Advertisement

Ancient Greek Made Easier : Databank: A marriage of the computer and classical literature helps researchers produce comprehensive, higher quality work more quickly than ever before.

Share via
TIMES STAFF WRITER

Ted Brunner puzzled over the Greek letters written on a tiny fragment of papyrus unearthed in Egypt.

Once it would have taken a decade or longer to sift through thousands of Greek texts for a hint to its origin. But last summer, all the UC Irvine professor of classics had to do was search his electronic databank, a 65-million word repository of more than 9,500 works by nearly 4,000 Greek authors from the time of Homer in 750 BC to AD 600.

“It was a real yummy,” the ebullient, German-born professor confided. “It turned out to be a fragment of Euripides’ play ‘Andromeda,’ for which no text has ever been found. . . . That could never have happened without the databank. No way!”

Advertisement

Brunner, 57, has devoted the last 20 years as director and builder of a modern-day Thesaurus Linguae Graecae. This marriage of the computer and Greek literature is a watershed for classical scholarship, making it possible for researchers to produce comprehensive, higher quality work more quickly than ever before. Even graduate students at universities with access to the TLG computer discs can now tackle topics that would have taken experienced scholars half a lifetime.

“It was a dream come true for classics scholars,” said Harvard University professor Gregory Nagy, president of the American Philological Assn., a national organization of Latin and Greek scholars.

“It makes it possible to study all sorts of questions that people wouldn’t have even asked before the TLG,” explained Nagy, an expert in Greek mythology and epic poetry who will speak at TLG’s 20th anniversary lecture and open house on Friday.

The UCI-based institute also has spawned a host of similar endeavors, including an effort at a UC Berkeley project to computerize a thesaurus of Buddhist texts. At Harvard, TLG users have now developed a multimedia computer project know as Perseus, which offers Greek texts, drawings and photographs of ruins, as well as ancient art for use as an undergraduate teaching tool.

“In the future, this is the way most people will have access to the major primary materials to study cultures,” predicted Donald J. Mastronarde, a professor of classics at Berkeley who has used TLG for several years.

Reflecting on his group’s pioneering effort to meld classics with computers over the last two decades, Brunner grinned.

Advertisement

“The day I die, I will say this has all been worthwhile,” said the chain-smoking, self-described former computer illiterate. “It has been the most incredible opportunity because this material is the beginning of Western civilization. And we have developed a tool, the means to understand and analyze it.”

The first Thesaurus Linguae Graecae was unveiled in 1572 by Stephanus, a member of a prominent European publishing family. Within a few years, though, his four-volume set was obsolete, outpaced by the discovery of more ancient texts and done in by abridged versions published by competitors.

Late 19th-Century classicists were determined to fill the void. A thesaurus of Latin literature was launched first in 1894 in Germany. Since there were only 9 million words, scholars figured the project could be completed in a decade. Yet the work still goes on, and it is now targeted for completion in 2025.

Since the output of ancient Greeks was estimated at 10 times that of Latin literati, wary Greek scholars began to regard a definitive treasury of Greek literature as an impossibility.

That changed on March 23, 1971, when graduate student Marianne McDonald, daughter of the founder of the Zenith Corp., offered then-UCI classics department chairman Brunner $1 million to create an electronic thesaurus of Greek.

“I was obviously sorely tempted to say, ‘Of course!’ But what I said instead was, ‘No,’ that I would take $25,000 and one year to study whether it was possible,” Brunner recalled.

Advertisement

McDonald, now an adjunct professor of Greek drama at UC San Diego, said the idea was borne of her own frustration.

While writing her doctoral dissertation on the concepts of happiness in the plays of the Athenian tragic poet Euripides, she used a concordance with dictionary-style definitions and some examples of sentences. What McDonald wanted was a search of every one of the four separate words used in ancient Greek for happiness . What she wanted was a computer that could sift through material quickly and still allow her to read the full context in which each word appeared.

UCI was barely 6 years old at the time, and its classics department, however promising, was still young. But McDonald chose UCI, in part, because “I needed it right then, so it was my greed to have this tool for myself,” she said. “And Ted had that wonderful inventive vision.”

Twenty years and more than $7 million later, the TLG has surpassed it’s original goal. Definitive texts of everything from the epic poems, the “Iliad” and the “Odyssey,” to the early Byzantine period of medieval Christian culture now are housed in the memory of two 440 megabyte, disc-drive computers at the UCI campus.

