Graphic Form of Characters

For the graphic form of characters in paleographic texts, an important question is how can the rich amount of information embedded in the graphic form of a character be accurately reflected and conveniently accessed in digital form? Here, by “graphic form,” we refer to the calligraphic tendencies of a character, the composition of the strokes, and the positioning of the graphic components, generally called the pianpang 偏旁. Such information, to be sure, is available from the photographic reproduction, but when it comes to a digital database, the question is how this information should be inputted and extracted. In many ways, this is a question that overlaps with the matter of the transcription of paleographic texts. The most important issue at stake is what is the standard adopted when a paleographic character is being inputted into a database? As for the transcription by the original editors, should this be retained, for the sake of consistency and the completeness of the record, even when there is clear evidence that it is imprecise or even mistaken? Or should one try to come up with a new transcription that then serves as the basis of the database? To look into the matter a bit more deeply, the transcription of a character is often inseparable from one's understanding of the variations of a character, in both diachronic and synchronic (e.g. trans-regional) terms; it is questionable whether it can be resolved in the single-step process that is the transcription, let alone dictate the production of a database. Such an understanding will ultimately lead a serious scholar back to the beginning of his or her reflection: newly excavated texts are by and large manuscripts, written by hand, and this is the feature that distinguishes them from the printed texts that we have become accustomed to.

In response to these questions, the various institutions strive for a method for conducting multi-level searches, and also to gradually develop a unified standard for the transcription. In terms of multi-level searches, Wuhan University has led the way in offering the feature of searching by graphic components. As noted by the scholars at that institution, this feature is especially crucial for those cases where one recognizes only one part of a character and not the rest. By accessing in a systematic manner all the characters featuring a particular graphic component, paleographers can identify whole series of characters, sometimes phonetic series. To address the same problem, the database of the manuscripts from the Shanghai Museum, created by Fudan University, has an index that itemizes every character and all the components. Such features can also be found in the database developed by East China Normal University.Footnote ⁴

In terms of setting a unified standard for the transcription of characters, Academia Sinica has for some time pushed for the creation of a database of character components, or Hanzi gouxing ziliao ku 漢字構形資料庫, and its database of bone and shell inscriptions shows the correspondence between the inscriptions and transcribed characters. In contrast, East China Normal University, with its principle of complete digitization, assigns a code to every character, which can then be inputted using a method devised by the same institution. Finally, Fudan University has a new idea for how to search for secondary sources, namely, by employing a unified and standardized transcription prepared by the architects. Together, these efforts raise the question of whether a single, practical, and plausible standard could be developed by the architects of these databases, so ultimately there would be a unified list of characters as well as an input method that transcends institutional boundaries.Footnote ⁵

Finally, it is possible to mention a feature developed by the Chinese University of Hong Kong, an input method by handwriting. Imagine if one wrote the paleographic form of a character, and upon input, received a series of characters that resemble one another, however “resemblance” is defined. The result would not only provide information relevant to the problem of decipherment at hand, but also draw attention to other aspects critical for the recognition of a character. To be more specific, if characters that resemble each other in terms of the calligraphy, stroke, and component could be digitally grouped in a particular manner, then such series of characters with similar or identical variation patterns could provide new ways for understanding the development of the writing system in general.Footnote ⁶

Relation between Word and Character

How do the characters in paleographic texts record the spoken words of Old Chinese? This is the most fundamental as well as the most difficult question in the study of newly excavated texts. In its database of the Shanghai Museum manuscripts, Fudan University provides a table indicating the correspondence of characters and words, or Zici duiying shuju biao 字詞對應數據表, and by allowing multi-functional searches, makes possible the building of a network of closely related characters and words.Footnote ⁷ In connection with this, the Fudan database indicates for each character the Old Chinese initial and rhyme group, following the reconstructions by Wang Li 王力 (1900–1986), and this is also a feature of the database created by East China Normal University.Footnote ⁸ It is not difficult to see that if closely or identically sounding words could be obtained through database searches, this would lead to unexpected results in the decipherment of a character. Related to this, one might raise the more daring question of whether it might ever be possible to search by semantic meaning, so words with close meanings could be linked?Footnote ⁹ If nothing else, this would allow one to observe phonological connections that might otherwise go unnoticed.

