![[Picture of Tov and Kraft at SBL on November 22, 1992]](/~jtreat/images/tov.gif)
ENIAC, the world's first large-scale digital computer, was created fifty years ago at the University of Pennsylvania. The Department of Religious Studies at the University of Pennsylvania and, more specifically, Professor Robert A. Kraft and his students have played a significant role in the introduction of computer technology into the humanities and the study of religion. It seems especially appropriate in this fiftieth year of ENIAC to chronicle some of these events.
The International Organization for Septuagint and Cognate Studies (IOSCS) from the time of its founding in 1968 put a high priority on creating a lexicon for the Greek Jewish Scriptures (commonly known as the "Septuagint"). The IOSCS co-sponsored a symposium on the subject in 1972; and in the symposium's published papers (Septuagintal Lexicography, edited by Robert Kraft), one can find a discussion of using a system of paper slips to organize the lexicographic data that would be needed to create such a lexicon.
Kraft had long been persuaded that "intelligent and imaginative use of computer technology was the most practical and efficient way to attack such a complex and massive collection of materials." At a meeting of the IOSCS steering committee in 1972 or 1973, Kraft suggested that it would make sense to use computer technology to store and analyze the data. Although few humanist scholars at the time were making much use of computers, the proposal made sense and John Wevers, IOSCS president, encouraged Kraft to investigate. In 1974, Emanuel Tov, Professor at the Hebrew University of Jerusalem, agreed to serve as editor elect for the proposed lexicon. Thus began an enduring collaboration between Penn and Hebrew University on a project that would pioneer new approaches to the study of ancient Greek scriptures: the Computer Assisted Tools for Septuagint Studies (CATSS). CATSS set out to create tools and an expandable data bank for Septuagint studies.
CATSS had the support of the Society of Biblical Literature and the scholarly community, but no funding. For the prepatory computer work, Kraft applied for a "feasibility grant" from the Program for Research Tools and Reference Works of the recently established National Endowment for the Humanities. Kraft recalls,
We were encouraged by the NEH to pursue these dreams (they saw the value), but the evaluation process they had established did not yet integrate computing aspects of proposals with more traditional scholarly aspects, which made things difficult for the applicants as well as for visionary NEH administrators.
The NEH approved the proposal and provided funds for a feasibility study in 1978-1979. At the time, Jack Abercrombie, who had computing experience, was writing his dissertation in archaeology under James Pritchard. Kraft hired him to visit computing centers in Europe and Israel to see what could be done. Abercrombie visited Emanuel Tov in Jerusalem, F. Poswick at Maredsous (Belgium), the Septuagint Project in Göttingen, and Susan Hockey at the Oxford Text Archive.
During 1979-1980, Bill Adler (an advanced grad student in the program) was studying in Jerusalem on a Penn-Israel Fellowship. Among other things, he learned programming (on punch cards!). He also helped establish closer ties with Tov and his students.
Meanwhile, Abercrombie finished his PhD and left to teach Religious Studies at Waynesburg College (PA). For three years (1979-1982), he developed programs and experimented with data structures on the Waynesburg DEC computer to support the developing goals of the Project. He worked primarily on the text-critically difficult material in 2Kings.
Building on the feasibility study of 1978-1979, the CATSS Project was able to get an NEH grant for a "pilot study" in 1980-1981. CATSS was able to purchase a machine-readable tape of Rahlfs' Septuaginta from the Thesaurus Linguae Graecae (TLG) project. When Adler returned to Penn in the Fall of 1980, he became coordinator of computer activities for the Project and directed the students involved in the work.
Penn appointed Tov as a visiting professor for 1980-1981. Working in close proximity during the year, Tov and Kraft collaborated in developing strategies, testing algorithms, and training young scholars.
About this time, David Packard (a classicist and the son of the founder of Hewlett-Packard) became interested in the Project. Packard's Ibycus system, located at the Institute for Advanced Studies in Princeton, had advanced capabilities to search and display Greek texts. The CATSS team arranged to install an Ibycus terminal with a direct line to the Princeton system. Adler recalls:
The project really got underway once we got the NEH grant and David Packard got involved. I can remember when he installed his MORPH program (written in IBM 360 assembler language) [on the mainframe at David Rittenhouse Labs]. All the techies at DRL were impressed by the program.This installation occurred in Spring 1981. Adler was able to learn the Ibycus system's Ibex programming language and began to teach others how to use it.
