"The overwhelming majority (of the material they want redacted) was in the manuscript that was cleared for publication in September 2004," said attorney Mark Zaid.
An official authorized to speak for the agency declined to comment on the specifics of the case, citing the ongoing legal proceedings, but said that the CIA was "a secret agency... fighting a war against terrorism."
"We will continue to guard our secrets," she said.
Open government advocates say the effort to stop publication is part of an increasingly aggressive effort by CIA Director Porter Goss to -- in Zaid's words -- "clampdown on any discussion of the agency's activities by current or former employees."
Some see it as part of a government-wide effort to chill public scrutiny of the executive branch's secretive national security agencies. Others say it is a reaction to a series of books by CIA veterans that have criticized or otherwise embarrassed the administration.
The lawsuit alleges that Paul Chretien, the chairman of the CIA board which reviews writing that current or former employees want to publish, quit last year due, "at least in part, to his opposition to the current CIA policies that infringe upon the First Amendment rights" of would-be authors.
Zaid represents Thomas Waters, one of the 150,000 people who applied to join the agency in the immediate aftermath of the Sept. 11 suicide hijackings, and one of the 100 or so of them who were admitted to the CIA Directorate of Operations training program in what he says was the largest, most diverse class in the CIA's history.
He left the CIA after two years for "unrelated family reasons," said Zaid, and is now an intelligence contractor with the Department of Defense.
Zaid says his client submitted a manuscript for pre-publication review in May 2004.
All CIA employees sign a secrecy agreement when they go to work for the agency that commits them to submit anything they intend to publish -- even after they leave -- for such a review.
He said the CIA's Publication Review Board, which checks all such manuscripts to ensure they contain no classified information, approved the text in late September 2004, after requiring redactions.
Waters then sold his manuscript to Dutton, which is part of Penguin Group, in December 2004. The book is scheduled for publication next month, and Internet retailer Amazon.com is already accepting advance orders.
Zaid says that throughout the course of last year, Waters submitted small changes and additions on three separate occasions, but heard nothing substantive until last month, when the board's new chairman wrote to him.
"Suddenly, they were asking for dozens of redactions of the exact same language they'd cleared before," said Zaid.
He said that, by Waters' count, about 167 pages of the manuscript had redactions of text previously cleared for publication, whereas only about 15 pages contained redacted material that the board had never seen before.
Some of the redactions were of information that was revealed -- and cleared for publication -- elsewhere in the manuscript, or in other books about the agency, Zaid said.
"I don't understand what these people think they're doing," said Zaid. He said the review process was "inconsistent," and some of the reactions were "ridiculous."
The official authorized to speak for the CIA denied that the process was inconsistent.
"For former employees, the (Publication Board Review) process is designed solely to prevent publication of classified material," said the official. "Classified is classified," she added, denying that there was any inconsistency.
That notion was pooh-poohed by open government advocates.
"The CIA practices an arbitrary and indiscriminate form of secrecy," said Steven Aftergood, of the Federation of Government Scientists. He added that the CIA's decisions about secrecy were often "utterly illogical and breed contempt for the (classification) system."
During the protracted negotiations in 2003 over the declassification of the report of the Joint Congressional Inquiry into the Sept. 11 attacks, Staff Director Eleanor Hill said at the time some U.S. intelligence agencies tried to insist on the redaction of material that was available on their own Web sites.
Aftergood said that without knowing what was in the book, it was difficult to comment on the specifics of this case. But he added that "under Goss, there has been a tightening of information" control at the CIA.
One former senior CIA official who left before Goss took over said he had "heard people suggest that (the new management team) are ... less permissive" about allowing publications.
He said that, during his time at the agency, it was "not a matter of being in favor of giving away secrets," but nonetheless, the Publication Review Board had tried to work with would-be authors "to help get things out in a timely fashion."
Zaid agreed, saying that the board had fought "big battles with (directorate of operations) people" over the rights of insiders to publish. "They were one of the best offices in the CIA," he said.
There were "historians, lawyers from other agencies" on the panel, he said. "They were not CIA types."
The former senior official said he had heard that now the board was "moving more slowly and taking a stricter view."
He acknowledged that some controversial books -- like "Imperial Hubris," the angry critique of U.S. strategy in the war against terrorism, written by the CIA's one time al-Qaida expert Michael Scheuer -- had hurt the agency.
"Our view was always that we were big enough and tough enough to take the criticism," he said.
Goss' defenders say that the suggestion of a clampdown is rebutted by the steady stream of books published by CIA insiders in recent years.
Two memoirs by female former spies have criticized the agency's treatment of women employees and its allegedly antediluvian management style.
Other recent publications, like Gary Schroen's "First In," and Gary Berntsen's "Jawbreaker," are not critical of the CIA directly, but nonetheless have been seen as embarrassing for the administration because of the way they highlight the failure of U.S. forces to capture Osama bin Laden during their conquest of Afghanistan.
The former official said that some in the agency believed the stream of books "had gotten out of control." They believed "too many people were publishing," and the agency's reputation was at risk, he said.
Shortly after Goss took over at the agency in 2004, new regulations were introduced governing what current employees could publish.
Zaid said that Goss' clampdown was "part of an administration policy" to curtail discussion about the activities of U.S. intelligence, and that in the process his client's rights were being trampled.
He said that, because of constitutional guarantees of freedom of expression, the agency could not prevent former officials from publishing material that was merely embarrassing or inconvenient.
"They are using classification to get round that," he said.