboo characters, multiple versions and periodic scholarly attempts to rationalize them etc. all make the idea of completely unraveling the textual history look difficult enough that, with Nietzsche, we may be tempted to drop the "myth of the original text."

Consider the current “best” theory of the Zhuangzi (and most classical texts) which says that they were compile, edited, revised and shaped over hundreds of years by communities of scholars. These texts communities maintained and “updated” their texts to keep them abreast of other texts and philosophical developments. Obviously, no one knows how this “committee” editing took place, how many copies of a text were made and circulated etc. We assume the scholars sometimes defected from some textual communities and joined others bringing their knowledge with them. This is arguably how philosophical discourse took place in ancient China. Now with that picture in mind—what counts as the original text? What author’s intentions or mind is being reconstructed in our search for meaning?

It is merely hard to justify much confidence that we can ever identify an original of a text successfully. Further, in my experience, nearly all justifications of textual emendation theories appeal to some kind of interpretive evidence.  They appeal to a theory contained in other parts of a text or the belief expressed by some sentence of a text. So if the justification for changing the text comes from the meaning of the text, it can hardly be the case that we have to decide how to change the text before we can interpret it.  Textual theory should not be regarded as epistemically prior to interpretive theory. Of course, any interpretation is interpretation of a text with certain words and if words are deleted, added or exchanged it should normally result in a change in meaning.

In applying the principle of humanity, we need not assume that we are reconstructing some aspect of the process of arriving at the original. We theoretically adopt the status of the audience for a text. When the text committee revises a text, it does so in consideration of shared rules of the language, which they assume readers will follow, so that the revision will have such an effect on their commitments—beliefs and normative attitudes.  Even where we think a book is layered with parts written by several people, we can strive for coherence in our interpretive system by looking at the community context. We cannot (epistemically) try to mimic the mental state of the successive groups of people doing the editing. 

You have read Munro's comparison of Graham's and Liu Xiao-gan's textual theories. These are not the only rival theories, but they do converge in a helpful way with most theories of the text going back to the Guo Xiang hypothesis. That is, Zhuangzi himself authored only the first seven chapters. There are some dissenters from that point of view, as both Graham and Liu point out. The most interesting is Fu Sinian's suggestion that we separate out the Qiwulun. Graham once remarked that if the Qiwulun was written by someone different from the author of the rest of the book, we would still want to believe that author was Zhuangzi. In the end, we don’t know if Zhuangzi wrote any of the inner chapter nor how much editing they endured at the hands of the textual committees for hundreds of years.

Their agreement even extends to a rough division of the rest of the book. Chapters 8-11 28-31 to related school (Graham's primitivists and Yangists, Liu anarchists); 12-16 to Graham's syncretists (Liu's Huang-Lao) 17-22 and 32-33 to Liu's "transmitters" (Graham's School of Zhuangzi and syncretists). And both divisions follow broadly the model of Kuan Feng. But Graham cuts the text into much smaller shreds and redistributes a lot of it as well as incorporating some back into the inner chapters. They also disagree on dating.

They differ less in method. Liu accuses Graham of depending on interpretive theory (his reading of the "theme" of various sections) while Liu focuses on other linguistic evidence (use of compounds and poetry styles). At first glance, this seems unfair. Graham accompanies his very complicated argument with even more lists of the frequency of use of a whole batch of characters and phrases than Liu. Still, I share Liu's worry because Graham gives no general theory of why items he selects are significant. Liu's does rest on the widely held theory that

  developed compounds gradually from an original pattern of single character words. One is suspicious that similar lists could be assembled to "prove" widely different clustering theories.

So it does appear that the deep reason for Graham's clustering must be interpretive. Graham finds what he identifies as distinctive views in different chapters. That prompted his search for evidence such as frequency of use lists. Liu's approach seems to rely less on a prior theory of what theory is expressed in the sections he separates, but he does use interpretive theory for confirming evidence and for the identification of each "school" and for dating.

With regard to this use of doctrine, my main worry is that both seem to use what I call the "bag-of-doctrines" test. We say a X-ist is someone who holds beliefs A, B and C and a Y-ist subscribes to B, D and E. As interpreters of philosophy, we should want to know why someone who would believe A would be led also to believe B and C. Perhaps when we do, we will see how D and E are further elaborations or reflections of the point, or what new insight might lead to them. In other words, we don't want to see doctrines just as random collections of beliefs, but as coherent, rationally produced beliefs. The "grab-bag" approach may be justifiable for political parties (conservatives and liberals) or religions (different articles of faith) but when we think the objects of interpretation are philosophers we will want more than a mere list of shared doctrines. If we have such a view, then it would be more appropriate to use it for ordering views. We could explain "access" to a certain belief by showing that it would follow from what Zhuangzi taught him. Of course, not all the students of a philosopher correctly develop his philosophy. Philosophy does not inevitably progress.

That kind of analysis would also help us answer a larger question. How are these textual theories relevant for our interpretation of Zhuangzi? The image of Zhuangzi presented by his own school or transmitters is relevant as long as they are competent at applying its principles. Graham's Syncretism, for example, might be an implication of Zhuangzi's philosophy even if he did not dwell on it extensively. If that were so, then we could take the syncretist writings more seriously as faithful to the tradition. If Graham is right, Zhuangzi started as a Yangist and "had a conversion experience" that made him a Daoist. If that is right, then the Yangist chapters would be dangerous guides to the principles illuminating Zhuangzi's thought. If Zhuangzi were a student of Laozi, and if Laozi were a mystic, this would strengthen the case for a mystical interpretation of Zhuangzi. Ditto for relativism if Zhuangzi were a student of Hui Shi and Hui Shi is a relativist. Graham notes that Chapter 2 (Qiwulun) shows mastery of ancient Chinese "logic" (should be "semantics") that is lacking from all the outer chapters. This may explain their tendency to ridicule Hui Shi while Zhuangzi speaks affectionately of him as a philosophical interlocutor with whom he has disagreement.

As you see from the course outline, I will look at syncretist and School of Zhuangzi chapters to help identify the filiation and influences on Zhuangzi and to get the general thrust of his theory. I can make this coherent either by accepting Liu's affiliation and dating, or using Graham's and making the point I just did about the coherence of syncretism with Zhuangzi's principles. Ditto for any influence of Hui Shi and Chinese semantics. The point is that whatever interpretive stance we take, we have to make it coherent with some plausible textual theory-and make our textual theory plausible on interpretive grounds as well.