When considering the question of the relation between word and character, it is important to note that a text sometimes has parallels with other texts, whether this means that it is quoted by another text or that two texts, to whatever extent, have overlapping content, vocabulary, or syntactic form. These correspondences are valuable sources of textual variants, allowing for comparisons on several levels. Touching on the complex issue of how to compare newly excavated texts and the transmitted literary record, these parallels are a reminder of how texts can vary across time and space and from one transmitter to another, thus providing an important resource for the discussion of textual transmission. It is not inconceivable that if such data were linked digitally, then the relevant information could be made searchable in an effective and comprehensive manner. On this point, the database by the Chinese University of Hong Kong provides the following feature: for a newly excavated text, if there is a relevant transmitted source, then this is cited and compared, resulting in emendations (indicated by a different color font), critical notes, or tables summarizing the differences. In addition, the Chinese University has published several reference works that compile these quoted and cited sources with respect to particular ancient texts, in a series entitled the Xian Qin Liang Han dianji yin jingdian ziliao huibian 先秦兩漢典籍引經典資料彙編. On this basis, one wonders if statistical methods or other tools could be utilized to come up with a more precise definition for parallel texts, which would in turn allow for the identification of additional relevant materials?

Context

The paleographic research of the past has generally focused on the decipherment of single characters. The type of reference work that best reflects this interest is the index of graphic form, or wenzi bian 文字編 (and the concordances that are comparatively fewer in number). Due to the growing variety and complexity of newly excavated texts, particularly literary and philosophical texts, scholars sometimes propose readings and interpretations that vary widely from one another. The context for placing these paleographic sources is increasingly becoming a challenge, and it has a direct impact on using the new evidence to make certain claims about language, thought, culture, and other aspects of Early China. Here “context” can refer to a bamboo slip, a whole text, various documents from the hand of a single scribe, a single tomb, or it can be a web or interpretative framework based on a particular researcher's understanding of language, thought, culture, and other historical backgrounds. How should these different “contexts” be reflected so they could meet the needs of scholars—this is a question that has no ready answer.

Through their databases, different institutions respond to this question in different ways. One response is to link a search result with the photograph or rubbing of the corresponding object, thus reproducing the context as it is reflected in that impression. This is the case for the Chinese University of Hong Kong in its database of the bone and shell inscriptions, bronze vessel inscriptions, and bamboo slips and silk. It is also true for the series of databases designed by Academia Sinica, which include not only the database of bone and shell inscriptions under discussion, but also a database for bronze inscriptions of the Shang and Zhou. In its database, East China Normal University divides the various levels of transcription and suggested reading, and it takes into account the correspondences as well as possible discrepancies between the paleographic text and later decipherments. It has also published a book series that gathers the various opinions of paleographic research, character by character, called the Gu wenzi kaoshi tiyao zonglan 古文字考釋提要總覽, with a corresponding searchable online database. Finally, Fudan University has a bibliography of secondary studies that links the relevant scholarship to a corresponding character of a paleographic text; and in its transcription and reading, it aims to incorporate the best and latest research.

Web Forums and Publication

Given the frantic pace at which newly excavated texts are coming to light, it is little surprise that more and more scholars are opting for web forums as opposed to print publication for disseminating their research. Rather than waiting for months or years for an article to come into print, one can share one's ideas almost instantaneously. Among the contents featured on these forums are research notes; previously published papers; announcements of projects, conferences, and other scholarly events; and publication notices. To name only some of the most active web forums:

• • Fudan University, Center for Research on Chinese Excavated Classics and Paleography (http://www.gwz.fudan.edu.cn/)

• • Wuhan University “Bamboo and Silk Manuscripts” (http://www.bsm.org.cn/)

• • Tsinghua University, Unearthed Research and Protection Center (http://www.tsinghua.edu.cn/publish/cetrp/)

• • Chinese Academy of Social Sciences, Institute of History (http://www.xianqin.org/)

• • “Bamboo Silk” (http://www.jianbo.org/)

• • “Confucius 2000” (http://www.confucius2000.com/)

