Chuang-tzŭ illustrates to perfection the kind of battering which a text may suffer between being written in one language and being transferred to another at the other end of the world some two thousand years later. In the first place ancient Chinese thinkers did not write books, they jotted down sayings, verses, stories, thoughts and by the third century BC composed essays, on bamboo strips which were tied together in sheets and rolled up in scrolls. A chapter of Chuang-tzŭ would have originated as an item or collection of items making up a single scroll. Collections of scrolls ascribed on good or bad authority to one author or school grew up gradually and did not assume a standard form until Liu Hsiang (77–6 BC) edited them for the Imperial library of the Han dynasty. From his reports on the books, some of which survive, we learn that each copy available to him would have a different number of scrolls, and his method was to collate them and cut out duplicates. We do not have his report on Chuang-tzŭ, but according to the bibliographical chapter of the Han history the Imperial library copy had 52 chapters. Our present Chuang-tzŭ is an abridgement in 33 chapters by Kuo Hsiang (died AD 312), who may be suspected of having tacked the more interesting parts of discarded chapters on to the ones which he retained. The book as we have it from him is divided into three parts.
1 INNER CHAPTERS (chapters 1–7)
Each consists of discontinuous episodes grouped round a common theme which is summed up in a three-word chapter title. This series is homogeneous in thought and style and generally recognised as substantially the work of Chuang-tzŭ himself.
(2) OUTER CHAPTERS (chapters 8–22)
None can be plausibly ascribed to Chuang-tzŭ. These have two-word titles which are mere labels, taken from words in the first sentence. They comprise:
(a) Four complete essays (chapters 8–10 and the first part of chapter 11) by an author idiosyncratic in thought and style whom we call the ‘Primitivist’ and date about 205 BC.
(b) Three chapters related by their titles, ‘Heaven and earth’, ‘Way of Heaven’ and ‘Circuits of Heaven’ (chapters 12–14), each starting with an exposition of ideas which we class as ‘Syncretist’ and believe to be those of the editors of the book, probably in the second century BC. But from the Primitivist chapter 11 as far as chapter 14 only the introductory passages are homogeneous; each chapter has been filled out with all kinds of miscellaneous material. Thus the last item of chapter 11 is Syncretist, of chapter 12 Primitivist.
(c) Two complete essays (chapters 15 and 16), the first Syncretist, the next unrelated to anything elsewhere in the book.
(d) Six chapters (chapters 17–22) in which the editor seems to be trying to group materials of multiple authorship around the same themes as in the Inner chapters. For this, as for much else in the book, we have no better label than ‘School of Chuang-tzŭ’.
(3) MIXED CHAPTERS (chapters 23–33)
Like the Outer chapters these have two-word titles which (except in chapters 28–31) are taken from the opening sentence. They are:
(a) Five ‘ragbag’ chapters (chapters 23–27), quite heterogeneous, much so badly fragmented as to suggest that they have been assembled from broken or misplaced strips in other scrolls. Some of this material looks like Chuang-tzŭ’s own writing, and bits of it can be fitted with varying degrees of plausibility into mutilated parts of the Inner chapters.
(b) The block we call the ‘Yangist miscellany’ (chapters 28–31), distinguished by chapter titles summing up their content. Su Shih (AD 1036–1101) already saw that they are not the work of Chuang-tzŭ, and a modern scholar, Kuan Feng, has pointed out that they are not even Taoist, and that most (in my own opinion, all) come from another school, that of Yang Chu. They probably date from a little before and after 200 BC.
(c) The collection ends with another ragbag chapter (chapter 32) and another Syncretist essay (chapter 33) with a title related to those of chapters 12–14, ‘Below in the empire’ (T’ien-hsia, literally ‘Below Heaven’). The titles are from the opening sentences as in chapters 8–27. Why have these been separated off and put at the end of the book? Probably because chapter 32 ends with a story about the death of Chuang-tzŭ, while chapter 33 is a general description and evaluation of the pre-Han philosophers down to Chuang-tzŭ, suitable as a conclusion to the book.
