DigMaster is an online experiment in the publication of archaeological materials. The intent is to test whether electronic publishing will meet several needs of the archaeological community, including a more robust presentation of archaeological data, prompt publication, and collaboration between researchers working on related excavation projects.

Contents

DigMaster as an Experiment in "Total Publication"
DigMaster as an Experiment in "Early Publication"
DigMaster as an Experiment in "Collaborative Publication"
DigMaster as an Experiment in "Three-D Publication"
DigMaster Database Construction
Conclusion

DigMaster, the online production of the Cobb Institute of Archaeology at Mississippi State University, is an effort to meet in a novel way for archaeology the obligation to disseminate archaeological information to the academy and to the public. It is an experiment in bringing ancient artifacts to life for both scholarly and non-professional inquiry.

Archaeology as a discipline - including the archaeology of the ancient Near East - has long recognized its responsibility to publish all discoveries. That imperative provides something of a dilemma, however, if not a full-blown crisis, because of the fact that even a modest season of field excavation will produce far too many artifacts (not to mention architecture, soil layers, and faunal and floral remains) to be managed and studied rapidly and fully and to be presented in a timely way by traditional modes of publication (other than as simple lists). That dilemma - the moral obligation to move discoveries into the public realm and the seeming impossibility of getting that done - is the result of at least two factors:

1. thousands of artifacts are recovered in the excavation of old world tell sites in relatively short periods of time;

2. many of these artifacts can best be presented only in visual or graphic format, a prospect normally much too expensive for traditional publication means.

Historically, the lag between recovery in the field and final publication is frequently ten to twenty years or longer. In the interim, data which might have assisted ongoing research remains inaccessible to scholars and public alike. More normally, the "spectacular" or the "unique" find will be published quickly, while the ordinary object represented by numerous examples (which presumably would tell most about activities, practices, values) languishes in the laboratory or in storage. The result is all too frequently an objet d'art without context, beauty without meaning, and a skewed reading of the culture from which the artifact derived.

DigMaster as an Experiment in "Total Publication"

In the 1992 and 1993 summer seasons, Lahav Research Project excavation teams at Tell Halif in southern Israel recovered (among other artifacts) 550 ceramic and stone figurines and fragments in Field IV on the western edge of the tell. By analogy with other collections of figurines we know that the Halif corpus of (mostly) 5th C. BCE artifacts functioned in the realm of cultic practice; they had initially been placed in a shrine or temple by supplicants as votaries or as votive offerings. They were subsequently ceremonially broken and appropriately disposed when cleared from the shrine.

Our investigations soon discovered parallels to the Halif corpus, several collections of hundreds of figurines, for example, from sites within 30 kilometers of Tell Halif. Unfortunately, though these contemporary collections were either catalogued in excavation reports or given separate publications, none was available as a base for comparative studies, because none depicted graphically (whether photograph or line drawing) more than ten percent of the hundreds or thousands discovered. Furthermore, the figurines that were published, statistically a small percentage, usually reflected an author's aesthetic predilection to show the "best" or "unique" examples. Because they are incomplete, the published collections of figurines are less than useful tools for comparative studies; as databases they are seriously biased. In one instance, the most frequently occurring figurine type, the horse and rider, in an array of 700 was not represented at all, whereas the unique or those with minor representation received full treatment.

The reasons for the publication of such small portions of large collections are clear: the ideal of total publication by traditional means would have produced a volume too costly to purchase; not until the advent of electronic publication could such an endeavor be undertaken. The unfortunate result is that in the literature no reliable database is available for comparative and typological studies of the Halif corpus.

By publishing all 550 figurines, DigMaster has addressed the obligation to publish the Lahav Research Project discoveries and has begun to address the need for a reliable, non-biased database.

DigMaster as an Experiment in "Early Publication"

The overlong lag between discovery and publication is an embarrassment for archaeology of the ancient Near East. Whereas heated issues surrounding the Dead Sea Scrolls and the "failure to publish" have been the center of media attention, in fact, the publication record for other excavations, whose materials are not nearly so complex, is equally as bad or worse. Twenty or thirty years between discovery and publication seems more the norm than an exaggeration of fact. Often there is good reason for delays in publication: final publication prior to an exposure in breadth, requiring years of fieldwork, would be premature and would miss appropriate context(s) of artifact remains. Often archaeologists teach in universities and excavate, as it were, "on the side", without sufficient research time to process large collections of materials quickly. Most frequently, host countries to the excavations do not permit exporting artifact collections, thus severely limiting access for the purposes of ongoing study.

