Toward a Digital Research Environment for Buddhist Studies

Toward a Digital Research Environment for Buddhist Studies
Nagasaki, Kiyonori , International Institute for Digital Humanities, nagasaki@dhii.jp
Tomabechi, Toru, International Institute for Digital Humanities, tomabechi@dhii.jp
Shimoda, Masahiro, University of Tokyo, shimoda@l.u-tokyo.ac.jp
In recent times, digital resources have taken on steadily greater importance in the field of Buddhist studies, with increasing numbers of digitized versions of Buddhist canonical texts, representation of material culture, and other objects of research becoming available on Web. However, despite the basic availability of such resources, most of them are not set up in an optimal way for usage by researchers; nor are they for the most part integrated with each other. For example, there does not yet exist a system that can operate with equal efficacy with philological data related to Sanskrit, Chinese, Tibetan language materials. Therefore, a comprehensive and concrete framework is needed. Although a method of text description is fairly well established in the form of TEI P5, neither the interfaces, nor the methods of presentation of results for digitized works are as yet satisfactory for the scholars of Buddhism. In this paper, we will present our approach to the establishment of requirements for various kinds of materials used in Buddhist studies and make some suggestions for the implementation of more functional interfaces as a Web research environment for such scholars.

At first, it must be noted that it is very difficult to define adequate requirements for the full range of the scholars of Buddhism, who come from a broad array of language training and methodological approaches. Thus, this paper will focus primarily on the fulfillment of the requirements for the scholars who are dealing with authentic scholarly digitized texts. In the field of Buddhist studies, where texts have been translated across a number of languages in different regions at various points in history, we have no recourse but to deal with several versions of a text, including transmitted, diffused, and translated variations at the same time (Fig.1).

(Fig. 1. An Example: “ मूलमध्यमककारिका (Madhyamakakārikā)" and some of the related texts)

Full Size Image

As discussed by Steinkellner (1988), it is often difficult to "recover" the original form of any given text, as they have been changed variously in their long tradition. In many cases, all that has come down to us is a translation that was preserved in the Chinese tradition since the 2nd century or in the Tibetan tradition since the 8th century. It is quite often the case that various witnesses are extant in both traditions. In such cases, various diplomatic texts in various languages must be compared in various units such as at the level of text, chapter, fascicle, sentence, word, syllable, or even character. Therefore, it is necessary for textual scholars to prepare an environment that delivers integrated views of a given text views. On the other hand, it is important to understand the background thought and beliefs reflected in each stage of the textual development—not only in Sanskrit and Pāli which are the closest to the original form—but also in Tibetan, Chinese. In this case, each annotation must be recorded in the context of its diffusion into other texts. In addition, modern translations of each text should be pointed to from such texts. Therefore, it is crucial for Buddhist textual studies to provide such information within a system of intertextual relationships. (See Fig. 1)
In order to realize such relationship in digitized resources, it is necessary to provide the data in a punctuation-neutral manner. This is because separation of words or sentences is not always obvious; punctuating text itself amounts to an independent and original scholarly contribution, especially in the cases of Sanskrit and classical Chinese manuscripts. Therefore, the basic concepts are:

being based on a unit that is not restricted to the legacy media.
inheritance of the legacy studies on the paper media.
DB providers prepare the space for the sharing of data units.
the users act as recipients of the units and some of them act as distributors of the units.
DB providers develop and distribute their own Web API so that users can make arbitrary links to access each others' data.
The above are realized as a collaborative research environment on the Web.
In addition, in order to preserve compatibility with past research results, all units should be identifiable in legacy media through traditional referencing methods such as T0001,01,0001a01 which means Taishō Daizōkyō, vol.1, page 1, register a (among a, b, and c), line 1. Then, they should be located by URI by implementation of Web API so that the other persons or applications can freely refer to arbitrary units through Web.

In order to realize such concepts, a collaborative editing system needs to be developed. Initially, users should be registered according to their own roles such as visitor, editor, or administrative editor. The role of the editor is that of inputting and checking of the data; the role of administrative editor is checking of the data and the determination of the distribution of the data. Both editor and administrative editor should be able to efficiently select arbitrary units with an easy method such as mouse click or drag on the text body without any restriction posed by the textual structure. Then they can append any necessary information to the units and link selected units to other ones (Fig. 2).

(Fig. 2. The procedures of inputting relationship data on an environment)

Full Size Image

In the final step, the administrative editor should check and determine whether each unit is ready for release. The role of visitor is only to browse the workspace to observe the progress of the work. On the other hand, anonymous users can browse only released data. They can efficiently view various information with various methods such as mouse click or drag on the text body and select (or ignore) any data on the basis of various properties such as their sources, contributors, and so on. In order to realize such a function on Web, AJAX will be fully utilized.
Finally, we will explain the Web Database function. The Web Database stores only relational data that includes one or two locations in the original materials, its annotation or relationship information, contributor's name, data composition and other related information. The writing rule of the location allows the user to use the information even without digitized textual material. However, if the material is originally in digitized form, the data must refer to a logical location such as the URI. Users don't need to be aware of separation between the database and the materials presented to them because to the viewer of the materials, they seem to be seamlessly integrated with the database. By this kind of method, it will be easier to integrate the other materials which are released on other Web sites into the environment. The database can provide arbitrary data according to the user's preference through Web API, as well as retrieve and show the data provided by other DBs. From here, the data can be published under various formats such as RDF. However, in principle, this kind of system would not be able to avoid the problem of overlap if there is an intent to publish its data fully including the original materials. So we are trying to do so according to TEI P5 by using a kind of stand-off markup.

(Fig. 3. The procedures of browsing the relational data on an environment)

Full Size Image

Although we have just begun to develop this approach, it has already been tested by some scholars and we have seen positive results. Actually, these early testers say that it is very useful for them because their accessibility and permissions are limited and explicitly shown. However, some enhancements have been suggested, which mainly concern the function of browsing. We will work on this until the time of DH2011, when we will able to present a much more matured system.

References:
Burnard, L. Bauman, S. 2007 P5: Guidelines for Electronic Text Encoding and Interchange, (link)

Caton, P. 2007 “Distributed Multivalent Encoding, ” Digital Humanities 2007, University of Illinois, Urbana-Champaign, 2-8 June, 2007 33-34

DeRose, S. 2004 “Overlap: A Review and a Horse, ” Extreme Markup Languages 2004, Montreal, 2-6 Aug., 2004 (link)

Lavagnino, J. 2009 “Access, ” Literary and Linguistic Computing, 24 (1) 63-76

Nagasaki, K. 2008 “A Collaboration System for the Philology of the Buddhist Study , ” Digital Humanities 2008, Oulu, Finland, 25-29 June, 2008 262-263

Rehm, G. Witt, A. 2008 “Aspects of Sustainability in Digital Humanities , ” Digital Humanities 2008, Oulu, Finland, 25-29 June, 2008 21-29

Renear, A. H. 2004 “Text Encoding, ” A Companion to Digital Humanities, Schreibman, S. Siemens, R. Unsworth, J. Blackwell Oxford 218-239

Steinkellner, E. 1988 “Methodological Remarks On The Constitution Of Sanskrit Texts From The Buddhist Pramāṇa-Tradition, ” Wiener Zeitschrift für die Kunde Südasiens, Band XXXII, 103-129