Closer to Du Bois' original vision, Gates, Appiah and Soyinka wanted their encyclopedia to be inclusionary, one that would bring in as many disciplines, cultures and perspectives as they could. "You have a diasporic black world, and the only way to put it back together again is symbolic. It's like Humpty Dumpty," Gates told Salon during a four-hour discussion. (Gates insisted that his comments be used only on a deep-background basis; all quotes used in this story had to be cleared with him.) "Whoever could edit the Encyclopedia Africana would provide symbolic order to the fragments created over the past 500 years. That is a major contribution."

Joining a trend of joint ventures between academics and corporations, Gates and company looked outside the university for funding. After a decades-long search for a backer, they landed a big fish -- Microsoft. The company agreed to put up $1 million and technical, design and production support of the CD-ROM. Another million came from Perseus, a publisher, which would produce a printed version of the encyclopedia. After much heartache -- and 25 years after the pact made at the restaurant -- the interactive CD-ROM encyclopedia, Encarta Africana, arrived on the shelves to near-unanimous praise last February.

Yet while the Africana debut has seen no dearth of press coverage, the story of its painful gestation has never been told. And it is a tale so fabulous -- including accusations of plagiarism, controversies over affirmative action, charges of academic royals oppressing scholarly serfs, Microsoft bottom-line mania, worker revolt and a Rashomon-like tangle of competing truths -- that one might even call it a creation myth for the contemporary university.

In today's university, academics in the science and technology departments can easily turn a profit from their intellectual work. With the help of a technology licensing office on campus, an academic can become an entrepreneur, often collaborating with a company to distribute his or her invention/discovery in the form of a marketable product.

More and more often, humanities professors like Gates can do the same thing. An illustrious academic -- who in the hierarchy of the new university star system has become something akin to royalty -- descends from the ivory tower to share his or her intellectual wealth with the public. But reaching the greatest number of people requires more than brilliance, it requires power. So the academic decides to work with a corporation. This transports the academic into a brave new world, where the old rules of the ivory tower no longer apply. There, intellectual property must be guarded fiercely; workers labor under strict deadlines with their eyes focused on the bottom line. Books must balance at the end of the year; there is no alumni fund to draw from when the project goes over budget. Still, the professor and the project often gain by their affiliation with the university -- there are laboratories, research materials, libraries, offices, computers and staff willing to help out in a jam. More importantly, there is a deep pool of well-educated cheap labor in the form of ambitious young graduate students and untenured academics. In short, though the project may have been driven by idealism, it quickly becomes subject to all the complexities and compromises of a business.

On a logistical level, Africana's 15-month gestation was fraught with complications. Grueling deadlines led to overexertion. Editors contended with bouts of plagiarism. Open revolt broke out among nearly a dozen staffers. All this turmoil eventually led to a crippling worker slowdown in the middle of the project. Most startlingly, a very low representation of African-Americans on the core editorial staff (four of 17 writers at the most) inspired a dozen employees to ask the management team to hire more African-Americans. "There's nothing like a common enemy to unite a lot of young, similarly minded people," one senior staffer told Salon Books, echoing a sentiment shared by several others. What made this project so beset with bitterness?

Perhaps it was the great expectations of Gates the public humanist. Though Gates is only one of four Afropedia partners, he is Africana's front man and biggest personality. Politically, Gates is neither radical nor neo-conservative, but rather a thoughtful, learned voice of black progressive liberalism who has consistently been able to translate his intellectual ideas into books and articles geared toward non-academics. Pundit Adolph Reed Jr. called Gates "the freelance advocate for black centrism," while Time once voted Gates one of the "Twenty-Five Most Influential Americans." An outspoken advocate of affirmative action, the 49-year-old West Virginia-bred Gates also has voiced his concerns about the responsibility of corporations to soften capitalism's rougher edges. "A more humane form of capitalism is about the best I think we can get," he told the Progressive last year. "Which might sound very reformist or conservative, but that's basically where I am." While the notions of better business practices and affirmative action may mean different things to different people, some Africana employees told Salon Books that when it came to working for Gates the CEO, they encountered a split between Gates' progressive theories and Africana's bottom-line practices.