Whatever the reasons, or excuses, for archaeology the result is the same: data relevant to ongoing field projects and to laboratory research become, in effect, inaccessible, precisely the situation encountered as study of the Halif figurines began. Though we learned other large collections existed, the majority of which were not published, frequently these collections were also otherwise inaccessible, misplaced, or permission to review them was denied. (We add here that some excavators readily granted access and even helped to locate collections.) In some instances, not even the promise not to publish these other figurines gained permission to reference them in comparing Halif styles with theirs.

The right to the publication of finds is important for an excavator. We cannot imagine a situation in which artifacts recovered through the planning and work of an excavator are given to another to publish without due reason. However, the value of access to archaeological data and artifact evidence also cannot be overstated. It remains unfortunate that excavation teams quite often must begin field or laboratory work without benefit of parallels, which are known to exist in other (non-published) collections, to guide fieldwork or to assist in evaluation.

DigMaster, as an experiment in "early publication", has attempted to address this embarrassment to the discipline. We placed the 550 figurines online within two and a half years of the Lahav Research Project's leaving the field. Every artifact in this corpus, along with all supporting documentation, was available in graphic form (drawings, black-and-white and color photographs) to scholars by December, 1995. Since that time the Web site has been expanded to include multiple photographs of each artifact and Apple QuickTime movies of 59 figurines in an experiment at three-dimensional representation. We continue experimentation in VRML models. Final publication, probably as a CD-ROM, will come in due course.

We believe that DigMaster has been a particularly successful experiment in this aspect, for it has demonstrated the viability of getting data (albeit "raw" data) into the hands of scholars early in the process toward final publication. Furthermore, given our experience with misplaced and inaccessible figurine collections, and given the fact that the Halif artifacts must soon (December 30, 1997) return to storage facilities in Israel, DigMaster has been successful in keeping the corpus in the hands of scholar colleagues. Granted that online photographs are "not the same" as the objects themselves for detailed studies, the Halif materials are, nonetheless, available as no other collection. DigMaster has raised to a new level the obligation to make known the discoveries of archaeology, now with the promise of accomplishing that responsibility in a timely fashion.

DigMaster as an Experiment in "Collaborative Publication"

One outgrowth of DigMaster's experience with incomplete databases is our attempt at collaborative publications. We have issued invitations to excavators with non-published or incompletely published collections to add their materials to DigMaster. In the summer of 1996 permission was granted by Dr. Amos Kloner, Israel Antiquities Authority, to photograph and to digitize the 39 Persian age figurines discovered at Maresha, a site some 24 km. northeast of Tell Halif. These contemporary artifacts, placed online in August, 1996, in a common, searchable database with the Halif figurines, allow a browser to summon, for example, all common types from both collections. Beginning in June, 1997, approximately another 1,000 ceramic figurines and statues discovered at ancient Marion (Polis), Cyprus, will be added to DigMaster in a common, searchable database, more than doubling the number of online figurines from the 5th C. BCE.

In this aspect, too, DigMaster has demonstrated the viability of several excavation teams sharing resources. This experiment in collaborative publication across traditional excavation boundaries places a renewed urgency on the obligation to publish completely and early. "Territorial concerns" (publication rights) in this venture have been superseded by possibilities now imaginable via electronic publication.

DigMaster as an Experiment in " Three-D Publication"

Even with color photographs of each of the 550 figurines (minimally two photographs each, some with as many as nine), and with line drawings of 338, plus verbal descriptions, the online presentation is still not yet as good for detailedstudy as the object itself in one's hand. In an attempt to overcome the limitations of "flat" photographs on screen, we created for 59 of the figurines Apple QuickTime "movies." The " movies," which consist of up to 32 linked, digitized color photographs, appear to show a rotating (three-dimensional) object to the viewer. These QuickTime files move the project significantly closer to a three-dimensional appearance and hopefully closer to being a more useful tool for researchers. We also have work ongoing with VRML technology. However, untilthe PC commonly becomes powerful enough to manage large VRML files quickly, or until the sizes of VRML files are reduced significantly, our online VRML representations of figurines must remain few.

DigMaster Database Construction

Planning for the Future
When preparing the DigMaster Database we planned for the future in many ways. Because we recognized early in its formation that we would eventually add different types of information to the database, in addition to the original text descriptions, it became imperative that these future additions and maintenance had to be simple and straightforward, not requiring advanced computer or programming knowledge. Entry of data into the database has therefore been made as easy as possible while retaining the power of the search and multiplatform/application entry of the data.

