be the individual researcher—the modern equivalent of the polymath—who has achieved single-handedly a deep understanding of two or more disciplines and the ability to integrate them. At the other end might be a structure of multiple government-funded programs staffed by thousands of scientists and engineers drawn to a goal as ambitious—and focused—as the search for life on Mars. Interdisciplinary structures may also be interinstitutional, sharing no common physical space, or they may be in physical centers or “collaboratories” of substantial size and life span.

Whatever their structure, interdisciplinary projects flourish in an environment that allows researchers to communicate, share ideas, and collaborate across disciplines. The flow of ideas and people is made possible by institutional policies that govern faculty appointments and salary lines, faculty recruitment, responsibility for tenure and promotion decisions, allocations of indirect-cost returns on grants, development of new course and curricular materials, and so on.

Many researchers want to pursue interdisciplinary work more actively, but what new structures can best support them? The committee envisioned two possible modes for the creation of new structures: an incremental mode, which builds on lessons learned in the recent past, and a more transformative mode in which change comes more rapidly and discontinuously with respect to existing structures and practices.

Given the diverse nature of interdisciplinary activities, the number of formats for IDR in the future is likely to reflect the growing complexity of research. Whatever format characterizes a given IDR project, especially in academic institutions, it must operate in the context of a larger, overarching institutional framework that in many ways defines and constrains it. It is important, therefore, to examine institutional organizations and traditions critically and to ask what kinds of changes are possible and helpful for IDR.

An older management structure of universities is a landscape of separate components, or “silos,” with weak coupling between them. A newer structure, which can already be discerned both in the United States and abroad and which has long been evident in industry and elsewhere, is more like a matrix, in which people move freely among disciplinary departments that are bridged and linked by interdisciplinary centers, offices, programs, courses, and curricula. There are many possible forms of coupling between departments and centers, including appointments, salary lines, distribution of indirect-cost returns, teaching assignments and course-teaching credits, curricula, and degree-granting.

A matrix structure (see Box 9-1) in a university might include many