Heaven, Li, and the Formation of the Zuozhuan
A Taeko Brooks

Oriens Extremus v44 (2003/04) 51-100

Click to Return to Publications Page

This long study establishes five layers in the Dzwo Jwan and relates them to the general intellectual history of the 04th century, moving from a ritual to an ethical focus, and finally to a cosmological one. The paper is carefully developed to make the stages in the discovery of this structure as clear as possible. The key stages in that argument are:

1. The Dzwo Jwan, the largest of all Warring States texts, consists of very diverse material, from short notes on ritual matters to very long narratives of battle and political intrigue.

2. One promising key to understanding these differences within the work is the Dzwo Jwan's treatment of the role of Heaven in human affairs. Here too there is diversity. Some passages assign Heaven an ethical concern, and others regard it as ethically indifferent. Among the ethically concerned passages, there is again a contrast between Heaven's waiting to reward and punish individual behavior, and its intervening proactively on behalf of the future of the state. These differences more strongly suggest ideological diversity.

3. That diversity is strongly confirmed by the existence of passages which not only assume one of these viewpoints, but at the same time argue against another viewpoint. If we assume that an objection is later than the thing to which it objects, then we can form a series in which B objects to an earlier A, C in turn objects to B, and so on. Associating with each layer all passages of like character, whether or not they directly criticize the preceding layer, we arrive at a solution in terms of five Dzwo Jwan layers. The first of them, Sacrificial Heaven, or Heaven simply as an object of sacrifice, is scarcely present. Passive Heaven is ethically responsive to the deeds of men; Active Heaven does not wait for men to act, but may act on its own, typically with a focus on the state rather than the individual; Transitional Heaven is a seeming halfway point between this and the next, and Natural Heaven is Heaven without any ethical connections, simply doing what it does. The names of the layers, with representative passages, are as follows:

Sample Passage
Sacrificial Heaven
Passive Ethical Heaven
Active Ethical Heaven
Transitional Heaven
Natural Heaven


4. Putting this result aside for the moment, we may next examine all passages containing the word Li ("Ritual"). Here again, we find ideological differences, and we also find that some Li passages explicitly criticize, or implicitly reject, viewpoints found in other Li passages. Grouping all Li passages according to these expressed differences, and arranging them according to the order in which their positions are successively criticized, we arrive at a five-layer model for the text in its Li aspect. Spirit Li passages are concerned with Li in the context of ancestral sacrifices, Human Li legislates rather for human conduct in general, Governmental Li is concerned with the structure of the state, Disputed Li contains passages where the speakers in a story may be reticent or openly uncertain about the nature of Li, and Cosmic Li concerns the order of nature without regard for the order of humankind. These Li layers, with sample passages, are as follows:

Sample Passage
Spirit Li
Human Li
Governmental Li
Disputed Li
Cosmic Li


5. Some passages contain both Heaven and Li, and in none of them does the order implied by one hypothesis violate the order implied by the other hypothesis. That is, the two hypotheses, though separately arrived at, are structurally compatible. Moreover, the philosophical outlook implied by a Heaven layer is manifestly compatible with that implied by the corresponding Li layer. Then it appears that our two hypotheses are really only aspects of a single hypothesis about the ideological evolution of the authors of the Dzwo Jwan. The combined five-layer hypothesis looks like this:

SacrificialSpirit or Ancestral
Passive EthicalHuman
Active EthicalGovernmental


The large movement on both sides of this table is from the auguristic-sacrificial world to a human-centered ethical world, and thence to a world where the state, not the individual, is the object of concern of Heaven (or the area for which Li legislates), to an uncertain or transitional phase, and finally ending in a cosmological order in which the doings of man have no particular place: they are not the object of the motions of Heaven or that which is determined by the prescriptions of Li.

This internal compatibility is the first strong argument in favor of the hypothesis.

A Typical Yi Hexagram

6. In support of this preliminary picture, further developmental sequences are identified in the Dzwo Jwan: (1) the usage of the word hwo (normally "someone," but sometimes also "perhaps"); (2) the relative age of hexagram analyses versus trigram analysis in Yi divination passages; and (2) the three terms for the so-called "hegemon," namely mvngju, hou-bwo, and ba. Of these, the expectations are the following: (1) it is intuitively likely that the more abstract usage "perhaps" postdates the more specific "someone" (notice the cognate primary pronoun mwo "no one"); (2) Hellmut Wilhelm and other students of the Yi have concluded that trigram analysis, though seemingly simpler, is actually a late development within the history of Yi interpretation; and (3) my own earlier study of the "hegemon" terms concluded that there were three successive versions of this construct (all of them in the minds of the Dzwo Jwan writers, since the Chun/Chyou text itself does not support the idea of a formal hegemon institution), each using by preference a different one of the three terms, in the order mvngju, Hou-bwo, and ba. These sequences are then those by which the order of passages forced by the above hypothesis may most appropriately be judged.

