Leaders of U.S. science and engineering, reacting to Defense Department restrictions on attendance at an unclassified technical symposium, say the action represents a new and possibly more worrisome phase in a six-year-old Pentagon campaign that threatens the free exchange of scientific information.

The Pentagon says the goal of its campaign is to keep U.S. high-technology data that relates to military or space matters from reaching channels to which Soviet scientists have access.

But the most immediate result, the scientists say, has been a "chilling effect" on science's valued tradition of sharing information.

In its action earlier this month, the Pentagon at first forbid the presentation of about 25 papers and then, when symposium organizers objected, compromised by allowing the papers to be given before an audience restricted to U.S. citizens and selected foreign scientists, all of whom had to sign a pledge not to make the data public. Pentagon officials stood at the doors collecting the pledges.

In previous actions, Pentagon pressure has persuaded scientists to withdraw their papers. Several scientific societies, hoping to avert Pentagon intervention, have closed their meetings to all but Americans presenting proof of citizenship.

"The Defense Department has embarked on a course that -- as patriotic and well-intentioned as it may seem -- may threaten the technological supremacy of the U.S.," said Richard J. Gowan, immediate past president of the Institute of Electrical and Electronic Engineers (IEEE).

Along with many others, Gowan argues that the U.S. military gains more from open scientific communication than it loses through data reaching the Soviets.

Robert L. Park, a University of Maryland physicist and director of the American Physical Society's office of public affairs, says the fact that the Soviets want to steal U.S. data is comforting confirmation that the United States is well ahead.

Soviet technology is behind, Park said, because it "suffers from the kind of restrictions the Defense Department is trying to impose here. Their scientists can't talk to one another or to their colleagues in other countries. I'm not worried when they steal our information. I'm going to start worrying when we start having to steal their information."

Although the Pentagon's actions affect only research that it funds, this includes almost every area of basic science and high technology, including many where civilian applications are as important as military applications and other areas with no obvious applications as yet.

Over the years the Defense Department has become a major funding source for scientists with no interest in military applications but who found, amid growing competition for funding, that the Pentagon would support their research.

None of the research the Pentagon campaign is restricting is classified. (Military officials say classification imposes cumbersome and costly restriction that go too far.) In fact, many of the scientists affected work at universities whose policies prohibit secret research. Thus, they find it troubling to be told their findings cannot be shared freely.

The campaign began in 1980, shortly after then-Sen. Henry M. Jackson (D-Wash.), a longtime critic of the transfer of high-technology hardware to the Soviet Union, denounced the Commerce Department for waiting until the Soviet invasion of Afghanistan to crack down on technology transfer.

Apparently goaded into action, the Commerce Department "swooped down," as some scientists put it, on an international meeting of the American Vacuum Society, declaring that the presentation of papers on a new form of computer memory to an international audience constituted export of technical data. Unless they had a license to export, the agency said, the scientists could face large fines and prison terms of up to 10 years.

Participants from Hungary, Poland, the Soviet Union and China were barred from the sessions. FBI agents were on hand to ensure that other foreign scientists signed pledges not to reveal the information to communist-bloc colleagues. During the conference, the bar was lifted for Chinese scientists.

At least 12 times since then, according to a list compiled by the American Association for the Advancement of Science, officials from the Commerce Department or the Defense Department have acted at the last minute to pull papers from meetings and to bar attendance of selected foreign scientists.

In 1982, for example, Pentagon officials ordered the Society of Photo-Optical Instrumentation Engineers -- the group affected this month at its meeting in Arlington -- to withdraw more than 100 of the 700 papers scheduled for delivery in a San Diego meeting even though another branch of the Defense Department had approved their presentation.

Also in 1982 the Air Force told organizers of an IEEE symposium in Washington that their program had to be canceled and all records destroyed. The conference chairman said he would comply if the Air Force picked up the $25,000 to $50,000 costs of doing so. One day later the Air Force withdrew its request.

Many of the actions have been taken under the Export Administration Act of 1979.

Other actions have been taken under terms of the contract that scientists doing applied research sign with the Defense Department. The contract obliges scientists to submit their papers to the Pentagon for approval before they present or publish them. Park said contractual restrictions are becoming more common, forcing some universities that once accepted unrestricted Defense contracts out of working for the military.

Contractual restrictions, however, do not apply to most Pentagon-funded basic research -- research aimed at gaining fundamental knowledge and not at developing particular equipment. It is here that scientists and civil libertarians see the new threat.

When the Pentagon acted against the meeting in Arlington earlier this month, it invoked a provision of the 1984 Defense Authorization Act. The provision was enacted to allow the Pentagon to withhold key technical data when responding to requests under the Freedom of Information Act. The Pentagon had asked for the provision on the grounds that Soviet agents, denied data under export laws, could get what they wanted by filing an FOI request.

As enacted, however, the law is not limited to FOI matters. Rather than simply ban certain papers at the Arlington meeting, as would have been possible under the contracts, Defense Department officials said the law empowers them to set up sessions with restricted attendance.

"This amounts to a clear violation of the intent of Congress," said Allan Adler of the American Civil Liberties Union. "They're taking this provision, which Congress said was to apply to Freedom of Information Act matters, and saying it gives them authority way beyond that."

Although the papers at the Arlington meeting dealt with applied research, Adler said "the danger now exists that the government could apply this to academic research," which is predominantly basic research.

Pentagon officials say this could never happen, that the new provision applies only to research covered by the export-control laws, which generally do not cover basic research.