An exclusive club of US scientists that advises the government on military matters has lost much of its funding after refusing to accept orders over whom it should take as a member.

The funding was cut after the group — known as the JASONs — rebuffed three nominees put forward by the Defense Advanced Research Projects Agency (DARPA), a division of the Pentagon devoted to innovative research and which provides about half of the think tank's money. Some JASON members say that they saw the nominations as an attempt to make political appointments.

The JASONs, a group of about 50 scientists, have met behind closed doors for the past 40 years to review the Pentagon's toughest technical problems. The team, which is primarily composed of prominent university researchers, has played an important role in the development of modern submarine warfare, for example, and in assuring the government that nuclear weapons can be safely maintained in the absence of testing.

The nomination problem began to unfold last summer, when Tony Tether, who was appointed as DARPA's director by President George Bush, suggested three new members intended to provide the JASONs with expertise in Internet-related technologies. The JASONs were not impressed with the nominees — two Silicon Valley executives and an engineer — and, after reviewing their credentials and willingness to participate, told DARPA that the group would not accept them.

In negotiations that followed, sources familiar with the situation say, the JASONs offered to let the new members join their meetings on an ad-hoc basis or, failing that, proposed that DARPA nominate alternative candidates. DARPA refused and allowed its contract with the JASONs to expire this year.

DARPA officials say that the new members would have brought fresh blood into the JASONs and that without them, the organization has become obsolete. “The JASONs were valuable during the cold war,” says a spokeswoman for the agency, “but after it ended, they didn't move towards new fields such as information technology.” Without the new appointments, the JASONs would be too heavily oriented towards physical sciences to be of help to DARPA, she says.

“The DARPA statement is just wrong,” says Steven Koonin, provost of the California Institute of Technology in Pasadena and current head of the JASONs. Koonin concedes that more than half of the JASONs are physicists, but points out that the group also includes prominent biologists, computer scientists and engineers. Over the past few years, he says, the topics that the group has researched include climate change and biowarfare. “It is amazing to me, given the heightened importance of national security, that we've been sidelined,” he says.

Steven Block, a biophysicist at Stanford University and JASON member, and other members of the group who declined to be identified, allege that DARPA's insistence on its three candidates was an attempt to make political appointments to a panel that strongly values its independence. “There seems to be an attempt to make the JASONs into a political patronage job,” Block says.

Others close to the dispute think that personal differences led to the split. Conflicts between the JASONs and DARPA officials have erupted before. “This should have been talked out, but somehow it became a matter of principle, and then it got out of hand,” a source close to the JASONs says.

The JASONs are now looking for a prime sponsor to replace DARPA. They are believed to be close to finding an alternative within the Pentagon, although they may also talk to other agencies, such as the Central Intelligence Agency and the Department of Energy.

The dispute between the two groups has worried observers inside and outside the US government. “We think it's a mistake,” says Steven Aftergood of the Federation of American Scientists, an organization that monitors issues relating to science and defence. Although Aftergood admits to being frustrated in the past with the veil of secrecy that surrounds the JASONs' work, he believes that they provide the Pentagon with much-needed independent advice. “They provide a useful reality check for programmes that otherwise might be running on enthusiasm,” he says.