The recent reversal of U.S. policy on restricting unclassified government research comes after prolonged efforts by Harvard and the academic and scientific community at large. Many educational associations have resisted the trend in recent years towards government control over how scientific papers produced at universities under government research grants are disseminated.

In recent years, several federal agencies and departments--especially the Department of Defense--have sought to boost government say over the publication of scientific research with military applications and over the results obtained in studies done under government contracts.

Government sponsorship is particulary important to Harvard, which gets 19 percent of its income from federal sources, including $39 million in Fiscal Year 1985 for projects by the Faculty of Arts and Sciences.

The other recent addition to the growing file of mutual accusations of excessive secrecy and excessive technology transfer came earlier this month in the form of a Pentagon report that said open research conducted at Harvard and at other universities and by private companies is being used by the Soviet military.

Officials here called the report spurious and vague, but a Pentagon spokesman said it showed a need for tighter security. Last week's White House announcement said efforts at tightening export controls and classifying more information would continue, but basic scientific research would not be affected.

The principle issues under contention, and Harvard's positions and actions on them, are as follows:

Increased Classification.

President Reagan and the Pentagon have significantly relaxed standards for determining when documents can be classified. In 1982, the President's Executive Order 12356 formally eliminated the requirement that classifiers take into account the public's need for the information in question.

In addition, the government has more say in what is classified and what isn't. Before, the classifier had to demonstrate that open access to the information would cause "identifiable damage" to national security. Now, the government can classify meterial if "its disclosure could reasonably be expected to cause damage to the national security."

The order also provides that any borderline decisions should be classified and that material can be reclassified after it is requested under the Freedom of Information Act. These decisions both reverse earlier policies.

Harvard's rules for research expressly prohibit work that is classified or otherwise kept from free dissemination. The more contracted research that gets classified, the less can be done here.

Observed Phillips Professor of Astronomy Alexander Dalgarno on the latest Pentagon report, "I advocate less classification. Secrecy is not the American way of doing things."

Pre-publication Review.

Administrators complain of a growing tendency within the government to regard work they commissioned as private property over which they can exercise rights of "correction" and "modification" by their own standards. Several contracts having no military applications--including studies of health policies and of the economic development of cities--have been offered to Harvard with the rights of publication withheld pending "approval" by the sponsoring federal agencies.

Read one clause in a contract proposed by the National Institute of Education, "Any presentation of any statistical or analytical material covered by this contract will be subject to review by the Government's Project officer before publication or dissemination for accuracy of factual data and interpretation."