At the forum hosted by Fudan University, it is not uncommon for a research note to be followed by readers' comments, rejoinders, and surrejoinders. This is not the case for every forum, though there is a great deal of cross-referencing among the research notes, an indication of the high level of frequency at which they are consulted. While some of the forums regularly bundle together their contents—both Wuhan University and Fudan University gather the research notes and make them available as compact discs or electronic files—a question that naturally arises is whether the forums would in the future introduce a peer-review process, which is ultimately what separates these online discussions from print publications. (For some of the institutions that already publish printed journals, the question can be put somewhat differently, as to whether they would integrate the forums with their journals.)Footnote ¹⁰ As long as the refereed articles remained distinct from other discussions taking place online, there is no risk of their conflicting with one another. The advantages would of course be accessibility for all, the saving of production costs, and the possibility that an article, once published, could provide the basis for further debates and discussions, a living document, so to speak.Footnote ¹¹

The major question with the web forums as they currently operate is of course their impact on scholarship. While the discussions often give the impression of diversity, perhaps even cacophony, they also have the paradoxical effect of impeding truly innovative ideas. Just as one idea is quickly echoed and variously referenced so it assumes the appearance of accepted consensus, another idea is submerged and forgotten, never receiving the consideration that it rightly or not rightly deserves. In this regard, what the sociologist James A. Evans suggests about online journals could just as well be true for web forums. Basing his research on a database of 34 million articles, their citations, and online availability, Evans argues that an increase in online availability actually has the adverse effect such that “the articles referenced tended to be more recent, fewer journals and articles were cited, and more of those citations were to fewer journals and articles.” According to Evans, “Searching online is more efficient and following hyperlinks quickly puts researchers in touch with prevailing opinion, but this may accelerate consensus and narrow the range of findings and ideas built upon.”Footnote ¹² While the situation for the study of newly excavated texts is somewhat unusual, with the web forums, journals, and the large number of electronic books circulating on the Internet occupying different levels of scholarly discourse, it is not immune from the problems described by Evans, and the implications are well worth pondering.

Potentials of the “Wiki” Approach

The success of Wikipedia and its popularity among students and the general audience have led many to consider more seriously the possibility of an open, collaborative approach to scholarship, particularly the preparation of such reference works as the encyclopedia.Footnote ¹³ Is this approach possible for the study of newly excavated texts? Although it is understandable that the architect of a database might not wish to open his or her work to just about everyone—Cheng Shaoxuan of Fudan University mentions the possibility of a vouching process that would allow certain scholars the privilege of modifying the database developed by Fudan—there are more fruitful areas where a collaborative approach might be attempted. One possibility is the compilation of bibliographies. As noted in the preceding section, publication has now reached such a frantic pace that original research is published daily as monographs, journal articles, chapters in edited volumes, and research notes at web forums (including the readers' comments) all clamor for one's attention. It is difficult for one person or even a group of scholars to keep track of all these, let alone read everything; and the problem is exacerbated as one moves beyond one's native language. A collaborative approach has the potential to solve many of these problems; in fact, it might be the only practical solution. With regard to the problem of vandalism that is often mentioned as a potential risk of collaborative approach, this has already been discussed in connection with Wikipedia and can easily be overcome by such simple measures as requiring registration only by one's real name and maintaining a certain level of administrative oversight.

Once a bibliography has been established, it can be organized according to the various principles of library and information science, and retrieved by searchable options, including keywords and subject headings. At the same time, from the user's perspective, if one could have the option of attaching a certain label (in the form of keywords or short phrases) to individual bibliographical items, a practice known as “tagging,” then this might open up even greater possibilities, one of which would be to reveal all kinds of interesting trends about how scholars receive and respond to the works of their peers.

A collaborative approach has great potential also in terms of pedagogy. One can envision such an approach for course syllabi, lecture notes, textbooks, and other areas where consensus and mainstream opinions are valued over individual innovation. While beginning students are not expected to know the full range of opinions on a certain question, they should have a sense of the most important and most historically interesting. This is where the input of others can be beneficial, and the result, as the product of collective labor, would demystify a field which many outsiders regard as arcane, out of date, and tirelessly quarrelsome, and present it in a more favorable light. Note that such an approach is fundamentally different from the designing of websites, which, like museum exhibitions, can reach and touch the general audience in ways unimaginable by the printed word. Commendable as such projects are (and many do make the effort to be balanced and even interactive), their approach is ultimately a top-down, one-directional creative process.