For the pilot study, the CATSS team worked on Ruth and Samuel-Kings, inputting the variants and producing rough data, including concordances. When Tov returned to Jerusalem, he explored the opportunities for computing there and established a project base. The Jerusalem team began work on a Prime mainframe (later moving to an IBM mainframe).
In 1982, CATSS received a major two-year grant from NEH and other sources (including the David and Lucille Packard Foundation). In June of that year, computer guru Abercrombie was hired to come back to Penn and work full time with the project, and the Project acquired its own mini-Ibycus (a Hewlett-Packard mini-computer running David Packard's Ibycus system). The Ibycus was installed in 111 Duhring Wing; several Ibycus terminals were placed in 110 Duhring and one in Classical Studies. For Adler, who had just received his Ph.D. and continued to work on the Project, acquiring the Ibycus "was important because it was less intimidating than working on the DRL mainframe." A local Ibycus was also considerably more convenient than the long distance hookup with Princeton.
The CATSS team began to work seriously on the Project's three major goals: (1) to align the Greek text with the Hebrew in parallel columns; (2) to provide a morphological analysis of each Greek word; and (3) to collect all published Greek variants. The team drafted a pilot publication, using Ruth as an example of the database and illustrating how it could be used. The book was eventually published as Abercrombie, Adler, Kraft, and Tov, Computer Assisted Tools for Septuagint Studies (CATSS): Volume 1, Ruth (Atlanta: Scholars Press, 1986).
Other major grants for the CATSS Project followed in the years 1984 to 1994 with funding from NEH; matching funds from the Packard Foundation, the Society of Biblical Literature, the Jewish Publication Society, and other sources; and significant investments in equipment and support staff by the University of Pennsylvania. The Jerusalem team received funding from Israel Academy of Sciences for the years 1982-1988.
By 1981, Parunak and Whitaker had put the Hebrew text into computer form (the "Michigan-Claremont" text), and CATSS already had Rahlfs' Greek text in computer form from TLG. In 1982, Abercrombie (in consultation with Tov, Adler, and Kraft) wrote a program on the Ibycus to make an initial rough alignment of the Greek with the Hebrew. The Penn team finished the preliminary, automatic alignment of the Greek and Hebrew texts in the next nine months. That such results could be produced in nine months was very impressive. The resulting aligned text was then sent to Tov's team in Jerusalem for verification and correction. Paul Lippi and Zippi Talshir played leading roles on that team. It took a year or two to correct the alignment.
In 1984-1985, Adler was back in Jerusalem, teaching at Hebrew University. While there, he wrote a program that collated the two electronic versions of the Hebrew text (the Michigan-Claremont and the Maredsous texts) and identified errors. In this way, Tov was able to work with a clean Hebrew text. The Jerusalem team continues to adjust and correct the parallel aligned text.
Ronit Shamgar of the Jerusalem team wrote a program on their IBM mainframe to concord the parallel alignment. B. Nieuwoudt worked on a DOS version, and Galen Marquis developed a Fourth Dimension application on the Macintosh called CATSSBase, which fully integrates the data files of the parallel alignment with the morphologically analyzed Hebrew and Greek.
Bill Adler coordinated the Morphological Analysis of the Greek Text. After Packard's MORPH program was up and running, the CATSS team generated morphological analyses of several books of the Greek. These were parceled out to workers, who manually corrected the printouts. From their corrections, Adler updated the dictionaries used by the MORPH program. The bigger the dictionary got, the better the results. By the time they got to the final stages, the program was operating at a high degree of efficiency. Adler produced the "Morph Creed" and other documentation, including "Computer Assisted Morphological Analysis of the Septuagint," published in Textus 11 (1984) and in the Ruth volume. The "Morph Creed" was later updated by Ted Bergren.
