It's a cheery thursday morning at the Manhattan headquarters of Merrill Lynch. The Wall Street Journal has just announced the brokerage firm's takeover of London money manager Mercury--the biggest overseas acquisition ever by a U.S. securities outfit. So how exactly is Merrill chairman and CEO David Komansky celebrating? By discussing Ph.D. dissertations with a small group of reporters, of course.

The occasion is Merrill Lynch's introduction of the Innovation Grants Competition, which will reward doctoral students who can persuasively argue for the market potential of their dissertation research. (Top prize: $50,000.) "The Ph.D. dissertation represents a very undervalued asset from the business point of view," intones Komansky at the orange-juice-and-pastries press conference. After all, he points out, students have launched a great many entrepreneurial ventures in the past--from FedEx, the germ of which was contained in founder Fred Smith's senior thesis, to Netscape, a product of Marc Andreessen and Jim Clark's undergraduate research at the University of Illinois. According to MIT's Technology Review, the combined assets of all the companies spun out of MIT would exceed the gross national product of most nations.

Given the potential profitability of dissertations, it's easy to speculate that the first-prize incentive for entrants is nothing compared to the possible incentive for the investment bankers themselves. "We are not approaching this as a commercial endeavor on our part," Komansky demurs. "But clearly, if something were to germinate from it that were attractive to us as an investment, we would want to participate."

It's no surprise that any investment banker worth his salt would want the early inside scoop on something like Stanford-born Hewlett-Packard. But the award applies to all disciplines in the arts and sciences. Is there really any reason why Merrill Lynch would be eager to get its hands on, say, the latest exegesis of Wittgenstein? The competition's director, MIT Media Lab research associate Michael Schrage, has an answer. "The student doing a thesis comparing the semiotics of Derrida, Nietzsche, and Heidegger might not think the work has market potential," he says. But think again. Schrage points out that some successful software has been designed based on a Heideggerian interpretation of the role of language: Terry Winograd and Fernando Flores, designers of a computer program called the Coordinator Workgroup Productivity System, write in Understanding Computers and Cognition that Heidegger's discussion of "thrownness" and "breakdown" provides a crucial design logic that artificial intelligence experts ignore at their peril.

Indeed, Schrage muses, even a dissertation analyzing the nature and role of discourse in Habermas, Husserl, and Foucault could be parlayed into an interactive technology product. "What are computer networks all about?" says Schrage. "Discourse!" And what better proof than the World Wide Web for the postmodernist assertion that the reader shapes a text as much as its author? "It's not that bold or creative a leap," he explains.

Then again, it's unlikely that the judges--a panel of nine entrepreneurs, including cyberpundit Esther Dyson, Xerox chief scientist John Seely Brown, and outgoing Rockefeller Foundation president Peter Goldmark--will be slogging through mountains of dissertations on Husserl. Enthuses Schrage, "The genius of this proposal is that we don't even want to read the theses." Applicants are instead required to submit only 3,000 words--250 of which will summarize their dissertation research, while the rest go to explicating its most significant commercial idea and detailing how that idea might be brought to market as a product or service. Part of the competition's purpose, he says, is to "make even cutting-edge knowledge graspable to those who would otherwise have no interest in it."

Although Schrage concedes that a dissertation on Dante would be a hard sell, he hasn't given up on Shakespeareans. (Donald Foster, after all, is an Elizabethan scholar who, through computer-assisted analysis, pegged Joe Klein as the author of Primary Colors. He has since been retained as a consultant to police departments.) Schrage can also imagine a political scientist who, influenced by University of Michigan game theorist Robert Axelrod, might devise "the Lotus 1-2-3 of negotiations software."

When pressed for more examples from the humanities and social sciences, Schrage cites Frank Sulloway's Born to Rebel. "To put it in the most crass possible way," he says, Sulloway's conclusions about the relative conservatism of firstborn children and the unconventionality of later-borns could be used "as the basis of a direct-mail campaign." Stodgy solicitations for Reader's Digest subscriptions could be sent to firstborns, for example, while hipster publications like Wired could laser in on their kid sisters.

David Komansky, after introducing the other speakers and providing an overview of the project, has passed much of the press conference quietly in the corner of the dining room. Maybe it's the afterglow of the Mercury deal, or maybe it's the tantalizing thought of all that untapped intellectual capital that has brought a satisfied smile to his lips.

"Interesting," he interjects in response to Schrage's Sulloway idea. "What you call crass, I call practical."

LAURA SECOR