To address these needs we relied on simple tools and programs that are readily available for low cost and are multiplatform in design. For database entry we originally used DBase and exported to a tab-delineated text file. Later in the process we adopted Claris FileMaker Pro, since it, too, is multiplatform and allows us to export data as text. Using a text file also facilitates the changing of records easily while online and logged onto the server. For photographs we used the GIF format for drawings and thumbnails of the objects and used the JPEG format since it offers greater compression of full color images. The search engine for the database was written in Perl so that it could be easily moved from platform to platform (without having to be recompiled) if the server changed in the future.

Data Organization
We initially collected the types of data that we originally wanted to distribute - text, pictures, and drawings. We anticipated that the supported data types would change as new technology became available and offered better ways to display our collection. For instance, we have since added 59 QuickTimeVR versions of artifacts and also have laser scanned VRML models of a few artifacts.

The collected photographs were subdivided and directories were created for each of the different types - thumbnail icons, line drawings, full-color large-scale photographs, and a miscellaneous directory. As data was collected, photographed, scanned, and entered into the database it was placed into the appropriate directories. Each artifact in DigMaster has a low resolution gray-scale thumbnail icon which represents it in artifact lists and appears on the top of its main page. These images are placed in the "icons" directory and labeled by artifact number (i.e. 1864.gif) as it is used in the database.

Multiple high resolution, full color photographs were taken of each artifact for documentation and research purposes. The number of photographs for each object varies, based on important features of the object, but the photos always show at least the obverse and reverse of every object, usually each of the four sides of the pieces. These photographs were put in the "picts" folder and labeled with their object numbers and sequential letters for those showing multiple views (i.e. 1864.jpg, 1864a.jpg, 1864b.jpg, etc.). QuickTimeVR files, VRML models, and drawings are treated in a similar manner, each type having its own directory.

Search

The text information for all of the objects was entered into a traditional database program which was then exported to a tab-delineated text file. Information, such as physical measurements, technical descriptions, type of material used in manufacture of the artifact, and researchers' notes, were recorded in the database. It is this text database that is then searched by the Perl program. Perl is quite well suited for handling text data, and the program for the search, while somewhat complicated, is actually relatively simple. It offers the ability to search for a keyword in the descriptions of an artifact, for a specific object number, or for strata level, etc. The power of a tab-delineated text search engine is in its simplicity. It allows a person to manage the data using any platform and allows remote management of the data via modem or network connection.

Conclusion

Our experience with DigMaster has illustrated that probably for the first time the discipline of archaeology may meet fully its obligation to publish findings. The experiment, we believe, has proved largely successful. First, perhaps most important, complete publication of large collections is now feasible. At approximately the same cost to produce a traditional book publication, which would necessarily be limited to a selection of artifacts presented in photographic form, DigMaster has come to include online every artifact in the Halif collection as well as all related information. Nothing has been omitted from the database. Second, the goal of forming the foundation for a reliable database was reached already with the completion of the 550th figurine from Halif. Along with the Maresha 39, the addition of the Marion 1,000 will greatly expand and multiply its reliability. Third, although none of the three figurine sets has yet reached final publication (and may not for several years), the artifacts have already served the needs of others working in the field. A statistical package, installed in December, 1996, has counted 26,174 users to the site. Fourth, traditional barriers between scholars have been in part successfully breached; three excavation teams have collaborated (or will have collaborated) in the formation of DigMaster.

One emotional factor discovered in the process of going online with the archaeological artifacts is the instant gratification of seeing one's work immediately available to any person with the basic electronic equipment, anywhere in the world. In fact, at the moment changes are saved to file, other people will see the latest "edition." Online publication is a powerful tool, and heady, too.

The Authors

Dr. Paul Jacobs, currently Head of the Department of Philosophy and Religion with a joint appointment in the Cobb Institute of Archaeology at Mississippi State University, is director of the Lahav Research Project Phase III excavations at Tell Halif in southern Israel. Jacobs has worked in excavations in Israel and elsewhere since 1970, in particular at the ancient sites Tell Gezer, Tell Halif, and Caesarea. Mail: Box JS, Mississippi State university, MS 39762 USA.

E-mail: pfj1@ra.msstate.edu

Christopher Holland is an undergraduate student attending Mississippi State University with a major in Art with an emphasis in Computer Animation and Graphic Design. He has been involved with the Web since its earliest days and has designed Web sites for both commercial and educational applications. Current projects include DigMaster and DigitalDarwins, an educational site for the National Museum of Natural History, Smithsonian Institution. He is also highly involved in computer graphics research.

E-mail: chris@sarc.msstate.edu

Copyright © 1997, First Monday