Fritz Knobloch spent a lot of time standardizing the morphological analysis. He recalls using the Ibycus line editor in those early days:
It might be difficult for a student today to imagine the "primitive" editing environment on the Ibycus. Changes were made by entering the line, and one could only go sequentially through the file; once line 2500 was edited, for example, you could forget about doing anything to the first 2499. Much of the morphing was done under those conditions (although later editing programs allowed non-sequential editing).
By the time Adler left Philadelphia in 1984, the Morph project had advanced a great deal. At that point, Ted Bergren took over the Morph project, running diagnostics on the entire database to make it consistent. When Bergren left to teach at the University of Richmond in the Summer of 1990, the Morphological Analysis was ready for use, although still in need of further corrections. (Bernard Taylor's use of this data for his Analytical Lexicon to the Septuagint (Grand Rapids: Zondervan, 1994) caught further inconsistencies and errors and will soon facilitate a better edition.)
Ben Wright was assigned to coordinate the Variants sub-project, the goal of which was to create an expandable database containing all published variant readings of the Greek Jewish Scriptures. Adler first enunciated the idea of working on the word level, and that was the way they chose to implement the files. The project was able to proceed rapidly in its early stages, largely due to Abercrombie's previous experiments with algorithms and data structures at Waynesburg. Kraft and the CATSS team developed the actual format, which is documented in the Ruth volume. Kraft and Wright manually encoded the Ruth variants to test the format. The Project acquired the use of a Kurzweil scanner and used it to perform Optical Character Recognition on the published critical apparatuses. Then Ibex programs were used to massage the data into the new format, and the CATSS workers painstakingly edited the files to correct them.
In 1989, Jay Treat, David Rech, Ted Bergren, and Alan Humm transferred the process of producing variant files from the mini-Ibycus to micro-computers and provided further nuances to the format. (This format has now been used in other projects, such as Joe Farrell's Vergil Project.)
Some of the Variants files are now available on the Internet and more will be there soon. Progress continues, if slowly, on this last goal of the CATSS Project.
Over the years, many people have worked on the Penn side of the CATSS Project. Besides Bob Kraft and Jack Abercrombie, the primary staff on the Penn team have been Bill Adler, Ted Bergren, Alan Humm, Maddy Lopez, Walter Mankowski, Jackie Pastis, David Rech, Jay Treat, and Ben Wright. Other Penn people involved in the Project are Ken Banner, Jim Blankenship, Ben Dunning, Patrick George, Max Grossman, Joe Gulka, Noel Hubler, Amey Hutchens, Allen Kerkeslager, Brad Kirkegaard, Fritz Knobloch, Mary LaRue, Lynn LiDonnici, Beth Lisi, Lenny Lopez, David Louder, Alan Lowenschuss, Sigrid Peterson, Gil Renberg, and Rob Rice. Several people from Westminster Theological Seminary have furthered CATSS: Dale Brueggeman, Steve Horine, Karen Jobes, Alan Lenzi, and Bill Stroup.
The Jerusalem team under the direction of Emanuel Tov continues to be an essential partner in CATSS. People who have worked on the Jerusalem team include: Paul Lippi, Zippora Talshir, Galen Marquis, Nehama Leiter, Kyong-Rae Kim, Frank Polak. Philadelphians who worked on the Jerusalem team are Ben Wright, Fritz Knobloch, and Bill Adler. The Jerusalem team's programmers are Ehud Manzuri and Ronit Shamgar.
The Morphologically Analyzed Hebrew project began at Westminster Theological Seminary as a sub-contract of CATSS under the direction of Al Groves and used the Hebrew morphological program package Richard Whitaker wrote for CATSS; the Hebrew Morph later became an independent project of Westminster. Other people outside Penn who have contributed significant materials include Richard Whitaker, Bernard Taylor, Hans Erbes, and Dick Saley.
The new technology made it necessary to train students in it, both informally and in courses. In 1981, Kraft and Adler started the first RelS 409 course, "Computing in the Humanities," using Apple II machines in the Math lab. (An older course had been offered by Lawrence Klein.) In 1983, Abercrombie started teaching RelS 410, "Computing Programming in Literary Analysis." When Adler left, Abercrombie took over teaching RelS 409 also. In later years, these computer classes have been taught by Rech, Humm, and Treat. RelS 409 has been particularly popular in the last few years, and has introduced many students to the wonders of the Internet. When students wonder why these are Religious Studies courses, we tell them about our department's historic involvement in computing.
