One weekend last February a Boston insurance company executive composed the tentative first chapter of a book he's been wanting to write for a long time. It is expected to come up for review today at a closed-door meeting in the Roosevelt Room of the White House.

The author is a former FBI official who was in charge of counterespionage investigations at the peak of the Cold War. His manuscript has touched off an unprecedented dispute within the rarefield bureaucracy in charge of government secrets.

The appropriate officials at the Justice Department have already voted to clear it for publication. But the National Securtiy Agency wants to overturn the decision.

"This is hot . . . It's very, very interesting material," said one knowledgeable government official. "The NSA didn't even want anybody else to read it."

The working title of the book, written by former FBI special agent Robert G. Lamphere, is "The FBI vs. the KBG - the Unseen War" and the secrets at issue are more than 30 years old. They involve the interception of Russian intelligence communications, reportedly involving the case of Ethel and Julius Rosenberg who were electrocuted in 1952 after their conviction for conspiracy to commit espionage for the Soviet Union in the time of war.

The dispute is now before the classification bureaucracy's court of last resort, the seven-member Interagency Classification Review Committee ICRC, which has been meeting monthly at the White House in relative obscurity since it was established in 1972 under an executive order signed by President Nixon.

Both sides have been putting forth their positions rather strongly," said RobertWells, a retired Army colonel who serves as the ICRC'S executive director. "I'm sure it will be given real full consideration."

The outlines of a backstage battle are contained in a set of letters written at the Justice Department last month and subsequently obtained by The Washington Post under the Freedom of Information Act.

Lamphere, now a vice president with the John Hancock Mutual Lite Insurance Co. in Boston, elaborated in a telephone interview.

I don't want to do anything that would hurt my government the former FBI man emphasized. He said he is still open to persuasion, but he made clear he has yet to be convinced that publication of the chapter - the first of a proposed nine-chapter book - would do any harm.

From 1947 to 1955. Lamphere said he served in and became supervisor of a special FBI unit in charge of counterespionage investigations that included the Rosenberg case and that of Judith Coplon, a onetime Justice Department analyst tried on espionage charges in 1949. In 1950, Lamphere went to England to interview British scientist Klaus Fuchs whose convinction for passing atomic secrets to Russia prefaced the still controversial Rosenberg case.

At one point during the telephone interview, Lamphere said he was tired of reading again and again over the years about how innocent the Rosenbergs were, but he declined to comment when asked if the chapter at issue involved them. Whatever its subject he said he wrote it in tentative form over asingle weekend last February simply "to indicate how I intended to use my knowledge" and sent it off to the FBI to see if it would cause any problems."It's something of a lead-in to readers to help them understand what the book is all about. I think it adds a lot - the spark that would make the book much more interesting." He said he hoped to use the information else-where in the book as well, but "typically they (government officials) have seized on this (the draft) as the be all and the end all, which it isn't."

The episode illustrates the Byzantine complexities of the government's classification and declassification system. Under the 1972 executive order issued by President Nixon - which is about to be replaced by another being drafted for President Carter - every major department and agency has its own classification review committee to handle requests for the declassification of government secrets.

Over them all its the Interagency Classification Review Committee currently headed by the archivist of the United States, with a staff of eight and a $170,000 a year budget to monitor the workings of the system. But it rarely has to bother with appeals.Since 1972 less than 3 per cent of the requests for declassification review prompted appeals to departmental review committees and less than 0.6 per cent got to the ICRC.

Officials say Lamphere could have gone to the courts instead, and probably still can, but for the moment, he said he is content to see how far he can get within the executive branch. One Justice Department expert speculated that the final decision might have to be made by President Carter.

Although it contains no government documents as such, the manuscript was apparently stamped "Top Secret" shortly after its arrival in Washington. The FBI saw that it involved communications intercepts within NSA's province and classified it at the code-breaking agency's behest.

The thing was really very confused," recalled one member of the Justice Department review committee. The FBI originally could have refered the material directly to NSA but instead they classified it themselves and it fell within jurisdiction of the Justice Department."

As a result, when Lamphere appealed the FBI action, the issue was bumped to the Justice Department committee which listened to a barrage of protests from NSA officials in August and remained distinctly unimpressed.

As the chapter contained information pertaining to the interception by the United States of communications intelligence, the National Security Agency was consulted as to whether classification of the document was warranted, the Justice Department review committee's chairman. Leon Ulman recounted in a Sept. 14, 1977, letter to acting ICRC chairman James B. Rhoads. Although the agency's recommendation was in favour of classifying the document, the Department Review Committee by a vote of 4 to 1 concluded that classification was not warranted inthis case." The vote was taken Aug. 6.

NSA Deputy Director B. K. Bulfham and his aides protested and won another meeting of the Justice Department committee last Aug. 16.

Ulman informed Buffham in another Sept. 14 letter that the Justice Department panel had decided to bump the issue to the ICRC as a result of the NSA's demands for continued secrecy, but he added that the Justice Department still felt "that the information does not warrant classification."

According to ICRC executive director Wells, the interagency panel will probably vote today simply on whether to accept the case. It may take another month or two to deal with the merits of it.

Fortunately for Lamphere, according to one Justice Department official the dispute will be decided under the Nixon order rather than the one under consideration by Carter. For secrets more than 30 years old, the Nixon order demands declassification unless continued secrecy is "essential to the national security."

The Carter draft order removes that test and allows continued secrecy for any matter whose disclosure would be "demonstrably harmful" to the national security.

"It's doubtful that the NSA can show demonstrable harm. But I don't think they could possibly show this continued secrecy is essential to national security," the Justice Department official said.

NSA officials declined to comment, but they are apparently concerned about the outcome, too, said Lamphere. "There have been some efforts to discourage me from continuing my appeal."