Washington — The cardboard box of documents is about the size of a piece of carry-on luggage. It is marked ''Tehran Embassy, General Records, 1951,'' and has just been cleared by National Archives declassifiers.
Many of the papers it holds are not exciting: transfer orders, trade complaints, a decree that ladies' attire at the Shah's marriage will be ''the dress worn at tea, with hat.''
Some, however, speak candidly of a society even then under great stress. ''The cooler heads are in the minority here at the present,'' one flimsy blue dispatch says. ''Anything may happen, but we are still optimistic that some workable arrangement will be made which at least will keep oil flowing to the West.''
Once-classified documents such as these are the raw materials of history. But over the last four years, the number of secret files being made public has been greatly reduced.
Under the Reagan administration, declassification of United States documents has dropped from 1980's record-high 90 million pages to 12 million pages last year. This tight grip on old secrets, historians complain, is greatly hindering the writing of post World War II history.
''In particular it is material dating from 1950 to 1954 that is being fought over,'' says Walter LaFeber, a Cornell history professor.
The United States, for instance, has let slip very little about its role in Guatemala's 1954 coup, Dr. LaFeber says. Material dealing with the 1953 uprising in Iran is still stamped ''secret,'' although US diplomats have written about their involvement in the action.
The roots of the dispute over classified documents lie in the fact that presidents have much control over what the government stamps ''secret,'' and which old files are dusted off and made public. Classification standards are set by executive order, not laws passed by Congress.
Ironically, it was President Nixon who first gave the public widespread access to once-sensitive government documents. In 1972, he signed an order establishing systematic declassification to clear out the piles of World War II documents then clogging government files.
President Carter pried open the file drawers even farther by requiring that government agencies review all 20-year-old documents for possible declassification.
Then, in 1982, President Reagan authorized a thorough renovation of the US classification system.
''This order enhances protection for national-security information without permitting excessive classification of documents by the government,'' Mr. Reagan said upon the occasion of the change.
Under Reagan, there are still three categories of classified information: confidential, secret, and top secret. (Only 2 percent of classified documents are top secret. Sixty-six percent are merely confidential. The Pentagon classifies the most documents, with the Central Intelligence Agency running a close second.)
About 7,000 government officials are authorized to wield classifying stamps. And lately those bureaucrats have actually been stamping less, by one important measure: classification of original documents declined 18 percent last year, points out Bob Wells, deputy director of the US Information Security Oversight Office.
But the unveiling of secret US documents over the last three years has been sharply curtailed. Fewer pages are being reviewed for release, and standards are stricter. In 1981, 91 percent of documents examined were declassified; last year , the figure fell to 63 percent.
''Carter may have gone too far in opening things up,'' says Dr. Sam Gammon, director of the American Historical Association, ''but our view is that Reagan has turned things around and headed them back by express.''
One reason the flow of declassified documents has slowed is that the National Archives has fallen on tough times. The Archives, a huge neoclassical barn near the Mall, is, under the Reagan policy, supposed to declassify most important government papers that are 30 years old. But ''we've lost 75 out of 100 declassifiers (to budget cutbacks),'' says one employee. ''We don't have the staff to do systematic declassification.''
Another cause of the declassification drop may be the fact that papers are being reviewed more closely. Over the last six years, agencies have written detailed guidelines. The Archives and others are now required to use these handbooks.
But it's the nature of the documents themselves that is a major factor in current declassification disputes.
Much of the paper work now coming up for classification scrutiny is from the early cold-war period. This period, government officials say, was the beginning of our current phase of history - and thus many of them are still relevant.
''We've passed (the papers) from World War II and Korea, which could be declassified in bulk,'' says Bob Wells of the Information Security Oversight Office. ''The closer we get to the present time, the more sensitive the information is.''
Historians working on the early '50s don't agree that many secrets from the period are still significant. Anna Nelson of George Washington University, who is writing on the workings of the National Security Council before 1960, says ''you're talking about documents 30 years old. There's not that much left.''
Cornell's Dr. LaFeber, a diplomatic historian, says, ''I have difficulty believing that what went on in '53-54 was more important than the (1945) Yalta Conference, which shaped our modern world. Yet the Yalta documents were released in 15 years.''
Critics of the Reagan declassification order also complain about the ability of the government to change its mind, and reclassify material already made public.
Administration officials say the flow of declassified information will increase somewhat in the future. They point out that classifiers must now mark classified documents with an automatic declassification date if possible. About 35 percent of papers classified last year carried such a date.
In any case, historians do not wish for a return to the old days, when there was no organized declassification at all. Back then, officials sometimes allowed favorite historians access to secrets, creating a scholarly privileged caste.
''That's the sort of thing we wouldn't tolerate today,'' says Anna Nelson of George Washington University.