| what's new | announcements | public programs |
| website information & statistics | copyrights & permissions | comments |

| website navigational aid |


Disjecta Membra in Araneo -

Scattered Remains on the Web


Alexandra A. O'Brien ©1997-2001

[This paper was given at the annual meeting of the American Research Center in Egypt
in Ann Arbor on Saturday 12th April 1997]

Recent years have seen an increase in the use of computer technology in the Humanities, primarily in the form of large, searchable text databases such as the ARTFL project at the University of Chicago. Artful - A R T F L is an abbreviation of "American and French Research on the Treasury of the French Language." Currently this database includes almost one hundred and fifteen million words in nearly two thousand texts. The texts making up the database are chosen to be broadly representative of French literature of 18th to 20th centuries, with some 16th century, Renaissance and Medieval material. The texts making up the database can be searched either in all or part; it is possible to define text groups by author, time period, genre, word for example. Within these defined texts groups, or using the entire database, one can search for a word, or word stem i.e. "femme", or words beginning with "fem-" F E M.

Similar projects are on-going in a field somewhat closer to ancient Egypt, that of Classics: the largest is at present the Perseus Project, which describes itself as "an evolving digital library which currently focuses upon the ancient Greek world." This on-line library includes art, archaeology, texts in Greek and in translation as well as secondary sources and tools for working with the texts.

Another project, the TLG (Thesaurus Linguae Graecae) has collected together an even larger body of texts, and includes over 70 million words (not much smaller than the ARTFL database already mentioned) encompassing all ancient Greek texts surviving from the period between Homer (8th century BC) and AD 600, plus further Greek texts from the period between 600 and 1453.

Other similar projects are beginning elsewhere, for example here at the University of Michigan, the Papyrus Collection has made much information and material available on-line such as the interesting article "Snapshots of Daily Life: The Family Letters of Paniskos" and scans of two of the letters, with translation s into English. Each sentence of the translation can be clicked on for a sentence by sentence Greek text and commentary.

So far, none of this material is in Egyptian. And although much of the Greek material comes from Egypt you may all be wondering what this has to do with Eg yptology.

Well, accessible from the Perseus homepage is the Duke Data Bank of Documentary Papyri, where nearly 500 papyri have been made available on the World Wide Web, most not previously published elsewhere. Amongst these papyri are over 40 in Demotic, scanned and available on the world wide web. In fact, some 64 Demotic papyri are available on the internet, most of them not previously published and remaining unedited. Many of them have not even been identified. The Papyrological Institute of the University of Lecce in Italy has made scans of 10 unidentified papyri available on-line. These texts can now be seen by anyone, anywhere in the world with access to the World Wide Web. Without this medium, it's probable that such unpublished material would rarely be seen by anyone outside the institution itself.

This world-wide availability of such texts underlines the versatility and potential of electronic images. The superiority of these over traditional paper publication is manifold:

On the other hand, digitally scanned images of papyri can be stored on one computer, but made available throughout the world to be viewed on other computer screens via the internet using the worldwide web. Furthermore, if appropriate, such images can be downloaded by an individual user, who can then manipulate the images themselves on their machine by cropping and zooming, for example, in order to magnify a problematic part of a text. Magnification like this is possible because texts are scanned at a higher resolution (more "dots per inch") than can be displayed on a monitor at one time (in other words, the original scan is much larger than a computer screen). [One way to gauge this is to compare the original publications of the Carlsberg Papyri (PP. Carlsberg 15, 16, 17, 18, 19, 20, 21, 22, 23 and 32) available on-line with the electronically recorded images of these papyri.]

In fact, simple magnification is just one way of manipulating an image. The Electronic Beowulf project has used various methods of scanning, along with manipulation of images once scanned to enhance legibility of text. This project of the British Library and the University of Kentucky is working to produce a full-colour electronic facsimile of the earliest surviving manuscript of Beowulf.

The scanning of this particular text brings its own set of problems, as the manuscript was damaged by fire in 1731 and then remained in it's burnt binding until the 19th century when it was restored and rebound. This was done by tracing the outline of each leaf, cutting out the centre of each tracing save about 2mm, and then pasting each leaf into it's individual frame. This preserved the manuscript well but hid much text under the retaining edges of the frame. However, it was found that by backlighting the leaves when digitally recording them, the hidden text could be seen.

Thus, while it is being scanned can also be used to enhance the electronic image: as said, backlighting can reveal covered text, ultraviolet light can reveal erasures and faded areas while coloured filters and adjusting contrast and brightness can also be used to facilitate reading difficult passages.

As well as producing and manipulating scans, these computer techniques can be used to enhance black and white photographs: just reducing the size of an image can drastically increase resolution as does heightening contrast.

This is illustrated by different versions of this extract from Onchsheshonqy put together from images in Professor Janet Johnson's on-line article "Computers, Gra phics, and Papyrology."

Scan of published photograph

Halftone image

"Clean" L ine drawing

Translit eration

At the top is the scan from the original published black and white photograph; from this scan a "greyscale" image is produced, from which in turn is generated below, a half-tone image and then, a "clean" line drawing. All together these imaging techniques can help to produce, at the bottom, a transliteration and, ultimately, a translation.

Another advantage is that fragments can be placed together on the screen. Robert Kraft writes of his putting together 4 fragments from the collection of the Center for Judaic Studies at the University of Pennsylvania. The mounting of these fragments prevented them from being placed together physically, so initially he used Xeroxes of the fragments to do so. This was ultimately unsatisfactory due to inconsistencies in both the quality of Xeroxes and in scales of published photographs. However, once he had access to electronic images of the fragments he could juxtapose them on the computer screen, where he would be able to move any or all the fragments around, magnify all or any one of them, store images of each fragment separately or produce a composite image of the reassembled fragments, itself capable of further adjustment.

