WASHINGTON -- In recent years, the government has told Americans it has credible evidence of impending terror attacks, of weapons of mass destruction in Iraq and of collaboration between al-Qaida and Saddam Hussein. But "credible" doesn't mean the same thing to every government official, and even credible information can be wrong.
The warning this week from Attorney General John Ashcroft of a possible impending al-Qaida attack in this country has set off a new debate among ordinary citizens, on the presidential campaign and inside the government over the meaning of "credible."
Al-Qaida's intent to launch a major attack in this country has been no secret for some time, former CIA counterintelligence chief Vince Cannistraro noted Thursday. "The question is their capabilities and that [information] we don't have."
To professional intelligence analysts, credibility grows out of history. "Specific credible intelligence" of such a summer attack would mean details of plans from a source, either human or electronic, that had provided reliable information in the past, Cannistraro said Thursday. He didn't see any evidence of that in Wednesday's announcement.
In addition, analysts would try to determine whether the informant had access to such data and what his motivations might be -- revenge, money, ideology or whatever, said Larry Mefford, former FBI counterterrorism chief. Then analysts try to see if all or parts of the information can be corroborated or fit with other known events.
But that makes it sound easier -- and more clear -- than it is.
Author Bob Woodward writes in a new book that National Security Agency director Michael Hayden tried to explain to his wife the difference between the black-and-white world of facts and the gray world of intelligence this way: "If it were a fact, it wouldn't be intelligence."
Sometimes intelligence isn't discovery of a missile but an attempt to peer into the mind of an enemy leader to learn when he will use it.
Inside the FBI, Mefford said, "'credible intelligence' is a term taken very seriously. It always referred to information we believed to be reliable."
Nevertheless "credible intelligence is a very subjective term," Mefford added. And the analysis of intelligence "is not a science; it's an art."
The FBI is still trying to improve the formal internal process by which it vets the reliability of sources, Mefford noted.
Recently, Secretary of State Colin Powell conceded the Bush administration accepted distorted reports of Iraqi weapons of mass destruction. Other administration and congressional officials said the misleading information came from Ahmed Chalabi, a prominent exile who wanted to get the United States to overthrow Saddam so he could return to Iraq.
In January, departing U.S. weapons inspector David Kay said he no longer thought such weapons existed in Iraq. He said it was clear by then that the CIA's problem was a lack of its own spies in Iraq who could provide credible information.
This March, however, Charles Duelfer, now the CIA's supervisor of the search, said his group regularly receives reports -- "some quite intriguing and credible" -- about possible weapons stashes hidden in Iraq. But if these informants are judged credible because of previous reliable reports, such earlier reports must be about something other than caches of weapons of mass destruction because none of those have been found.
This debate also has played out in congressional hearings. The CIA's Tenet told the Senate Armed Services Committee that policy-makers are entitled to flexibility in describing intelligence and don't use precisely the same words as intelligence analysts.
But Sen. Edward M. Kennedy, D-Mass., angry over what he saw as Bush administration exaggeration of the justification for the Iraq war, replied, "I'm not talking about parsing words. We're talking about words that are basically warmongering."
Associated Press writer Katherine Pfleger Shrader contributed to this report.