Scale in such an enterprise is important: we knew that we could not, of course, include every piece of primary evidence about ancient Greece, but what we wanted to assemble was a "critical mass" of materials. We wanted the database to have enough overall depth that it would support wide-ranging study of many topics. For the general student of ancient Greece, Perseus 2.0 constitutes a compact digital library with many key materials: 24,000 images, with information on 1420 vases, 366 sculptures, 384 buildings, 179 sites, and 524 coins. Perseus 2.0 holds two thirds of the surviving literature up to the death of Alexander the Great in Greek with English translation, the Liddell-Scott Intermediate Greek English Lexicon, a color Atlas of the Greek World, articles on ancient authors and archaeology, an Overview of Greek History from Homer to the death of Alexander the Great, and a 2600-entry Bibliography. Thus, Perseus contains a representative selection of primary data and a substantial amount of secondary documentation; however, it was not conceived as a replacement for the University library. Perseus 2.0 is, rather, a tool which complements traditional resources, and it may be compared to a book shelf stocked with standard texts and references, with the ability inherent to the electronic medium to search and organize information quickly.
After considerable thought, however, we chose to build as general a tool as we could rather than a resource that covered one or more topics of ancient Greece in detail. In part, our decision reflected the prejudices of faculty who told us that they wanted better access to primary materials--texts, maps, pictures of art objects, site plans, searching aids and other basic tools--so that they could construct their own arguments and narratives. At the same time, we sensed that neatly packaged electronic books were premature: all of us, in every field, are still learning how best to use new interactive tools. We realized early on that no one really understood how electronic editions of source texts should be designed or how catalogue entries for museum objects should be organized so as to exploit the strengths of this new medium. We chose to concentrate on what we called already in 1986 the "infrastructure" necessary to help people -- from elementary school students to scholars -- explore the ancient Greek world in ways that were not feasible with print tools.
Second, we were also determined to build a database that would be accessible to the widest possible audience. It would, in some ways, have been much easier to build Perseus on a high-end workstation, but neither these high-end machines nor the expensive software components that we would need were widely accessible to our target audience. While we hoped to construct something of interest to computer scientists specializing in Hypertext systems, we wanted Perseus to extend, not restrict, the audience for ancient Greek culture. Our goal was to create a database for a hardware platform that cost less than $3,000. At present, the cost of a system to run Perseus -- and run it well -- is now between $1,500 and $2,000. When in 1987 we began serious development, MS-DOS dominated the PC world and the Macintosh was the only feasible platform for our work. HyperCard (which we first encountered in May 1987 as beta software called WildCard), for all its limitations, provided a reasonable delivery environment. HyperCard is, however, slow, limited in its capabilities and, worst of all, for all practical purposes restricted to the Macintosh environment. Perseus 2.0 is, in all probability, the last version of the database that will use HyperCard. Nevertheless, Hypercard has allowed us to provide a wide range of features on a relatively inexpensive hardware platform.
Our primary challenge, perhaps, has been reconciling the emphasis on infrastructure and the determination to reach the widest possible audience. The majority of our efforts have gone into building a database that would not only serve a broad audience immediately but that could also grow more useful over time. As a result, there is a sharp distinction between Perseus 2.0--the monumental database that we can distribute inexpensively on CD-ROM (compact disc read-only memory)--and the "real" Perseus database, which consists of many applications and databases stored in a variety of formats, none of which is tied to any one application or operating system.
All texts have, for example, been stored in a general purpose format called SGML (Standard Generalized Markup Language) that can facilitate sophisticated text analysis (e.g., compare the vocabularies of two different characters in a play or search for poetry quoted by Plato). HyperCard cannot, however, support such work, and we have had to strip our texts of this SGML encoding before publishing them.
Likewise, when we first began work assembling pictures, videodiscs were the best medium for publishing images and the infrastructure needed to support libraries of digital images was years away from our audience. We chose, however, to collect 35 mm slides, even though the film was vastly more expensive than videotape. The slides, when transferred to video, did not produce images as sharp as those based on 3/4 videotape. Now, hoewever, as videodiscs have begun to disappear and digital images are common, our digitized slides are far clearer and more detailed than any video-based images could be, and we can, in future years, redigitize these slides to provide images of even greater quality. If we had used video images rather than slides to produce the best possible videodisc in 1992, we would have constructed a resource that was obsolete in 1995. The maps, linguistic databases and other elements of Perseus each reflect self-standing projects that exploit or, in some cases, define standards. Perseus 2.0 is, in many ways, the tip of an iceberg. Even if we were not to add any more materials and were only to republish Perseus in newer software packages, its functionality could increase dramatically over the years.
First, those who are not specialists in ancient Greek culture have, in working with pre-release versions of Perseus 2.0, found it to be a handy and fairly comprehensive environment within which to explore aspects of ancient Greece. Thomas Martin's Overview of Classical Greek History has been expanded since Perseus 1.0 and now contains more than two thousand new links to other materials in the database. The Overview was designed to provide a succinct introduction both to many aspects of Greek history, literature, society and culture and to the resources available generally in the database.