The idealistic project was beset with birthing pains. As an associate professor at Yale in 1979, Gates proposed the idea to Charles Van Doren, the Encyclopedia Britannica editor who was infamous for the "Quiz Show" controversy in which he and NBC were involved in a fixing scam. Van Doren, who had since climbed his way to the position of vice president of editorial at Encyclopedia Britannica, took an interest in Gates' proposal. Portentously, however, Van Doren told Gates the project would cost $20 million. Though Van Doren remained a consultant for the group, Gates and Appiah soon were looking elsewhere for support. Eventually they would be rejected by publishers at Random House, Simon & Schuster and by new-media companies Voyager and Prodigy. So in 1997 when Microsoft agreed to underwrite the project, the team jumped at the opportunity.

Genetically speaking, Encarta Africana had some formidable DNA. Gates, along with Appiah and Soyinka, linked forces with musician and media entrepreneur Quincy Jones and Time Warner lawyer Martin Payson (the only white person on the board) to start a private company, Afropedia L.L.C. Gates and Appiah would co-edit the content, and Microsoft, which had two Encarta encyclopedias' worth of experience behind it, would integrate the content with the latest audiovisual technology, the costs of which would go on Microsoft's ledger. With Gates of Harvard and Gates of Microsoft leading the way, the project seemed like an inspired marriage of new technology, academia and progressive thinking.

But the $2 million budget was tiny compared to Van Doren's $20 million estimate. It's possible that attempting a project of this size with such modest funding was a mistake from the start. Since the budget has not been made public, it would be difficult to guess. However small the payment, Africana received its advance only after the editors submitted a specified number of words. In other words, the two sides agreed on a book-publishing model in which advances were to be paid back through royalties. In terms of finances, Afropedia didn't have a lot of breathing room. Insiders claim that co-editors Gates and Appiah received $100,000 each off the top, but this is not a particularly lavish fee for heading up such an undertaking.

Despite the formidable profiles of Gates and Appiah, however, the project was still steeped in uncertainty. For one thing, Microsoft partnership was a new one for all involved. Microsoft had never had a joint venture of this nature. It was ultimately responsible for supervising the project, yet Afropedia had the final say in the content. Meanwhile, the leaders of Afropedia had their own worries. As a collection of academics, they were venturing into unknown corporate waters. "We didn't have a model for this," says Pat Sullivan, Africana's managing editor and an Afro-America studies lecturer. "No one had ever done this before." Encarta may have been called an encyclopedia, but it functioned with the structure of a magazine working on a book-publishing deadline. Unlike most encyclopedias, Encartas are completed at a breakneck pace -- French Encarta and Italian Encarta were both finished in one year.

In order to achieve its staggering goal of 2.25 million words, Africana had three models of encyclopedias on which to base itself: the Samuel Johnson Dictionary model, in which one woebegotten genius writes the whole thing; the Encyclopedia of Islam model, a drawn-out process that involves only experts; or the Encyclopedia of Social Sciences model, which had a core staff that wrote everything. Africana synthesized the latter two: a core crew of writers who would generate 40 percent of the content. Experts would contribute the remaining 60 percent. A few intellectual stars would oversee the whole process.

To lure writers, Africana posted advertisements around Cambridge offering 15 cents a word to temporary writers. The writers who signed up on salary were mostly students between undergraduate and advanced degrees and would receive somewhere between $25,000 and $28,000; editors, some of whom were Harvard fellows or professors, received salaries in the $50,000 range. (For a university town, Cambridge can be costly. One-bedroom apartments generally rent for around $1,000 a month, which would be half a writer's salary, before taxes.) Writers would have benefits packages only if they had a previous arrangement through Harvard as a student. Editors, if they were not already covered by Harvard, did receive benefits packages.