The ability to magnify parts of a text allows careful alignment of fibres of papyrus, or cracks in an ostracon, again facilitating reassembly of fragmentary texts.

Work on a Demotic Lycopolite census at the Centre for the Study of Ancient Documents at Oxford has used the juxtaposition of Demotic text with transcription to keep track of work on the document. The scan of the work in progress then, preserves the original text, the first transliteration by Griffith and the notes of the scholars presently working on the text. It has been possible for those involved in the work to improve readings of damaged text by "playing with the image" - by zooming in and out of difficult passages and altering the contrast in order to highlight faded ink.

Of course, manipulation of images as described produces results in seconds on a computer screen, certainly much faster than could possibly be produced using conventional photography, even if the same technique could be used (which is often not the case).

It's not only papyri that can be scanned, other inscribed objects have also been treated in this way: potsherds, wooden boards and lead tablets, as well as the vellum of Beowulf already mentioned. Again, damaged artifacts are particularly suited to the medium of electronic imaging, the manipulation techniques of which can be used to reconstruct text as we have seen.

Once generated, electronic images can be stored on CD-ROM and sold commercially to individual or multiple users. However, making images available on the world wide web to be viewed using a web browser such as Netscape, removes the problem of platform incompatibility - in other words, this removes the need to produce a "front end" programme for viewing images from a CD-ROM on either a Macintosh or a PC machine and avoids either the need to produce a range of CDs for use with different machines or, if only one type is produced, prevents limiting availability to users with certain machines.

If it's necessary to charge for access to images, it is possible to lock the site and sell the password to subscribers (a method already tried and tested by sites selling access to what might be termed "adult" images!) in fact, this will be the means of accessing the TLG in the not too distant future, and is how the British Library plans to allow access to Beowulf.

Now to return to a couple of items mentioned in the introduction: Perseus and the Thesaurus Linguae Graecae demonstrate the potential, not just for storing and accessing images, but also for processing quantities of textual information. Both Perseus and the TLG have varied and complex search capabilities, one can search through 100s if not 1000s of documents for a particular word, searching either all the documents in the database or limiting the search, for example to a geographical region or to a particular author. One CD has enough capacity to hold the entire database of the TLG, that's over 70 million words of Greek text as I already said! Searches of this kind could not be done perfectly without the computer, even if one had the time! Clearly, the potential for computerization is huge - both in terms of ease of access to quantities of text and in terms of electronic imaging. [Hand Out] I have produced a world wide web page where I have gathered and indexed all the Demotic documents currently available on the web. By facilitating access to these images I seek to encourage use and appreciation of this medium, the resources that are available and the potential for the future.

This web page contains links to 64 Demotic papyri, including legal documents, documentary texts, letters, religious and astronomical texts, as well as one inscription on a Faiyum portrait.

Because the scan of the portrait is at a much higher resolution than can be shown on the screen it's possible to view the inscription magnified to the point where it is legible (to anyone capable of reading it that is!)

As well as texts, there are links to articles, directories, institutions, projects (such as the Demotic Dictionary Project at the University of Chicago) and collections - including the Duke Documentary Papyri Database, where one can access the catalogue record of a papyrus (here P. Duke 883) , and from there scans of the text itself (this is a link to a 150 dpi scan of P. Duke 883). Also the Carlsberg Papyri (a link to P Carlsberg 32) and the Papyrological Institute of the University of Lecce, (a link to P. Lecce 6) all of which have made scans of papyri available on the web. There are also handlists and catalogues available on-line, such as the catalogue of the Beinecke collection at Yale, and a list of the papyri held by the University of Pennsylvania Museum.

This is just a sample of the material available. I should add that (apart from the black and white publications) all the images I have shown today were found on the world wide, as were all the articles I referred to and used to write this paper. These sites are listed on the handout and can all be linked to from the World Wide Web page, Demotic Texts on the World Wide Web.

Hopefully use of electronic media will continue to increase, if only for the sake of the documents themselves - it being possible for scans to be widely and easily accessible, as well as being a means of preserving these documents for the future. We can prevent exposure of such delicate artifacts to further damage by producing a permanent electronic facsimile which can be disseminated across the globe. It is clear that the world wide web will eventually be a standard place of publication, despite the tendency of some to underestimate it. In fact, in the not too distant future, the TLG will be accessible by subscription on the World Wide Web, which has clearly superseded the CD-ROM. With more scholars able to use these images and access large text databases, perhaps we can make greater strides in understanding the ancient texts and raise the profile of the field by introducing it to a wider audience.


The ARTFL Project. ARTFL Overview.

Crane, Gregory R. About Perseus.

Johnson, Janet H. "Computers, Gra phics, and Papyrology."

Kiernan, Kevin S. The Electronic Beowulf. Digital Preservation, Restoration, and Dissemination of Medieval Manuscripts

Kraft, Robert A. Manipulating Digitized Images of Some Unpublished Papyri

Antti Nurminen. Recording, Processing and Archiving Carbonized Papyri

O'Brien, Alexandra A. Demotic Texts Published on the World Wide Web.

Olsen, Mark. ARTFL Project. Frequently Asked Questions.

TLG. General information about the TLG project.

Thompson, Dorothy J. Digitising a Lycopolite census.

Van Minnen, Peter. The Duke Data Bank of Documentary Papyri.

Winter, J. D. "Snapshots of Daily Life: The Family Letters of Paniskos"

Return to Demotic Texts on the World Wide Web

Return to Research Archives

Return to Abzu

©1997-2001 Alexandra A. O'Brien
Revised: February 6th, 2001