Second, Perseus 2.0 is one prototype for a digital library, but it was designed to augment, rather than replace, traditional tools. Perseus 2.0 will offer to specialists a useful compendium of many source materials, but the individual pieces of Perseus 2.0 will probably be most useful for this group: the specialist will find in Perseus a series of interlinked individual tools, a collection of "books" rather than a comprehensive library. Already in Perseus 1.0, we were able to offer a number of tools which constituted an advance upon what had previously been published. The retrieval tools, for example, allowed users to search the texts included in Perseus 1.0 in ways that no other system allowed. Similarly, Perseus 1.0 included a substantial body of original visual materials commissioned by the project: the 1,000 color pictures of 500 Greek coins from the Dewing Collection, the 450 images of the Aegina Pediments in Munich and the 111 views of the Harvard Cleophrades Crater, to take only three examples, each could have constituted a separate publication that contributed to the study of Greek art and archaeology. Perseus 2.0 contains many more new resources for the specialist. The philological tools not only cover more authors but have themselves been refined and expanded. The coverage of art and archaeology has grown enormously--Perseus 2.0 contains, for example, information on more than ten times as many vases than did 1.0. And where Perseus 1.0 contained only outline maps of ancient Greece, Perseus 2.0 contains roughly a thousand color satellite images and topographic maps.
-- Online help, keyed to the part of Perseus in which you are located
-- Search tools for art and archaeology
-- the Browser and Lookup, designed to make it easier to find out what is in the databases
-- the Search Saver, a means of storing and keeping at hand the results of an operation
-- new Philological tools which place at your disposal an advanced and sophisticated means of searching both Greek and English texts, and processing and organizing philological data.
With the exception of the Atlas (which has been completely rebuilt), no aspect of the Perseus database has grown more dramatically than our collection of vases. Perseus 2.0 contains 1,420 vases illustrated with more than 14,500 images. A grant from the NEA allowed us to commission six essays by leading experts in the field which introduce the styles of major Greek Vase Painters as well as the collection of vases included in Perseus 2.0. These essays which total c. 40,000 words (roughly 160 pages of double spaced typescript) contain specific links to 428 separate illustrations, many of them new photographs taken by Maria Daniels, the Perseus Project Photographer.
Of particular interest will be chapters in this Guide devoted to Search Tools (chapter 5), the Atlas (chapter 7), and Philological tools (chapter 8, part 2).
Two issues in upgrading versions are covered, Converting Paths from Version 1.0 to 2.0 (section 10.2.8) and the CD Swapper (section 10.4) built to accommodate this multi-cd version of Perseus.
Full-screen images in the Concise Edition include all pictures linked to the Historical Overview, to the Encyclopedia, to the Essays, to vases from Boston Museum of Fine Arts cataloged by Caskey and Beazley, to over 900 site/architecture images from the Large Site Plans and to the Paths that are distributed with Perseus 2.0. In addition, all of the approximately 3600 sculpture images have been included. This brings the total to approximately 6100 full-screen images in the Concise Perseus.
These images are kept in five folders called Unversal Images on the Perseus CD. To get a sense of what is there, you can browse through them with the Slide Shower (described in section 4.6.5).
For information on upgrading to the Full Edition of Perseus, please contact Yale University Press, 203-432-7620 (phone), 203-432-2394 (fax), yupmkt@yalevm.cis.yale.edu (e-mail).
Perseus for Windows: Support from the Annenberg/CPB Projects has allowed us to begin work on a new platform-independent version of Perseus that will run both on the Macintosh and under the Windows operating system.
Perseus on the World Wide Web: Perhaps the most exciting development in recent years has been the rise of the Internet. We have established a Perseus World Wide Web site (http://www.perseus.tufts.edu), which contains a variety of useful tools and the latest developments in Perseus. The Web site includes copies of documentation, a growing body of source materials (some of which are not in Perseus 2.0), reports on the actual use of Perseus in teaching and research, links to other relevant sites (such as the Perseus Atlas Project at Holy Cross) and various other services.
Perseus Atlas Project: The maps available in Perseus 2.0 were developed by Professor D. Neel Smith of Holy Cross. His work on using state-of-the-art "Geographic Information Systems" tools and datasets such as the satellite imagery and the Digital Chart of the World continues. One goal of the Perseus Atlas Project is to create an online service whereby this material could be accessed and analyzed directly over the Net.
Evaluation of Perseus in Teaching and Learning: A grant from the Fund to Improve Post-Secondary Education has allowed us to study the impact of Perseus on teaching and learning at more than a dozen institutions. Results from this can be found on the World Wide Web Site.
The Electronic Liddell-Scott-Jones Greek Lexicon: Grants from the National Endowment for the Humanities and the National Science Foundation have made it possible for us to begin creating an electronic version of the standard Greek lexicon as well as other more specialized lexicographic tools.
Ancient Science: Support from the National Science Foundation has allowed us to begin work on materials central to the history science. This support will allow us to offer a complete version of Aristotle as well as of other key texts (such as Euclid's Elements).