According to Sullivan, who hired many of the writers, management told writers initially that Gates and Appiah would switch off semesters to take their leaves. For most of the first semester, Gates went to Africa, where he was filming a BBC-PBS program, "The Lost Wonders of the African World," portions of which appeared in Africana. Appiah left for the second semester. Africana's core staff gathered each day on 10 Divinity St. at Vanserg, the Harvard building where Afropedia rented space. Harvard refused to say whether the Afropedia project received a special cut-rate rent. A dim building with a makeshift, temporary feel, Vanserg rents to pot-luck tenants such as classes from divinity-school overflow, a piano repair shop, a day-care center and the Harvard-Yenching Institute.

Problems in Vanserg developed within the first two months in the summer of 1997. When some writers discovered that editors were given medical benefits, disenchantment grew. Stephen Hendricks, an Africana writer and a former Microsoft employee, draws a parallel between Gates of Harvard and Gates of Microsoft.

Henry Louis Gates "told his employees that the lack of benefits was not something that he dreamed up, that such is the modus operandi of the software industry, in which growth is fueled by disposable, long-term temps," Hendricks told Salon Books. "Certainly this is at least partly true of Microsoft, from which Mr. [Henry Louis] Gates took his cue."

On one level, the analogy makes sense. With low-paying jobs and few benefits, the young Africana writers were being given their first taste of life in the modern corporation. Some Africana employees believe that Africana had learned the purported "permatemp" style of management from Microsoft. (One-third of Microsoft employees don't receive long-term benefits, even though they work for the company on a long-term basis.) On another level, however, the arrangement also resembled the new model of university labor, which has become notorious for its unabashed exploitation of non-tenure-track scholars, researchers and faculty. A common complaint from the ghetto of adjunct teachers is that they are paid far worse per hour than the janitors that clean their offices; moreover, the janitors often enjoy unions and job stability.

Since Encarta Africana was privately owned, initially some writers did not have Harvard privileges and could not access the university's vaunted libraries. Instead, they had to make due with their limited resources -- the small African-American Studies department library in Barker Center (a 15-minute walk from Vanserg), Macmillan Company's Encyclopedia of African-American Culture and History (edited by Africana contributor Cornel West) and Lexis-Nexis, which some staffers say was accessed by a password from a Yale student.

With a large deadline looming at the end of August, the staff grew desperate. Writers had to meet the first milestone by August, a numbing 250,000 words that the group had to submit to Microsoft so that Africana would get its first advance. "It was as if they said, don't worry about quality, just get quantity," the senior employee said. According to one observer close to the project, the writing turned out to be lousy. "I was appalled by the quality of material. The entries were woodenly conceived. They had linear chronologies," he said. "These writers were not very experienced. They were at the low end of the freelance chain." When writers turned over their sources, an editor discovered that some entries were barely rephrased versions of the entries from the Macmillan. In the end, Africana had to hire a temporary staff to rewrite the plagiarized sections, all of which were purged and replaced. "There were huge, huge mistakes that never would have eluded Skip [Gates] had he seen them," the senior staffer contended.

When Suzanne Friedberg arrived in September to edit the Africa section of the encyclopedia, she inherited a troubled legacy. "Everything was done in extreme time pressure," she said. Friedberg, now a professor of geography at Dartmouth College, said that the deadlines were daunting. "I'm on the tenure track now at an Ivy League school, and the encyclopedia was harder." Even though some writers were nabbed for plagiarism, they still simmered with discontent. "Some writers stirred up trouble," Friedberg said. "The medical insurance concern -- that was legitimate. But some bad writers just liked to complain."

In a controversial move among core staff members, Africana received permission from Harvard to hire Peter Glenshaw, the assistant director of the W.E.B. Du Bois Institute, the same place where Gates is director. (Glenshaw, who has since left Harvard, did not respond to Salon Books' calls.) While keeping up with his arduous duties with the Du Bois, he had to manage the content and employee relations of the encyclopedia. As a friend or admirer of Gates he may have agreed to the job out of respect for the project, but as Gates' deputy at Du Bois, how much latitude did he actually have? He now had to commute between two jobs.