In 1983, the Religious Studies department required all graduate students entering its program from the Fall of 1984 onward to demonstrate "computer literacy." This was the nation's first requirement for computer literacy in a graduate program in Religious Studies (and probably the first in the humanities).
The CATSS staff also felt an obligation to foster computer literacy among humanities professionals outside the University. They participated actively in such pioneering electronic forums as the Humanist discussion group, which began in 1987.
In April 1984, Kraft published "In Quest of Computer Literacy" in Council on the Study of Religion Bulletin. This signalled the start of the Offline column which appeared under his byline for the next 10 years in that bulletin and in the newly established Religious Studies News.
In late 1984, Abercrombie started Online, an electronic newsletter for computing in the humanities, available on Bitnet. By March of 1985, 113 institutions were connected to Online. Online was active until about 1991.
In September 1984, the University Press published Abercrombie's Computer Programs for Literary Analysis (Philadelphia: University of Pennsylvania Press), including diskettes of several of the programs for Macintosh or DOS machines.
Offline began as a service of the SBL Computer Assisted Research Group (CARG), another organization in which CATSS personnel were greatly influential. Kraft chaired the CARG group from 1984 to 1988. The 1984 SBL CARG session heard reports from several pioneers: Gramcord, Thesaurus Linguae Graecae, Nota Bene, Linguist's Software, GreekKeys, John Hurd, Richard Whitaker, F Poswick, Alan Segal, and many others including CATSS. Nearly 100 people filled out "User Group Registration" forms at the main CARG desk run by Jackie Pastis.
In March of 1985, Offline was trying to entice people into becoming "informed about the 'computer revolution,' and 'the new technology.'" Kraft suggested replacing typewriters with computers. "Because on the computer you type the material only once, and then make whatever modifications are necessary without having to retype what does not need to be changed." The column predicted:
Services traditionally available only by visiting a library are becoming more accessible to computer users at their own desks (e.g. searches of bibliographies, indices, abstracts, articles, newspapers, encyclopedias) and it will become increasingly possible to access large data banks of relevant texts as central archives are created for this purpose. Communication with other computer users by means of telephone lines is commonplace.
In "Offline 15" (dated 1 Sept 1987), the word "hypertext" was explained. The same issue bore witness to the explosion of computers into the world of the humanities:
OFFLINE is but the tip of the iceberg when it comes to finding out what is going on in the computer world. The mind boggles at the number of publications, columns, newsletters, etc., that are available.
For the critical first ten years, Kraft used Offline to get humanists on board with technology. After ten years, he passed the editorship on to others. In "Offline 45" (16 May 1994), Patrick Durusau said:
Robert Kraft, through the Offline column and other efforts, has brought the seamless integration of computers into religious studies closer to reality. We thank him for introducing us to the potential and practice of humanities computing and for helping us to negotiate the passage to this new medium of research and communication.
In the mid-1980s, microcomputers became widely available and affordable. Computing in the humanities exploded. There was suddenly a voracious demand for electronic files of biblical and related texts. The CATSS staff rose to meet the demand. As expressed by a 1989 grant proposal,
Technological progress and the steady growth of a data-starved scholarly clientele increasingly focused CATSS attention on the issues of distribution and convenient access to the materials that were becoming available. Thus the Greek textual variant sub-project took a back seat to the more pressing issues of distributing the various texts and tools ... and making them easier to use on microcomputers.
In 1983, a committee of humanities professors at Penn decided to try to create a humanities umbrella organization under which such projects as CATSS (the main active and funded project at the time) could operate, develop software, seek funding, and so on. Developing naturally out of CATSS, the Center began an informal existence under a variety of names ("Center for the Computer Analysis of Humanistic Texts," "the embryonic Center for Computer Assisted Textual Research," "Center for Computing and the Humanities," and "Facility for Computer Analysis of Text"). By May of 1986, it officially became the Center for Computer Analysis of Text (CCAT). On hot sticky days in the summer of 1989, Treat called it the Sweatshop for Computer Analysis of Text; but over the years CATSS and CCAT helped to finance many graduate educations that would not otherwise have been possible -- his own included.
