MED As An Experiment In Networked Scholarship

At the center of the Project is the idea that the Internet can be used as a powerful tool in scholarly research. Over and above the properties of nanosecond speed and massive storage that make computers enormously useful instruments for information processing, the Internet introduces qualities of information access and exchange that are particularly useful to scholars in the humanities and social sciences. To adapt an aphorism from theology, the Internet is a sphere whose center is everywhere and circumference is nowhere. Anyone with access to a terminal on the 'net has arrived at a "place" where information and people, though they be physically located thousands of miles away, are gathered, as if in the same building. Unlike the traditional library, inforamtion stored in this place is simultaneously and universally distributed. And unlike a conference or a seminar, the exchange of that infornmation is on-going and open to all.

The consequences of this new mode of communication are profound but perhaps difficult to envision. After all, the Internet has been around for twenty- five years and hasn't yet had a profound effect on the conduct of humanistic research. Nonetheless, as any observer of the 'net will point out, the full potential of the medium has yet to be fulfilled, and the time is upon us when a critical mass of participants will change all that. As is often pointed out, we are in the midst of a revolution in communication no less profound than that of the print revolution initiated by Gutenberg in the fifteenth century.

To see how the Internet might enhance epigraphic research, consider the following scenario. A student of the Classic Maya wishes to research a particular motif in the hieroglyphic texts, say the practice of auto- sacrifice. The student, who, let us say, attends a university in the United States, wishes to explore this motif as it appears in textual narratives. In order to pursue this research, the student will have to acquire as large a body of hieroglyphically inscribed artifacts (or their facsimiles) as she can, in order to explore the various patterns and contexts associated with the motif. The larger the body of information, the more clearly will emerge the patterns of association, substition and absence, and the more accurate will be her research.

At the outset of her research, our student encounters a problem. Depending on where she lives and studies, access to this information will be limited. Aside from the various books she has been able to buy and check out, she will have to go to other libraries, museums and collections--in Philadelphia, Austin, Cambridge, Washington DC, San Francisco, not to mention Mexico, Guatemala, and Europe--to complete her work. Of course, due to constraints of time and money, she will not be able to go to all of these places, but instead will go to some of them and make the best of what she can acquire.

After she has amassed her material, she is faced with another problem. Because she is not yet an expert epigrapher, she would like to ground her research in the work of authorities in the field. Sensitive to the complexity of epigraphic decipherment, translation and interpretation, she would like to know not only what gloss an epigrapher has provided for each block in a text, but the precise interpretive moves and decisions that led to the gloss. To what extent did the epigrapher apply the structural method of Berlin, Proskouriakoff and Kelly? What Thompson numbers were associated with each glyph? What phonetic or semantic values were assigned to each glyph? How were these values selected and combined to produce the words and sentences attributed to the inscription?

Here our student encounters a serious problem. The published material contains precious few examples of translations that detail the stages of glyphic identification and evaluation. Except for a few sources, like the Research Reports on Ancient Maya Writing (Center for Maya Research) and site reports like the Excavations at Seibal (Peabody Museum, Harvard), the translations available to her are incomplete. To remedy this, she limits her body of material to those texts that she can account for, either from published sources or by contacting epigraphers with specific questions.

But here she encounters another difficulty. She finds that epigraphers disagree on their transcriptions of the same text. Disagreements are of course to be expected, and indeed applauded, but she finds that occassionally there are discrepancies at the level of identification. That is, a given glyph may be classified as a token of two or more glyph types. (Roughly speaking, a glyph type corresponds to a glyph number in Thompson's scheme.) To remedy this problem, she creates, from the material at her disposal, a set of standard transcriptions that contain the varying identifications.

After having limited her material to those texts and images that meet her criteria, our student then produces a set of interpretations and, based on these, arrives at a set of hypotheses to account for the meaning of the motif she is studying. At this point she wishes to submit her ideas to the criticism of others working in the field. Again, she encounters limitations. Desiring to augment the criticism of her advisors, who may not be specialists in her subject, she must wait for a conference or journal to present her work, or for the responses of those to whom she has sent it. These things take time and, unless she is in close contact with the specialists whose evaluations she desires--those individuals able to provide that precious resource unassumingly referred to as "personal communication"--they do not always succeed in allowing for the kind of immediate dialogue that is so important for the advancement of scholarly research.