As the next 250,000-word milestone approached, editors' hours increased, as did those of the writers. Twelve-hour days were the norm, along with weekend shifts. According to Friedberg, there was a huge editorial bottleneck: the work was lined up to be edited, but the pieces needed a great deal of work. Writers were disappointed that they didn't receive feedback, and editors didn't appreciate the quality of writing. Africana had itself a situation that was headed for internal combustion.

On Oct. 3, 1997, writer Hendricks and 11 other staffers sent a memo to Gates, Glenshaw and the rest of the management team suggesting significant editorial and personnel changes for Africana. The memo specifically asked Africana for a clearer mission statement, benefits for writers -- which would include medical and dental coverage -- and an employee-matched retirement plan. As Hendricks explains, "We wanted to hold it to a higher standard." Employees also wanted more specialists, and last of all, suggested that Africana hire more people of color.

Some editors now maintain that had Africana spent $10,000 more on writer's salaries, it could have hired more seasoned candidates and tempted more African-American scholars in the process. Among writers on the core staff, African-American representation never reached more than four out of 17 and none of the core editing staff was black. "We took a look around and said, 'Jesus, we're 90 percent here and we're not comfortable with this,'" said the senior staffer about the paucity of blacks. "I wouldn't buy an encyclopedia about women if it were written by men." It seems that despite Gates' formidable reputation, few blacks applied to work on Africana. Moreover, the editors whom Salon Books spoke with say that they were never given any directive by Gates to pursue African-American applicants. Part of the affirmative-action agenda involves seeking out applicants who have been previously denied opportunity.

Yet making an airtight case of hypocrisy against Gates isn't so easy. How far does an employer have to pursue it? How much time, for instance, should be spent searching for diverse candidates before such searches are deemed inefficient? And if such searches don't yield competitive candidates, how important is it to give an opportunity to a worker who is not qualified, but might rise to the occasion if given a chance? Still, it's remarkable that Gates -- a black luminary -- wasn't simply surrounded by bright, ambitious young African-American scholars who could foresee what their participation on this project might mean to them or their risumi.

Does the fact that Gates was somehow stymied by the problem of affirmative action hiring say more about him, the shortage of highly educated black candidates willing to work for peanuts or the very problems inherent in affirmative action itself? After all, it may be an easy practice to embrace in the abstract, but when you're running a fast company, who has the time for theory?

For meetings, Gates would have the staff convene in Barker Center, a humanities building where the Afro-American Studies Department is based. Indeed, every staffer interviewed for this story contended that they never once saw Gates step foot in the central Africana office in Vanserg itself, despite the fact that Gates' own house is right across the street. Not surprisingly, such aloof management only exacerbated worker resentments. In October, he drew the staff together and gave them an ultimatum: For those who don't like the project, there's the door. Gates also said that plagiarism would not be tolerated. The Oct. 3 petition was rejected out of hand -- and on the delicate issue of hiring more African-Americans, Gates apparently told staffers: "Affirmative action? I'm Mr. Affirmative Action. You think I'm not all for affirmative action? But look, what we hire here are qualified people, people who can do the work. White people can do this work, and black people can do this work."

Relations on the Africana staff deteriorated further with the arrival of Microsoft project manager Keith Senzel in November. Senzel, a 26-year-old from Yale who graduated a year behind Hendricks, had a mandate to improve efficiency with the greatest attention to quality. Or, in other words, whip the staff into shape. Senzel had fulfilled this kind of role for Encarta for its French and Italian CD-ROMs, both of which were completed in one year. "He was a Microserf before he got made," Hendricks said of his old schoolmate. Senzel said that compared to other Encarta products, Africana did not have an unusually rigorous set of deadlines. "There's a pretty aggressive schedule on all of them," he said.

Nonetheless, Senzel found some difficulties. "They hired so many people all at once and so many of the people were not that good," he said. "The writers were not very good. They couldn't write in an encyclopedic style." Yet some staffers told Salon Books that Senzel talked down to the writers. "We were like, 'Who are you and where are you from?'" the senior employee recalls. "Their attitude was, your content sucks." When it came time for the staff Christmas party, Senzel received a last-minute invitation but did not attend. During the party's Secret Santa exchanges, Peter Glenshaw -- who had won no popularity points as Gates' henchman -- received Dale Carnegie's personality primer, "How to Win Friends and Influence People," from an anonymous staffer. Glenshaw left the party soon after.