By 1984, the Center was distributing a wide variety of biblical and other texts. Kraft made special arrangements with copyright holders so that the texts could be distributed at cost for non-commercial, educational uses. In December of 1987, CCAT offered its materials on a CD-ROM produced by the Packard Humanities Institute. Bergren and Treat assisted Kraft in preparing the numerous materials for this CD, which included texts as divergent as Kierkgaard and the Rig Veda. By mid-1990, the 500 copies of the original CD were sold out.
Pastis coordinated CCAT distribution in the early years. In later years, Rech and Humm have handled distribution on DOS machines. Treat initiated distribution on Macintosh computers and is still involved in it. Mankowski, Treat, and Humm created software to make the files usable on DOS machines, and Treat created software and fonts for Macintoshes.
In "Offline 3" (dated October 1984), the Center offered scanning services on a Kurzweil Data Entry Machine (KDEM) for Greek, Coptic, Hebrew, and other languages. Center personnel used to go to what is now Wharton Reprographics at night to use their scanner after they had closed. When Wharton no longer wanted their machine, Kraft persuaded them to donate it to CATSS. Somewhere around 1989, Rech was instrumental in acquiring a newer, more efficient scanner, the KDEM 4000. For several years, Rech was our scanning wizard.
In 1984, Abercrombie moved out of the CATSS Project and became Coordinator of Computer Education and then Assistant Dean for Computing and the Humanities at Penn. Abercrombie and his programming assistant Todd Kraft focused on the needs of the University for computer-assisted language instruction, training, and technical development, while Bob Kraft focused on external services (scanning and distribution). In 1988-1989, Penn formally separated CATSS from its offspring umbrella organization CCAT; but "the CCAT was out of the bag," so to speak, and most people still know Penn's text distribution services as "CCAT." Since then, CCAT text distribution has been done through secondary distributors, including several CATSS personnel working independently. Distribution on disk has become less important as the files have become available across the Internet. Many of the files can be accessed on the textual archives of the CCAT workstation:
Several commercial packages distribute our materials. The modest proceeds from secondary distribution of CATSS materials are sufficient to underwrite continued steady progress toward completion of the Variants project.
In the Spring of 1988, Kraft offered his graduate seminar on computer-assisted papyrology. Kraft and some of his graduate students have been active in the Text Encoding Initiative and in the development of SGML standards for describing manuscripts. He continues to find innovative ways to use new technologies in research and instruction. For examples, see his web page (at http://ccat.sas.upenn.edu/rs/rak/kraft.html), where he keeps links to publications and class materials, including useful resources for Western World Religions.
On March 18, 1994, Kraft's graduate students surprised him with the preliminary publication of Documents in the University of Pennsylvania Center for Judaic Studies: Texts Published in Honor of Robert A. Kraft, containing digitized papyrus fragments on CD-ROM. This is the first known attempt to honor a scholar by publishing manuscripts on CD-ROM. The CJS Papyrus Project is still in process, under the direction of Kass Evans, one of the department's graduate students.
In the early 1990s, Guy Welbon produced the Asian Civilization CD-ROM, containing images relating to religion and other aspects of Asian civilization. His creative use of computer images in the classroom has set new standards for instruction.
In the last few years, Jim O'Donnell (Professor in both Classical Studies and Religious Studies) has directed CCAT and has been a strong advocate for the use of new technology in the service of instruction. His presentations across the campus and across the country have interested a growing number of people in "the shiny new toys" of technology. His web site (formerly, http://ccat.sas.upenn.edu/jod/; more recently, http://www9.georgetown.edu/faculty/jod/) is one of the most popular on the Web and contains impressive materials for instruction and research. Since 1990 he has co-edited Bryn Mawr Classical Review, and since 1995 he is co-editor of NewJour. During 1995-1996, he co-chaired the task force for restructuring computing services at Penn, and this year he is interim Vice Provost for Information Systems and Computing.