The themes of Chuang-tzŭ’s own Inner chapters are summed up in their titles:
Chapter 1, ‘Going rambling without a destination’ – the joy of soaring above the realm of conventional judgements and practical concerns;
Chapter 2, ‘The sorting which evens things out’ – criticism of analytic thinking in terms of right and wrong alternatives;
Chapter 3, ‘What matters in the nurture of life’ – the knack of living spontaneously;
Chapter 4, ‘Worldly business among men’ – living in the world without compromising with it;
Chapter 5, ‘The signs of fullness of Power’ – evaluating the Power in a man without regard for conventional opinion;
Chapter 6, ‘The teacher who is the ultimate ancestor’ – reconciliation with the loss of individual identity in death;
Chapter 7, ‘Responding to emperors and kings’ – the government of the empire.
In the Chinese text these are the only chapters which have titles of not two but three words, resembling those of the Confucian apocrypha (wei-shu) of the Former Han dynasty (from 206 BC to AD 24). The terms ‘even things out’, the ‘ultimate ancestor’ and ‘emperors and kings’ do not occur in the Inner chapters themselves, only in the latest, the Syncretist stratum of the book. Probably Chuang-tzŭ left behind only disjointed pieces, mixed up perhaps with his disciples’ records of his oral teaching, and it was a Syncretist editor of the second century BC who devised the headings, grouped the relics under them, and relegated the unusable bits to the Mixed chapters. In the last of the Inner chapters, centred on a theme in which Chuang-tzŭ was hardly interested, the government of the empire, one has an especially strong impression, not of an author approaching his topic from different directions, but of an editor going to great pains to find even remotely relevant passages.99
We can at least say that we have the Chinese text of Chuang-tzŭ very much as it was about AD 300, when it was already becoming difficult to understand. Kuo Hsiang, who at that time prepared the abridgement of the text which survives, also wrote the first commentary extant in full. This and other annotations and commentaries written between the third and seventh centuries AD provide explanations of varying plausibility for the many words which had already become obsolete, together with philosophical comments inevitably tinged with the neo-Taoism or Buddhism of their own times. Studied in Chinese with these traditional aids, the book has been something of a lucky dip, in which magnificent episodes mingle with obscure or unintelligible fragments for which the commentators have nothing to offer us but guesses. Further progress requires the full apparatus of modern scholarship, which Chinese, Japanese and Western specialists have only recently begun to deploy, to distinguish the different strata in the writing, restore the corrupt or dislocated text, add to the deplorably little yet known about the grammar of the classical language, make sense of the philosophical terminology. We need also to relate the book to the other philosophical literature of the age. For example, to understand Chuang-tzŭ’s criticisms of disputation we have to explore the only surviving manual of disputation, the Mohist Canons, which textually is in even worse shape than Chuang-tzŭ itself. There are similar difficulties of course with all those ancient writers, Western and Eastern, who miraculously remain living voices although their words survive only in mutilated texts in imperfectly understood languages, like the statues in museums which are still living presences in spite of their broken noses and stumps of arms. However, sinology in the West has a much shorter and sparser tradition than Greek or Hebrew studies, and it will be a long time before we can share the confidence of translators of Homer or Genesis that the preliminary problems, if not solved, have at least been explored up to the limits of present knowledge.
