Almost everyone knows about the Radcliffe Fellows. These scholars, artists, writers, and scientists — 45 to 50 a year — spend two semesters of intellectual exploration at Harvard University, sponsored by the Radcliffe Institute for Advanced Study.

Key to the experience is the fruitful clash and collaboration that results when so many disciplines are brought together. This intellectual commons has been a feature of the institute since its transformation from a college in 1999.

Fellows this year will explore a broad range of interests, from experimental poetry and holographic algorithms to the locomotion of mollusks and the physics of falling leaves.

Less well-known are Radcliffe’s Exploratory and Advanced Seminars, closed interdisciplinary workshops that have been in place since 2002. They are a fellowship experience in miniature — intensive investigations of emerging topics by small groups over the course of just a few days. Harvard faculty and former Radcliffe Fellows are invited to submit proposals for seminars they would like to offer and lead.

Exploratory seminars are a first look at what cross-discipline collaborations might engender. Advanced seminars usually result in a paper, a book, or some other substantial representation of scholarship.

“We open conversations,” said Barbara J. Grosz, dean of the institute and Higgins Professor of Natural Sciences at the School of Engineering and Applied Sciences. “The goal is to seed and jump-start new collaborations.”

The E&A seminars, as they are called, begin with streamlined but competitive applications. “We’re looking for the excitement of an idea,” said Grosz. Selections are made every April for an annual funding cycle that begins July 1.

Seminar groups are small — rarely more than 20 — and usually include invited experts from outside Harvard. The focus is on scholarly inquiry that stretches boundaries beyond a single academic specialty. “Truly successful interdisciplinary collaborations,” said Grosz, “start small.”

“We want the faculty to focus on intellectual content,” she said, while Radcliffe handles every logistical detail. “We’re a convening force. We want our fellows to hit the ground running. And in the same way we want everyone in these seminars to just come and do their work. We make it happen for them seamlessly.”

Radcliffe approved 15 seminars for the 2007-08 season. Topics are broad and sometimes surprising, including sessions on women’s interfaith initiatives, the war on drugs, deep time, and the promise of mycorrhizal fungi as organic fertilizers.

Two seminars have taken place already, including an exploratory one on China’s role in the making of European modernity. From July 31 to Aug. 4, a small group of international experts in literature, history, art, and economics convened at Radcliffe.

Organizer Eun Kyung Min RI ’05, who teaches literature at Seoul National University, wrote that the seminar opened up “new dialogues across disciplinary boundaries.” The future may bring a specialized conference or a book she said. And editors at the journal Eighteenth Century Studies have already pledged a special issue inspired by the cross-disciplinary seminar. “The intellectual sparks that started at Radcliffe, I hope, will start a fire,” Min wrote.

In an exploratory seminar this November, Harvard’s Patrick J. Wolfe — assistant professor of electrical engineering on the Gordon McKay Endowment — won’t be starting any fires. But he and a small interdisciplinary group of scholars will be adding the heat of inquiry to a puzzle that has eluded engineers, computer scientists, and statisticians for years: a fundamental mathematical way of modeling natural sounds, especially speech.

A probabilistic model of this sort could make speech-recognition systems better at screening out variables such as regional accents. On the other hand, understanding those same variables could make the next generation of voice-print readers more sensitive to the complexity of individual speech.

The coming seminar “will strengthen collaboration within the group,” which this spring was awarded a National Science Foundation grant for collaborative research, said Wolfe, a veteran of a related Radcliffe workshop in 2005. “And I suspect new ideas will arise.”

Radcliffe seminars are “a great tool for intellectual discovery,” he said; they’re more intimate and efficient than large-scale conferences and an important way for junior faculty to get exposure and even open a dialogue with senior scholars. Wolfe’s first seminar, in 2005, was conducted in collaboration with Xiao-Li Meng, Harvard’s Whipple V.N. Jones Professor of Statistics and chair of the Department of Statistics — and a frequent seminar participant.

“I view it as my annual intellectual retreat,” said Meng, who will help co-direct a November seminar, his third. “It gives me an opportunity to listen to and think about topics that are generally new to me.”

One of the advanced seminars this season, in April, will explore cultural creativity in the Ethiopian diaspora. About a dozen scholars will investigate Ethiopian influences on religion, music, art, and literature in America. From 1974 to 1991, Ethiopian civil war forced hundreds of thousands to migrate, and most settled in the United States.

Formal invitations go out this month, said Ethiopian music scholar and seminar organizer Kay Kaufman Shelemay, Harvard’s G. Gordon Watts Professor of Music and Professor of African and African American Studies, who is a 2007-08 Radcliffe Fellow.

“It’s about interdisciplinary as one can get,” she said of the invited group — experts in literature, art history, classic Ethiopian studies, sociology, music, and communications. “The Radcliffe advanced seminar gives us the opportunity to invite these colleagues, who would otherwise not be able to come together.”

A book of essays will appear within a couple of years, based on papers delivered at the Radcliffe seminar. So far, scholars of Ethiopian studies have paid little attention to the cultural implications of the diaspora.

Taking part in the seminar — and a public conference and performance to follow – will be famed Ethiopian musician, arranger, and composer Mulatu Astatke, the father of what is known as Ethio-jazz and a Radcliffe Fellow this year.

“He’s part of the story,” said Shelemay, who’s doing a series of oral-history interviews with Astatke in collaboration with Radcliffe Fellow Steven Kaplan M.T.S. ’77, who teaches religion and African Studies at the Hebrew University of Jerusalem. “His career spans the entire period of the Ethiopian diaspora, and more.”

The 46 Radcliffe seminars of past years have looked at the formation of stars and the hormonal-cancer risk in women. They’ve explored obesity, the seasonality of infectious diseases, the physics of neutron-star winds, and state responses to terrorism. Many have spun off into articles, books, grant proposals, and larger conferences.

One past seminar has already created a significant legacy. In 2004, a consortium of 22 biologists used two days at Radcliffe to explore a proposal for whole-genome sequencing of cichlid fish. This family of creatures has radiated rapidly, spinning off hundreds of related species in what is, in evolutionary terms, a very short time.

Last fall, the National Human Genome Research Institute approved the sequencing of four African species of cichlids, just two years after the Radcliffe meeting. Once done (as early as next summer), the sequencing will allow scientists to identify the cichlid genes related to sensory systems and behavior. Unpacking the inferences of gene function could have implications for understanding the human genome.

“All of genomics is ultimately comparative,” said seminar co-organizer Hans Hoffman. He spent five years at Harvard’s Bauer Center for Genomics Research and is now at the University of Texas, Austin.

Hoffman praised Radcliffe’s “neutral territory,” which allowed a wide-ranging and efficient collaboration to unfold quickly. Without it, he said, the Cichlid Genome Consortium “may never have seen the light of day.”