The Religious Studies department was among the first departments at Penn to have a World Wide Web site. (Its address was: http://ccat.sas.upenn.edu/rs/; in 2007, it became http://www.sas.upenn.edu/religious_studies). In 1995, the department put descriptions of all of its classes on its web site, and many of them now have extensive class materials available. Thanks to Eddy Breuer, Barbara von Schlegel, and William Grassie, some very impressive materials are now available for Western World Religions, Islamic History, and Science and the Sacred.
The CATSS project began as a way to facilitate the construction of a lexicon for the Greek Jewish Scriptures. The Project itself did not create a lexicon of the Septuagint, but Johan Lust is currently publishing a lexicon of the Septuagint using CATSS data. Instead, the CATSS project set out to provide the tools and data needed for any serious research of the Greek Jewish scriptures. The Parallel Aligned Hebrew-Aramaic and Greek Texts of Jewish Scripture, the Morphologically Analyzed Greek Jewish Scripture, and Westminster's Morphologically Analyzed Hebrew Bible are complete (apart from occasional correction of errors). The Variants project is still in process. These materials are an indispensable part of contemporary scholarship.
The people who have been part of the CATSS project have influenced the way computers are used in the humanities -- out of all proportion to the size of our program. Bob Kraft has served on various advisory committees at Penn and on boards outside the university, including CARG, TEI, and the CETH advisory board. Until January of this year, Jack Abercrombie served as Executive Director of Educational Technology Services at Penn; since then, as Vice Provost of Computing at Brandeis University. Ben Wright and Ted Bergren coordinate the electronic Ioudaios discussion, and, until recently, Bill Adler co-chaired Ioudaios Reviews. As manager of the Prep Center, Jay Treat assists Penn faculty in using technology for instruction. Hundreds of people use programs created at CATSS, and thousands use the texts.
In 1981, Adler and Abercrombie estimated that only about seven people in the humanities at Penn had touched computers, and five of those were working on the CATSS project. In 1981, the CATSS team were pioneers in humanities computing. With the explosion of personal computers in the mid-1980s, the situation changed. Computers became a standard part of the scholarly tool kit. Many humanities departments at Penn are now using computers pro-actively; one thinks especially of Classical Studies and the English Department.
Penn's influence on computing in the humanities is largely due to Kraft's willingness to share what CATSS was doing at Penn. One frequently sees him showing yet another person how to use the equipment, and team members have regularly made presentations at AAR-SBL meetings. The Religious Studies department contributed to the widespread use of computers in the humanities at Penn and other institutions.
In a 1989 grant proposal, Kraft could describe the Project's contributions in the following terms,
CCAT ... has trained a talented crew of young humanist computer-conversant scholars, has created and collected textual data and established routes for inexpensive dissemination, has developed programs for manipulating the data in various ways, has become a central source of information about computing in the humanities, and has helped lay the basis for effective cooperation and coordination of similar efforts throughout the western world.
The mini-Ibycus is not currently operational. Its 9-track tapes hang neatly on the shelves in Duhring 111, a vast archive from the busy days of the 1980s. Replaced by dozens of workstations, the DRL mainframe ceased operations a year ago. The Kurzweil scanner in Duhring is still in service, but most people are doing OCR on personal computers now. CCAT and CATSS still receive requests for text files, on disk and over the Internet, but there is less demand than a decade ago because these texts can be found in so many places, both commercial and non-commercial. The World Wide Web makes dissemination of data painless. Many scholars are now doing what we at CCAT once did in relative isolation. This is all to the good. A whole generation of humanists has been given the tools and the firstfruits of the electronic age.
This article was written with the help of Bob Kraft, Emanuel Tov, Jack Abercrombie, Bill Adler, Ben Wright, Ted Bergren, and Fritz Knobloch. I thank them warmly for their contributions. Most communication took place over the Internet. Most detailed information derives from electronic copies of Offline, from grant proposals, and from an occasional book or article. I have given preference to the evidence of official documentation over living memory or oral tradition when they conflict, especially with regard to dates. For further information on the early history of CATSS, see the report, “Computer Assisted Tools for Septuagint Studies,” published in the 1981 issue of the Bulletin of the IOSCS.
This article appeared in the 1996 edition of the Boardman Bulletin.
URLs updated November 8, 2012 and May 25, 2016
Jay Treat, alumnus