How should we set about turning this extraordinary hotchpotch into English? There are several complete versions, three of them by the best translators of Classical Chinese of their respective generations, James Legge, H. A. Giles and Burton Watson.100 Giles’s version, first published in 1889, has a place on the margins of literary history, as a minor influence on English and American writers from Oscar Wilde to Henry Miller.101 Watson’s, which appeared in 1968, is admirable for his consistent treatment of the main philosophical terms as well as for his deftness in picking the apt and vivid word. All have the limitation that for the most part they follow the traditional commentators without coming to grips with the outstanding textual, linguistic and philosophical problems. This conservatism has not prevented them from doing justice to some of Chuang-tzŭ’s best passages, but all three come to grief on a single basic error of policy. They treat Chuang-tzŭ as though it were what is nowadays understood by a ‘book’, and present it as written in prose and divided into chapters composed of paragraphs; and they assume that, however disjointed, mutilated, even frankly unintelligible the original may be, however much its parts may differ in date, in thought and style, it is their duty to trudge forward from sentence to sentence, disguising the breaks, blurring the differences, assimilating the verse to the prose, in order to sustain the illusion of a smooth flow. The trouble with this approach is not simply that it forces the translator to break down now and then and write nonsense, but that it throws him into a defensive stance which corrupts his whole style. He finds himself slipping against his will into a bland evasive English which does not commit him as to whether any one sentence follows on to the last—an English which no serious, let alone great artist could ever write. Any literary skill he may possess will become positively dangerous, enabling him to weave sense and nonsense into a seamless robe. A quite eerie effect is that the smoother the English the more Chuang-tzŭ will assume the persona of someone who could have written that English—intermittently lively, more often verbiage, expressing even at its most coherent incompatible opinions from 200 years of Chinese intellectual history. It is in the best translations that Chuang-tzŭ suffers a strange mutation into a whimsical, garrulous old wiseacre to whose ramblings you listen with half an ear in the confidence that every now and then he will startle you awake with a vivid phrase, a striking aphorism or a marvellous story. But this image of the great Taoist, at once affectionate and profoundly insulting, has no relation to Chuang-tzŭ or any other writer in the book, no relation to anything except the situation of a translator cracking under the multiple strains of his craft.
The problem is how to break out of the crippling convention of complete, sentence-by-sentence prose translation. Arthur Waley, the greatest of translators from Chinese, and the only one who always knew what he was doing, was a lover of Chuang-tzŭ who had the good sense to offer only carefully selected extracts embedded in exposition, in his Three ways of thought in ancient China. If we are to risk something more ambitious, it is our duty to remember that the ideal of integral translation is in this case meaningless, and that the danger is that the more of the book we think we are translating the less of it will be conveyed.
In the present venture I start from the following decisions:
(1) It is pointless to offer integral translations except of homogeneous blocks in the corpus, such as the Inner chapters and the Primitivist and Yangist sequences; but these must be presented complete. So must the other chapters which are single essays (chapters 15, 16 and 33), and any episode extracted from chapters which are miscellanea; everything translated must have its entire identifiable context.
(2) The reader must be informed of what is specific to the thought of Chuang-tzŭ, the Primitivist, the Yangists, and the Syncretists, so that he can distinguish them and find his way around them.
(3) Only the chapters which are true essays (such as the Primitivist’s) should be treated as consecutive paragraphed prose. The Inner chapters are collections of isolated episodes probably grouped together by a later editor, and including, for example, sequences of rhymed quatrains, stories in which speakers may burst into song, didactic verses with scattered prose comments, strings of aphorisms, provisional formulations of ideas followed by criticisms, propositions which Chuang-tzŭ (or a disciple perhaps) proceeds to annotate phrase by phrase. Each requires a corresponding form in English, with a typographic layout suited to its structure.
(4) Prose must be translated as prose and verse as verse. It may seem surprising that all the translators except Waley neglect this surely self-evident principle. However, in a Chinese text verse is not distinguished from prose by its layout on the page, and the formal differences between them are a matter of degree and are obscured by, for example, difficulties in identifying rhymes in the ancient pronunciation. Might it not after all be sufficient to render rhymed passages as poetic prose? In the not infrequent cases of sentences of uneven length which are rhymed but by every other criterion are prose, this defence is quite reasonable. But we also have verse interludes in the four-word measures of the old Book of songs, which interrupt the linear development of the thought to allow interactions between parallel members of a couplet or between corresponding lines in successive quatrains. Unless spaced out line by line these will appear as soggy patches in the prose where Chuang-tzŭ seems to have forgotten what he is talking about. Rhyme as such is not the essential test, for the same applies to strings of parallelised aphorisms which differ from didactic verses only in being unrhymed.
