Vol. X, No. 2

Fall, 1997

The Perseus Web Site: A Review

Lee T. Pearcy
Alex Pearson
The Episcopal Academy

Perseus (http://www.Perseus.tufts.edu), Gregory Crane, ed., July, 1997.

We began looking at the Perseus site through Netscape on a Gateway 2000 multi-media computer with an MMX chip, a 21" monitor, and a PCI bus, connected to the Internet by a T1 line. Our actual bandwidth was about 500K/second, appreciably faster than most dialup connections.

L.: How are we going to start this review, Alex?

A.: Starting points ought to be unambiguous, oughtn't they? Shouldn't every building have an obvious main entrance?

L.: Perhaps. But is there a starting point for a web?

A.: Or a review of a Web site?

L.: In theory you can start anywhere in a web, or on the Web, and end up where you started. You browse.

A.: Yes. Browsing is different from searching. You search a database, you browse a Web site. This Perseus Web site (http://www.Perseus.tufts.edu), though, is a database disguised as a Web site.

L.: Perseus 2.0 also exists on CD-ROM, and the CD-ROM version has already been reviewed in the CSA Newsletter ( H. Eiteljorg, II, "Perseus 2.0 - A Review," Spring, 1997). The CD-ROM version is available only for Macintosh, in either a four-CD version or a concise version on one CD. The full version costs $350.00. The Web site has 14,000 images against the CD-ROM's 25,000. Is there some advantage to having the same database, or a reduced set of it, available as a Web site?

A.: It comes down to access, I think. The Web version is platform-independent. It's there on the Internet for anyone who has a computer and a browser. Students and teachers may find the Perseus Web site easier to get to than the CD-ROMs, and if their institution already has access to the Internet, they won't have to make a separate purchase. With the four CD-ROMs in hand, you're likely to spend some time changing CDs. On the Web site, you're limited only by the speed of your connection. And if you aren't a Mac user, the Web site is the only way to gain access to Perseus.

L.: It sounds as though each manifestation of the Perseus database has its advantages and disadvantages. The actual process of browsing the Web site, though . . .

A.: I have to insist on that crucial difference between browsing and searching. Perseus isn't a site for browsing. It's an online database for looking up texts and images from ancient Greece. By linking to Perseus you can create a site, or a selection of texts and images, for browsing, but Perseus itself isn't organized for browsing.

L.: Fair enough. Let's say, then, that the actual process of searching for something on the Perseus Web site can begin with the Perseus Lookup Tool, a screen that allows a global search of the entire Perseus database: Greek texts, English texts, images, catalogs of archaeological sites and material objects, an encyclopedia, and bibliographies. Or you can begin at any of several other points: an English index searches English texts and the art and archaeology subordinate databases; other tools allow searches of catalogs of sites, vases, sculptures, coins, and architecture; there are tools to search Greek texts and lexica as well.

How accessible is all this, really? What about people who don't have access to this expensive hardware here?

A.: Well, Lee, I can give you an example of what it's like to look for something on Perseus without a fancy connection. Much to the credit of the technicians at Perseus, I was able to work at home last night using only a 33 megahertz 486 machine with a 14.4-kilobyte- per-second connection. They have kept the graphics to a bare minimum and use no backgrounds. It's just good old-fashioned raw information.

I wanted to help my wife investigate what her translation of Plato's Phaedo 66b was based on. It used the translation "contaminated soul" for sumpephurmenê psychê, and she wondered whether the participle had any medical connotations in Greek.

The problem of getting a Greek font working on my machine produced all the same emotions as upgrading or reconfiguring any piece of hardware or software. I finally got SGreek to work, though I never succeeded in loading SMK Greek keys. Perhaps this was a function of using a Windows 95 machine instead of a Mac. In any case, the Greek fonts are not produced by the Perseus people, but by independent vendors to which the project links. Caveat Browsor.

Perseus found four instances of my verb in its canon of Classical Greek literature. The authors, texts, editions, and translations included in the canon are available from the Perseus bibliography page. It gave me pause to view through my browser the date "1903" on the Burnet edition of the Phaedo. (Burnet never could have guessed what was to come.) Unlike Burnet's edition, Perseus provides no apparatus criticus. Nor could I find out how textual variants were handled by the search tool.

L.: Perseus does, however, employ a robot parser to look up every instance of the word, regardless of what form the word takes-a significant technical feat. Type in pherô and you will get forms of oisô as well.

Once I had the citations, I could easily go to each, read it in Greek, read it in translation, and look up the word on-line in either the little Liddell, the middle Liddell, or the Great Scott. Very handy. Perseus also produced a word frequency table of all the authors in the Perseus canon. Very impressive. The site offers morphological links for every word of text: click on the word and see a screen where the word is parsed, dialects in which it appears, and its frequency of use in the designated author. The fact that sumphurô is used .05 times per 10,000 words was not very illuminating to me, but word frequency is used to good effect in certain types of scholarship. The fact that my word is only used four times in the Perseus canon and only once outside of Plato (at Euripides, Medea 1199), was interesting. I don't think I could have found that piece of information in any other way.

