Presidents of the nation's leading scientific and engineering societies have broken with the Pentagon over its efforts to restrict access to unclassified research, declaring that their organizations will no longer sponsor restricted sessions at their meetings.

The effect of the presidents' actions would be to shut out from their society meetings the papers of any defense-funded scientists working in "sensitive" but nevertheless unclassified areas. In effect, the organizations, which range from the American Association for the Advancement of Science to the American Association of Engineering Societies, are now refusing to deal with any papers the Pentagon may restrict.

In a Sept. 17 letter to Defense Secretary Caspar W. Weinberger, the group accused the Defense Department of creating a new system of classification on research. It said, in effect, that if the Pentagon wants certain subjects restricted, it should take them out of open meetings or set up classified meetings.

"Responsibility for implementing controls for such information must lie with the government and not with our organizations," the letter said. "Therefore, our organizations will not be responsible for, nor will they sponsor, closed or restricted-access technical sessions at meetings or conferences conducted under their auspices."

The 17 societies have a membership of more than 2 million scientists and engineers.

The executives maintained that "it is in our mutual best interest to minimize the amount of unclassified information that is subject to these controls . . . . We will be pleased to work with you to define measures that will achieve this goal."

"We're still formulating a response," said Col. Donald I. Carter, acting deputy undersecretary for defense research and advanced technology. "We had been working with the societies over the last three or four months. I guess they were unhappy with the progress we were making." The crux of the issue, Carter said, lies in the 1984 Defense Authorization Act and the Export Administration Act that enable the Pentagon to determine what "technical data" merits special protection.

"Compare it to proprietary data," Carter said. "If we were a private company investing in research and development . . , we would hold the data proprietary to assure we got first use."

What should be noted, Carter said, is that "technical data is different from scientific information. Scientific information is basic principles and scientific phenomenon; technical data is engineering data -- taking scientific data and converting them into useful systems."

Carter said his office is placing restrictions on "critical technologies" data that are subject to export control so they will not fall into the hands of Soviet bloc countries.

A new Pentagon report indicates that the Soviet Union has extensive efforts under way to acquire sensitive U.S. technologies and research, and this is the main reason for what has become a new level of classification for researchers.

At the Society of Photo-optical Instrumentation Engineers meeting in April, the Pentagon used the rule to block presentation of 25 papers. It eventually allowed the papers to be presented to U.S. citizens and selected foreign scientists who signed nondisclosure agreements.

"What we're saying is the extent to which export control limitations are placed on information go well beyond what we would see is the intent of export controls," said Russell C. Drew, a vice president of the Institute of Electrical and Electronics Engineeers. "We're not going to be a party to carrying out these restrictions. We maintain certain guidelines on the free exchange of information and we're not going to bend them or break them just to meet current Defense Department desires."