In the end, after going through one or two rounds of criticism by these methods, our student produces her study, after which it is distributed to those who may be interested by it, to those libraries that will accept it, or, if her work is good enough and she is well connected, to a publisher. Over the next few years, she will receive responses from people who have picked up a copy of her work. Even in the best of circumstances, she does not expect a response from her work in less than a year.

To borrow the language of economics, this scenario describes the cycle of the production, distribution and consumption of scholarly research. In general, it describes how scholars today do work given the means of communication available to them--the infrastructure of books and buildings that are circulated and in which scholars circulate as they produce their work. The scenario also describes how research is limited by barriers which do not become visible except by contrast to what is made possible by another means of communication, such as that provided by the Internet. At each stage in her research our student encountered limited access to information, specifically to the raw data, interpreted data and interpretive criticism. In each case, the restrictions arose from the fact that the information is stored in a medium which resists easy distribution.

Now consider what effect the Internet might have on this cycle. Assume that a site or number of sites have been set up where researchers can submit and retrieve information of various kinds, such as scanned images and transcriptions of epigrapraphic texts, translations of those texts, and criticism of those translations. Also found at these sites are various glyph dictionaries (such as Kurbjuhn's Catalogue), along with dictionaries of the relevant Mayan languages, as well as various experimental text-processing programs that have been created to aid in the task of translation. Assume further that important epigraphers contribute to these sites, and that they, along with numerous others, have contributed for some time.

With this rudimentary infrastructure of networked scholarship in place, each point of the cycle of research will be greatly enhanced. Our student will be able to access immediately a large body of data in the form of scanned facsimiles and tanscribed texts. She will also be able to access existing translations of these texts, along with, perhaps, criticisms of those translations. And she will be able to communicate the results of her work as it evolves, through the use of electronic mail with participants and by posting notes in archives reserved for such ephemera. As for the distribution and publication of her results, she can get a head start by posting her work to an appropriate archive. In short, at all points in our scenario where access to data was restricted by limitations of having to travel to a particular place at a particular time, our student gains enormously in saved time and quantity of information.

In addition to these benefits, the prospect of networked scholarship offers other, unexpected advantages that arise from the constraints that medium places on message. The Internet, after all, is made up of computers, and therefore differs from spoken and printed communication not only by virtue of its fast and vast distributive capacity, but by the formal requirements it places on information that is exchanged on it. Any networked document will have necessarily the properties of replicability, searchability and transformability. What this means for epigraphic research is clearest in the area of hieroglyphic transcriptions. More valuable perhaps than a scanned image of a text, which has limited processing potential, a properly transcribed text can be "read" by a variety of translation routines, from programs which simply parse texts or replace Thompson numbers with phonetic or semantic values, to "smart" programs designed to imitate the rule-governed aspects of decipherment and translation.

Equally valuable to epigraphy is the potential that electronic transcriptions offer for statistical analyses of collections of texts. For example, a text archive, which might easily contain the entire corpus of inscriptions, can be searched as a single document, and patterns of glyphic distribution, substitution and combination can be detected with great speed and accuracy. Other possibilities include the classification of hieroglyphic texts on the basis of their statistical "signatures", as has been detected in statistical analyses of novels, letters and other documents.

Whatever the role such text processing techniques may play in one's epigraphic research, there are other, more subtle features of networked scholarship that will have a decisive effect on that research. These derive from the nature of the medium itself. Although one accesses an archive on the Internet privately and as an individual, as if it were a bookshelf in one's study--and in a very real sense it is--the archive is in an equally real sense a public and collectively authored entity. In principle, all transcriptions are submitted individually and edited collectively. The sharedness of the medium means that transcriptions will tend to be standardized according to the consensus of participants. It also means that the difference between transcription and translation will be more marked--transcription being consensual and translation individual--with the result that the medium will encourage epigraphers to be explicit about the steps they make in producing a translation. Thus, if it is true that epigraphic practice cannot be reduced to algorithms--and it is difficult not to think so--it will be clear at what points rule-governed behavior gives way to intuition and "play," in the positive, hermeneutic sense of that term.

The above should suffice to demonstrate the point that the Internet can enhance significantly the process of epigraphic research. The immediacy and universality of access to primary and secondary sources, the centralization and standardization of that information, the explication of the interpretive process--these are factors that epigraphers should recognize as essential conditions for the growth of their research. It is with the purpose of creating these conditions, and fulfilling the potential that networked scholarship holds for Classic Mayan epigraphic research, that the Mayan Epigraphic Database Project has been created.