As it turns out, the word clearly means "to knead" or "to mix," not "to contaminate." I reflected whether perhaps its usage in later medical writers might have contaminated the translator, but I had no way of looking them up since Perseus does not give much space to post-Classical works. The site editors state: "The major authors of the classical period are represented, as well as some later authors whose works are useful for the study of the fifth century B.C. Perseus 2.0 contains the works of Aeschines, Aeschylus, Andocides, Antiphon, Apollodorus, Aristophanes, Aristotle, Bacchylides, Demades, Demosthenes, Dinarchus, Diodorus, Euripides, Herodotus, Hesiod, Homer, the Homeric Hymns, Hyperides, Isaeus, Isocrates, Lycurgus, Lysias, Pausanias, Pindar, Plato, relevant parts of Plutarch, Sophocles, Strabo, Thucydides, and Xenophon." (http://www.Perseus.tufts.edu/aboutperseus.html) Note that we were not successful when we tried to find this quote at the cited URL -- ed.

L.: The limitation to fifth- and fourth-century works-the Classical period, as they say-makes sense, and Perseus has plans to expand into later Greek works and even into Roman material on the same or a related site, I understand. It sounds as though the Perseus Web site can be extremely useful for philological research. Did you look for any architectural or archaeological material?

A.: I did. But first, even though I have tried to avoid technical observations about the site, I must mention the use of what is called the "Perseus Lookup Tool" to generate links. Any page on this site is swarming with links. At first I marveled that someone could go through and key in all those links. But they aren't really links at all, they simply send you to the Perseus Lookup Tool to find all instances of the keyword you clicked on in the Perseus database. We are accustomed to links that lead us somewhere definite, that narrow down our search. These links spread the search.

For example, a page that describes the statue of Kleobis and Biton at Delphi has links for the following terms: "Delphi," "Museum," "Scale," and "sculpture." Follow any of these links, and you get all instances of those words in the database, not, as you would expect, something about this particular sculpture, its scale, its connection to Delphi. The device works well with some links, such as the link "Hdt 1.31" in the subject description which takes you to the exact spot in Herodotus where the statues are mentioned. I was delighted at this. But linked words such as "statue" and "island" are useless.

My pursuit of Kleobis and Biton ultimately proved disappointing. Although the statues are fully described and catalogued, I could not find pictures of them. No doubt, copyright issues are holding up development of visual materials. Entering the site via the "Art & Archaeology" button will take you to the pictures that are currently available. It seems that cataloguing was done first in the hope that permission will eventually be procured for the release of most if not all the images catalogued.

The copyright issue is interesting and pretty clear. You cannot copy the image files legally, but you can link to them. If you are producing a hypertext research paper, you link to sources when you cite them. (And if you try typing a URL in Word '97 it automatically turns it into a link, which will automatically fire up Internet Explorer and take you there when clicked - an interesting commentary on the new nature of citation.) In the "old" way of doing things, this would be the equivalent of carting in a dolly of library books feathered with yellow stick-'ems to accompany each research paper. The hypertext research paper is not yet widely used; so not many will be able to take advantage of the images in Perseus. Will a student, for example, be able to print on paper Perseus' picture of a site map of the Acropolis with the citation, "http://www.Perseus.tufts.edu/cgi-bin/image?lookup=PS0024&type=plan?"

Issues of authorship and authority are still being clarified on the Internet. You can't put a blue cloth binding with gold insignia on a Web page. Perseus documents all the texts, editions, and translations it uses. Artifacts and images as well as texts are clearly catalogued in the manner to which Classicists are accustomed. I did come across a few entries in something called the "Perseus Encyclopedia" which seemed authorless, although they had good citations and bibliographies.

What more I would like to have found, and what may be possible in the near future, would be something like this: the text of the Kleobis and Biton section in Herodotus, a link to some commentaries which would have pointed out the existence of the statues, and then a link to images of the statues. Technologically, Perseus can already do this. It is just a question of gathering the materials.

L.: And of course any user is free to gather the materials from Perseus, in the form of hypertext links in a document he or she creates.

A.: Exactly. If funding continues, the Perseus project could eventually surpass many university libraries in quality and scope, making classical research a cheap, one-stop shop for anyone who has a computer and a connection to the Internet. What a way to subvert the social order.

L.: Interesting that you should end up wanting Perseus to be a site that taught you something or showed you something about Kleobis and Biton, since you insisted on the distinction between browsing a Web site and searching a database. With a few exceptions, like the page on the much-hyped Hercules that Perseus featured on its index page this summer, the Perseus site does not set out to provide the kind of guidance that you describe. It's not a site for browsers, remember. It's an on-line database. The user has to structure the inquiry. As you say, it's a library, not a class.

A.: This doesn't seem much like a book review, Lee.

L.: It shouldn't. A book is a finished product, and a reviewer can describe the same thing that readers of the book will find when they buy it or check it out from their library. We've been talking about the July version of the Perseus Web site. By the time this review sees print, that version will be archived somewhere, and searchers who point their browsers to http://www.Perseus.tufts.edu will find something a little, or perhaps a lot, different from what we saw.

For an index of other CD and Web site reviews available on the Web pages of the CSA Newsletter, see the review index.

Next Article: Reconstructing the Theater of Dionysus in Athens

Table of Contents for the Fall, 1997 issue of the CSA Newsletter (Vol. X, no. 2)

Master Index Table of Contents for all CSA Newsletter issues on the Web

Go to CSA Home Page