Even though the Oct. 3 memo was a flop, employees received a $500 Christmas bonus. They also received a one-week vacation. "It defused the tension a bit," someone on the project recalled. But after the break the malaise intensified. A new system of editing software was implemented, which slowed down the process. While Microsoft contends that the program, E-Edit, was an improvement, some staffers contend that it was a bad replacement for some new technology that didn't roll out on time. Production fell off by an estimated 80 to 90 percent.

When Gates returned from a trip in April, he asked for three staffers to meet with him -- a group he facetiously called "the bitch-and-moan committee." As a result of the negotiations, Gates implemented a new incentive program for the writers: If they exceeded the 2,500-word-per-week minimum by 80 percent they would receive a bonus equal to 5 percent of their weekly salaries. If they exceeded by 100 percent, they would receive a 15 percent bonus. Although this amounted to minuscule sums for most writers and just made the work harder, it successfully disarmed the writers and set Africana back on track. Gates found a silent fifth partner to fund the peace treaty.

Pat Sullivan had her own take on the strife. "You have to be committed to the idea of the project. Only that could get you through the rough months." Indeed, Africana seemed to punish its staff with the rigors of a high school yearbook in which the editor or editors receive most of the glory. "Africana was very good about what it accomplished with the text," says Friedberg. "It was the first thing of its kind. It filled a hole. In that sense I'm glad I worked on it. It was just riven by bad management."

Was that all it was -- a case of bad management by admittedly neophyte corporate executives? From an internal point of view, Encarta Africana seems like a collision of too many interests, a confluence of paradoxical intentions. "It is an excellent example of the broad convergence of the recent trends, all sinister: target marketing, identity politics and the academic star system," observes New York University media professor and corporate critic Mark Crispin Miller. If an ivory tower company reaches out to fresh-faced graduates for employees, it will face the culture it wrought. If you start a company at Harvard and need some sharp young workers, no doubt you will hire egalitarian idealists nurtured on the idea of multiculturalism -- and conversely, a sense of their own entitlement. Another paradox: If this joint-venture has a target market for African-Americans, is it hypocritical or democratic that it is not prepared by African-Americans?

Even among the so-called target market, not everyone is pleased about the outcome of Africana. Dr. Raymond Winbush, a professor at Fisk University, Du Bois' alma mater, had trouble with the product. Winbush takes exception to what he sees as the encyclopedia's underrepresentation of Afrocentrics -- a large black constituency who believe that Africa should be the focus of all African-American study. "It's like a modern history of computing without mentioning Bill Gates," he says. Winbush is working on a rival project, Encyclopedia Africana, due in 2009, which he claims has its real roots in Du Bois' ultimate dream.

Masao Miyoshi, a literature professor at the University of California at San Diego, sees the Africana model as symbolic of a greater problem. "This is simply outrageous. When you work for a professor, you get paid -- what -- $10,000 or $20,000 a year for a 40-hour week? But this project exists outside the university. No claim could be made that this is an instructive program. Many people with skills and graduate degrees who don't have jobs gravitate towards these kind of work," says Miyoshi, who edited "Cultures of Globalization" with Marxist critic Fredric Jameson. "That's where Skip has an opportunity. This is the kind of thing that goes on today at universities."

Despite the 15 months of upheaval, the interactive CD-ROM changed the stodgy image of an encyclopedia. Rather than an old-fashioned A-Z set of volumes, it features live video, virtual tours and more than 3,300 entries. Africana's arrival was greeted with great exultation. "I wanted to hug the FedEx guy," a Time writer gushed in her review. Gates also seems happy with the way Africana turned out. "I'm really proud of this product. I would do it the same way over again." As universities and their star professors become more involved with these kinds of quick and furious projects, and more overeducated young men and women float in postgraduate limbo, no doubt Gates will still have many willing takers for the next time around.