Classical Chinese Texts
 Analytical Procedure

This page supplements the Preface by giving a more detailed outline of the principles and working assumptions with which we have approached the complex problem of the Warring States texts. A fuller presentation of the principles of text analysis will be found in the Philology section of this site; this brief sketch is merely by way of orientation and first acquaintance.

Philological Method

In general, we have attempted to apply to classical Chinese materials the methods which have proved fruitful for the analysis and criticism of classical Western texts, but with suitable adjustments for the nature of the Chinese materials, and with appropriate extensions of technique for problems which have not so far been successfully addressed in either tradition. The chief areas of extension are in taking more systematic account of text growth, and in dealing more adequately with directionality problems, which occur during both the formation and corruption phases in the life of a text.

We have avoided the constraints upon interpretation which are imposed by traditional orthodoxy, with its agenda of Imperial justification, which have had a damping effect on research. We have also avoided all questions of "relevance" to contemporary issues, since those belong to a different world. We do not accept responsibility for the moral education of future generations. We ask only one question: What actually happened in the classical period?

Program of Analysis

Considerable experience with the material has suggested something like a standard procedure for studying an early Chinese text. Among the guidelines of this standard procedure are the following, given roughly in order of application:

1. Ignore the Author. The least firm fact about the typical Warring States text is the identity of its author. Titles can also be misleading. What we have is the text, and our investigation should begin with the text, leaving author questions for later. Arguments based on the supposed character of the author tend to be analytically ineffective. We more commonly find ourselves, at the end of an investigation, inferring the character of the author or authors from the text.

2. Test for Homogeneity. Given the possibility of composite origin or accretional growth, the first question to ask of a text is whether it appears to be homogeneous in style, grammar, and other linguistic features, and also in its content, assumptions, and general intellectual thrust. Once the pattern of similarity and difference is clear, our Text Typology list becomes useful in offering a range of models of text formation.

2a. If it IS Homogeneous, it probably has a single date. Ask what texts it is aware of, and what texts react to it. Such relationship patterns create a relative date. Sometimes it is possible to relate the text to an event, thus anchoring it to an absolute date. Whether the text relates to an event or to other texts, the question we try to answer is, Why was this written? What problem did it address? For what audience?

2b. If it is NOT Homogeneous, divide and retest until you get homogeneous segments. (Trying to date material that has not been previously examined for homogeneity is the commonest error of the older scholarship). Proceed with each of the resulting segments as in 2a, above. The first texts with which to compare each segment of text are the other segments of that text. Do they imply a rationale of addition over time? A principle of anthology arrangement? How did our text come about? What institutional continuity or advocational purpose can be discerned behind that process?

3. Determine Directionality. Contemporary texts or passages will often share an issue, or even a phrase or story embodying an approach to that issue. Parallel versions invite a hypothesis of indebtedness, but in which direction? Skill in directionality determinations leads eventually to the accurate relative dating of a large number of more or less contemporary documents.

4. Beware of Obliquity. The Warring States philosophical texts often credited their doctrines to earlier founding figures, or to other ancient authorities, real or imagined. In trying to seem plausible as ancient pronouncements, they usually avoided direct contemporary reference. This general fact introduces several complications:

4a. Antiquity Arguments. These were increasingly relied on as the WS period progressed. Implicit claims of superior antiquity ("Our founder is older than your founder") or even explicit ones ("Your antiquity is not antique enough," see MZ 48:4) occur in the WS literature. All antiquity statements have a rhetorical function, and all are suspect for precisely that reason.

4b. Archaisms. As time goes on, writers display increasing skill with prose, and in handling longer narratives. But any literary form or device, once it exists, becomes permanently available as part of the repertoire. Later writers seeking an antique effect may intentionally imitate the form or language of early texts. Length, form, and rhyme as such are thus not assured indicators of early date.

4c. Covert Allusions. Apart from legal or governmental works, open reference to, contemporary concerns is rare in this literature. Issues are often transposed into other terms and times. It can take some subtlety on the investigator's part to detect the real references, and to restore the text's connection with them.

5. Study the Corpus. The initial goal is to assemble a set of relative datings, a body of texts whose interrelations and sequence of composition are known, whether or not any one text can be dated to any particular year. This set of datings is the literary life of the Warring States, held (as it were) in chronological suspension. Each part of it illumines the nature of the other parts, and in principle, the whole can be better understood than any of its parts. As the relatively dated construct grows larger, it should eventually be possible to fit it as a whole into the framework of absolute chronology, as known from evidence outside the texts.

The final methodological caution is to save everything. A major error of the old scholarship was to discard inauthentic texts as unworthy of attention. It would be an equal fallacy to discard the later layers of some school tradition, and pay attention only to the oldest or core layers. All textual layers are of interest, for the period when they were written. All are part of history.

Concluding Appraisal

The above steps are meant to take account of the internal evidence. Once a hypothesis has been reached in this way, the remaining tests of that hypothesis are three:

In estimating plausibility, we rely on a general expectation that the movements of history, though they are not predictable in detail, are rational in tendency. Simple versions of an idea should precede elaborate ones; refutations should follow assertions. Developments do not begin for no reason, or run backward once they have begun, or have gaps of four hundred years in their middle. In general, a correctly described historical situation will be intelligible as a development.

The present solutions for individual texts are offered as the best that we have so far been able to arrive at, in satisfaction of these criteria.

Classical Chinese Texts is Copyright © 1993- by E Bruce and A Taeko Brooks

To Text Typology Page

Comments to The Project / Exit to CCT Index Page