(5) The order of the episodes in the Inner chapters and the other homogeneous blocks can for the most part be left as it is. But it is not sacred, since there is no reason to suppose that Chuang-tzŭ ever did put his jottings in a definitive order. The occasional passages which break recognisable continuities may be moved to more suitable contexts. We may also use Mixed chapter fragments which seem on internal grounds to be Chuang-tzŭ’s to fill gaps in the Inner chapters. All such transpositions will be identified.102
(6) With the mass of varied materials in the Outer and Mixed chapters which in the present state of scholarship can only be classed as ‘School of Chuang-tzŭ’, the most convenient policy is to select and group by topic. Every episode chosen must be presented complete, as must such cycles as the tales about Chuang-tzŭ and about the meetings of Confucius and Old Tan. Otherwise, I recognise no duty to supply even easily translatable episodes which do not seem to me to add to the philosophical or literary value of the book. In theory it would be possible to reproduce and regroup all the items, concluding with a collection of scraps decorated with scholarly question marks and rows of dots; but to go on adding to the translated passages brings diminishing returns. A reader curious to know what I have left out (which amounts to a fifth of the book, and since tastes differ may well include some people’s favourite passages) will probably also want to check my rearrangements, and will be better served by the existing versions than by an expanded translation on the present lines.
(7) The ideal version would, like the original, have items which are delightful and illuminating at first reading, and others which are elliptical, difficult, enigmatic, to be skipped or to be wrestled with in the light of introduction and notes. But it would never amble between sense and nonsense in that inconsequential meander to a faultering rhythm which should always be taken as a warning that the translator is losing his grip, yet is so easily mistaken for the mysterious workings of the Oriental mind. Not that I am confident of having attained that ideal myself, for some of the worst problems are raised by the Inner chapters, which I am committed to translate entire. I offer in advance my apologies to the ghost of Chuang-tzŭ.
There remains one obstinate problem for which only compromise solutions are possible. How is one to do equal justice to Chuang-tzŭ as a philosopher and as a poet? Most versions show a bias towards one side or the other. A primarily literary translator (such as Giles or Watson) will probably have some liking for the Taoist view of life but also a Taoist distaste for the analysis of concepts, without which he cannot select and manipulate his English equivalents effectively. More intellectual translators (such as Legge, or the great historian of Chinese philosophy Fung Yu-lan, who published a version of the Inner chapters) are inclined to neglect the literary aspect as though it were mere decoration of the ideas. But a Taoist is a thinker who despises thoughts, yet values, and finds the imagery and rhythm to convey, any spontaneously emerging process of thinking which he senses is orienting him in the direction of the Way. My own private final test of whether translation is really working is whether it catches any of the extraordinary rhythmic energy of Chuang-tzŭ’s writing, not merely for the lift of the heart which it gives but because to lose it falsifies the pace and shifts and stresses of his thinking.
In the Chinese original the thinker and the poet are one. But in the effort to make him speak our language we are constantly faced with a choice between the ugly expression and the inexact. If we are to come to grips with the key terms of Taoist vocabulary we need regular equivalents, which will often sound awkward, and will assume their full meaning only after being experienced repeatedly in context. In ‘The sorting which evens things out’ one may translate, as in non-philosophical contexts, shih pi by ‘this’ and ‘that’, shih fei by ‘right’ and ‘wrong’, and the English will flow very smoothly (that precious sparing of the nerves from the shock of jarring words for which translators are tempted to sacrifice every other value); but the argument, which turns on the single word shih, will drop out of sight. Having vowed not to give the impression that Chuang-tzŭ writes mellifluous nonsense, I render shih pi by ‘it’ and ‘other’ and shih fei by ‘That’s it’ and ‘That’s not’. These expressions are not easy to manipulate, especially since Chuang-tzŭ appears to make a distinction, crucial throughout the chapter, between two kinds of shih – yin shih (‘going by circumstance and shih-ing’) and wei shih (‘deeming and shih-ing’). These I am reduced to rendering ‘the “That’s it” which goes by circumstance’ and ‘the “That’s it” which deems’. Here the coherence of the conceptual structure has to be respected at any cost. But the cost is considerable, for Chuang-tzŭ is writing as a poet just as much here as anywhere else. The translator of a complex text is a juggler with a dozen balls to keep in simultaneous flight, and some of them are always bouncing on the floor.