Already, Brunner and his team of nine are at work on Phase 2: Greek writings produced from AD 600 to the fall of Constantinople in 1453. Before the end of the decade, Brunner believes that TLG could conceivably include all the important works of history, jurisprudence and writings tracing the development of the orthodox church.

“I’m terribly pleased by how it has turned out,” said McDonald, who uses the TLG computer disc at her Rancho Santa Fe home to prepare courses or provide research for her recently published book, “Ancient Sun, Modern Light: Greek Drama on the Modern Stage.”

Advertisement

During his year of research before accepting McDonald’s grant, Brunner learned the standard method of thesaurus compilation: each word is individually entered on a slip of paper, alphabetized, then painstakingly analyzed and cross-referenced in voluminous detail.

“I realized that by God, it could not be done the old way. All you have to do is look at the people still laboring away at the Latin project,” Brunner said.

Trouble was, Brunner knew next to nothing about technology. “My big problem was that I could not think of a computer as dealing with anything but numbers,” he said.

By July 1, 1972, when Brunner accepted the grant, he had learned enough to buy two computers and hire a systems analyst from Dartmouth College to develop an encoding system to enter Greek letters and accent marks into the computers.

Brunner then hired a key punch operator. “That same evening, I stood over her, and it became clear to me that it would take hundreds of years to do it this way.”

Next, Brunner and his team turned to commercial data-entry firms. They tried companies in Greece, Singapore and South Korea, among others, demanding that they have no more than one error per 25,000 letters. Ironically, the Greek firm was the first to back out of the project because the data-entry clerks were too familiar with modern Greek to meet the accuracy targets in the classical tongue.

Advertisement

A South Korean firm eventually got the contract and began producing Greek texts on magnetic computer tape for TLG in 1973. This fall, TLG shifted to a data-entry firm in China. Brunner figures that it has cost $0.0002 per keystroke to compile the 65.5 million words.

One key to TLG, Brunner said, has been his long collaboration with a former Stanford University schoolmate, David W. Packard Jr., another scion like McDonald who was bitten by the classics bug and went on to get a doctorate in the subject.

Packard, whose father founded Hewlett-Packard Co., brought with him his knowledge of computers and eventually helped to design a personal computer called Ibycus and the software specifically for TLG data searches. (Packard and the Packard Humanities Institute in Los Altos, Calif., recently finished work on a Latin thesaurus that will be available to scholars in the same Ibycus format.)

Gradually, others have developed software to make the data work with other personal computers, especially Macintoshes. Now, two young computer experts are nearly finished with what they hope will be a fast and efficient software package for IBM-compatible computers.

Brunner’s chief duty is to raise money for the thesaurus project, which costs about $1,000 a day to operate. A big challenge will be to keep the technology current. And that will take money.

“Unless a way is found to keep the system up to date, it will turn into an antique curio, like those books,” Brunner said, gesturing to leather-bound volumes of the Stephanus thesaurus. “A book is like a jar of pickles; it will stay on the shelf for a long, long time. An electronic databank will fall apart or become outdated if it is not . . . kept current.”

Advertisement

The University of California now provides about $100,000 of that amount annually. To date, the National Endowment for the Humanities has given $3.8 million in support. The project also has received donations from the David and Lucile Packard Foundation, the Mellon Foundation and many other smaller donors.

McDonald, a prominent philanthropist, is active on the group’s executive board. And despite her new project of giving a Dublin university a similar $1-million fund to develop a computer thesaurus in Gaelic, the native tongue of Ireland, she has donated nearly another $1 million to TLG to ensure its survival.

“It’s enormously valuable,” McDonald said. “Now, I just hope it can continue and expand.”

As Brunner considers the round of fund raising and the seemingly constant haggling over copyright issues raised by this new electronic technology, he finds himself drawn to the library to steal a few minutes at the research he loves.

The “Andromeda” find, from a photographic copy of the fragment of papyrus discovered in Egypt, remains a source of delight.

“It still sends shivers down my back,” Brunner said.

But those days are few and far between.

“It’s funny. Here I am, building it, and yet I have to steal time to work with it,” he said wistfully. “I could write an article a day if I had the time.”

Instead, he speaks of the need to find an energetic, computer-minded classics scholar who will continue building TLG.

Advertisement

“I just don’t want to face the prospect of looking down from Heaven or looking up from Hell and see the project croaking,” he said. “I’ve invested too much of myself in this databank.”

Advertisement