7. For all three developmental sequences, the predicted order is also the actual order generated by the Dzwo Jwan layer model here proposed. This independent developmental identity is the second strong argument in favor of the hypothesis.

8. Eric Henry's data on the "gentleman" versus the "Confucius" passages in the Dzwo Jwan offers a further opportunity to test the proposed model. It turns out that the distribution of these terms according to the model is highly suggestive: the "gentleman" passages occur throughout the text, but the first two layers here proposed include only "gentleman" passages, which would seem to be the earlier usage, whereas the "Confucius" passages occur only in, and predominate in, the later layers. Here too, the proposed model seems capable of shedding further light on a theory which has already proved its utility.

9. The earliest layer of the Dzwo Jwan is probably defined by the characteristically short passages in which the Dzwo Jwan asks why a given event was recorded (written; shu) or not recorded (bushu). These attempts to detect the conventions of the Chun/Chyou imply a commentary still trying to understand its text, and differ greatly from the more openly philosophical, and literarily expansive, passages which are typical of the later layers. The implication is that the Dzwo Jwan enterprise began as an attempt to read the ritual implications of the Chun/Chyou text, and progressed only later to political theory.

The Chinese Constellations

10. The final layer of the Dzwo Jwan is marked not only by philosophical changes from previous layers, but by several traits which point to a Chi rather than a Lu context. These include the addition of Chi Hwan-gung to the list of recognized hegemons, predictions of the Chvn (later Tyen) family's success as usurpers in Chi, thus in effect legitimizing that usurpation, as Lu theorists would be unlikely to do; and traces of what may be Dzou Yen's astral/terrestrial correspondence theories, these (as Shr Ji 74 tells us) being regarded at the time as typical products of Chi thought. This anchors, not the whole of the text, but rather the last layer of it, where many observers on other grounds (including linguistic and astronomical) had previously put it, namely, toward the end of the 04th century.

11. Finally, the intellectual developments implied by the Dzwo Jwan layer theory are compared with the whole course of 04th century intellectual evolution, as seen in successively later layers of the other major texts of that period: the Gwandz, the Mwodz, and the middle Analects. The result of that comparison is that the Dzwo Jwan development turns out to closely parallel the developments in the other texts, according to previous studies of the dating of their individual chapters by Rickett (Gwandz) and by myself (Mwodz, Analects). Not only, then, are the Heaven and Li trajectories within Dzwo Jwan compatible with each other, and thus aspects of a unitary development of thought in the Dzwo Jwan authors, but it also appears that that development is in turn part of the course of eastern Chinese thought in general, over almost the whole course of the 04th century. This fit of the Dzwo Jwan model within a previously arrived at model for early thought at large, a model emerging from the researches of more than one person, is the third strong argument in favor of the hypothesis.

12. I will end by quoting the final two paragraphs of the article, under the heading "Conclusion:"

To continue in this way would be to write an intellectual history of the 04th century. In the space here available, I hope at least to have given grounds for thinking that the leading ideas in that century emerged logically, rooted in social changes which themselves occurred in a historically intelligible sequence, and that the various theoretical texts of the period took note of them, in various ways and with varying resistance, but in the same sequence. I suggest in conclusion that the Dzwo Jwan is itself an 04th century theoretical text - one in which basic ideas such as the relation between Heaven and man are continually modified in response to a changing political context, and in which the nature and direction of history itself is explored. By the end of its growth, the Dzwo Jwan has become the first work of Chinese historiography, attempting to render the past intelligible, and to prescribe effectively for the future.

The Dzwo Jwan has always been admired for its narrative facility. I hope that it may also come to be respected as a uniquely continuous record of reflections on history, documenting, as no other text of the period does in equal detail, the emergence of the first fully mature Chinese view of cosmos and society - a view that was to be definitive for the Empire which lay less than a century in the future.

A Taeko Brooks

Back to Articles Index Page

1 Sept 2006 / Contact The Project / Exit to Publications Page