Come to think of it, why would the Wiki approach not work for databases? It is without question that a database, like all other reference works and systems of knowledge, requires structure. But if such structure could be laid down in a way that permits the participation of its users, then this would only benefit the field as a whole. Thus, for a problematic character in a newly excavated text, it is no longer one or several scholars who are responsible for noting the large number of readings that have been proposed; everyone can chip in. As long as there is a mechanism to distinguish the different opinions, so the view of the original editors is differentiated from that of the architect, and in turn from the other proposals, then the work avoids the risk of alienating any of its users. The result would resemble the type of reference work, the gulin 詁林 or jishi 集釋 that scholars have become accustomed to for the study of paleographic texts, but it would be an ever-expanding one, with an endless list of addenda already incorporated into the database itself.Footnote ¹⁴

Future Prospects

For the present discussion as well as the study of newly excavated texts in general, the focus is and always will be on the reading of texts. While the written record of language cannot be examined in isolation from archaeology, art history, and the materiality of the writing surfaces, among other concerns, the priorities are beyond questioning. For a time when new approaches and methodologies are constantly being advanced and debated, with the field flooded by an ever-expanding body of data, we believe this point is worth emphasizing. As one looks on to a new age, one must never forget that the digital resources are intended to reinforce and supplement the traditional approaches, never to replace them.

Here it is helpful to look to history and recall that the reproduction, management, and processing of ancient textual sources has been an ongoing project, traceable, at least in the modern period, to the Sinological Index Series published under the direction of William Hung (Hong Ye 洪業, 1893–1980). From 1931 to 1950, Hung and his colleagues at the Harvard-Yenching Institute engaged in the itemization of the linguistic data of seventy-seven ancient works, basing their work on the critical texts that they had established and making the efficient and comprehensive use of these sources a reality.Footnote ¹⁵ Such endeavors, as Hung explains in his essay “On Indexing” (Yinde shuo 引得說), were not unprecedented and followed in the footsteps of the scholars of a previous generation.Footnote ¹⁶ In fact, if one included dictionaries, rime books, encyclopedias, bibliographies, and various guides and reference works, all of which reflect the diverse ways in which traditional scholarship confronted the seemingly boundless ocean that is the transmitted literary record, then the effort to effectively manage large quantities of textual information may go back even farther, and it can perhaps be understood as a recurring theme in the possession, preservation, editing, sorting and even censoring of texts. Such is the paradox that in attempting to manage information, we end up creating more of it. Such creations, like all textual creations, say as much about a literary tradition as they do about ourselves.

So where do we go from here?

There are, to be sure, many problems and challenges presented by the new resources, and some of these, such as the “narrowing” of scholarship, we alluded to above. In our view, what is most significant about these new developments is the following. If, in the past, we have relied on authorities and standard accounts, we will continue to do so, for that is how we acquire knowledge and make sense of the world. But the new resources also suggest a conception of scholarship and much else that is fundamentally different from what we are used to. Rather than a field dominated by a handful of towering figures, we now have multiple voices; rather than one character, one reading, we now have a dozen proposals, maybe more, for how the text should be understood; rather than a stable corpus, we now have an ever-expanding body of data; rather than a fixed standard text, we now have cross-referencing links pointing to widely diverging interests and agenda; rather than a center of research based in Beijing, Paris, Kyoto, or Berkeley, we now have scholars, graduate students, and amateurs working from all over the world. As much as old habits will endure and traditional approaches continue to be upheld—and we emphasize again that we do not advocate that they should be abandoned—it seems that these efforts now take place within a context that is ultimately diverse, multi-polar, democratic, and non-hierarchical, with the participation of the whole scholarly community. Any claim for authority and final answer will be as much a textual creation as the databases and other resources described here. Instead of asking what the correct reading of a character is, perhaps we could also ask what are the conditions that give rise to the multiplicity of meaning, the rationale underlying each proposal, the personalities behind them, and the various historical contexts in which they are embedded. If this implies anarchy, it also means that we will look more carefully at differences, potentials, and the spaces between the lines. The mode of inquiry that results is more open-ended, and scholarship becomes a process whereby each person, as an individual, makes his or her contribution to the community at large. For a discipline so steeped in tradition, and in a world overshadowed by authoritarianism in various forms and guises, such an understanding, as with open access, freedom of speech, and other related causes, is ultimately to empower the individual and ensure a more humane way of life, and it is worth striving for.