Seth Waxman, arguing on behalf of 37 detainees at Guantanamo BayCuba, told the Supreme Court Wednesday that each of these men "maintains, as this Court explained [in an earlier case] that he is quote 'innocent of all wrongdoing.' " Waxman, a former solicitor general, is the kind of oral advocate who can sell justices their own underwear. And in one fashion or another, he is likely to win this case.

But neither the justices nor the public should take quite at face value the insistence, however passionately and eloquently argued, that large numbers of innocents populate Guantanamo. To focus just on the 37 before the court this week, most Americans would not use the word "innocent" to describe Bashir Nasir Al Marwalah, who told a review panel at Guantanamo that he had traveled from his home in Yemen to Afghanistan to train at an Al Qaeda camp in 2000, returned in the summer of 2001 for more specialized training as a sniper, and served the Taliban on the back lines. Asked "Are you a member of al Qaeda?" he responded: "I don't know. I know I am an Arab fighter." 

They might choose a different word as well to describe Muktar Al Warafi, who openly acknowledged traveling to Afghanistan in response to fatwas, affiliating with the Taliban, training at a camp, and then going to the front lines "just to visit."

They might have thought another detainee, Abdul Latif Elbanna, was admitting something substantial, rather than professing innocence, when he described helping bring an Al Qaeda figure hiding in London his wife and son. Perhaps another detainee, Musab Omar Ali Al Mudwani,was professing innocence when he acknowledged going to Afghanistan, training at an Al Qaeda camp and seeing Osama Bin Laden twice--since he claimed to have been tricked into doing so. Numerous others of the 37 acknowledge staying in Taliban or Al Qaeda guest houses, training, or taking other forms of assistance from the Taliban.

Waxman conceded, immediately after stating that all of the detainees before the court deny all wrongdoing, that the facts of their cases differ a lot from one another. "[I]t may well be," he said, that habeas corpusreview of the type he advocates "would reveal perhaps that some of these [detainees] are lawfully detained." Indeed, it may. In many cases, the court would need look no further than the detainees' own words--one of many reasons the Bush administration's allergy to judicial review of detentions has been so self-defeating.

But the broader debate over Guantanamo has suffered greatly from overbroad claims of erroneous detentions there. The New York Times referred in an editorial to "hundreds of innocent men ... jailed at Guantanamo Bay without charges or rudimentary rights"--a statement it cannot possibly support. We hear endless stories about relief workers, instructors in the Koran, and victims of mistaken identity swept up and sold for bounty by the Northern Alliance to gullible Americans led by a malicious administration. There's an element of truth here, of course. A few certain cases of egregious error have surfaced. And others present wrenching conflicts between fairness, justice, and security interests. For example, Waxman's own clients are a group of six Algerian-born men who were living in Bosnia and arrested on suspicion of plotting to blow up the American embassy inSarajevo. After the Bosnian Supreme Court ordered them released, however, the authorities turned them over to the U.S. military, which whisked them off to Guantanamo. All claim to be innocent of everything. And the military's allegations against them have never faced any real test.

But there is also an element of romantic fantasy in the belief that large numbers of Guantanamo detainees are there by mistake. For the past six months, I have been studying declassified materials from Defense Department reviews of Guantanamo cases: transcripts and records of the much derided panels known as the Combatant Status Review Tribunals (CSRT) and Administrative Review Boards (ARB). While my data are not yet fully complete, the general picture of the people who went through these reviews, many of whom have been released, is clear: About a third of detainees admit facts that offer significant--though not always adequate--support to the government's contention that they are "enemy combatants." About a third deny everything. And about a third make no statements at all. (While some military interrogations have been coercive and the CIA has probably crossed over into outright torture, nobody has alleged that these particular statements, which took place in hearings before panels of officers, were involuntary.) This approximate ratio has remained relatively consistent even as the population at Guantanamo has shrunk.

The admissions vary a great deal. Some detainees proudly declare their Al Qaeda membership and terrorist activity. Many more admit fighting for the Taliban. Some Afghans claim plausibly to have been forced into Taliban service. And a fairly large group denies membership or belligerence but admits to some lesser degree of affiliation--staying in Taliban or Al Qaeda housing or taking training, for example--that is at least suggestive.

The denials vary a great deal as well. In some cases, they have the ring of truth. In other cases, they are so patently absurd as to warrant quick dismissal; in one case, a detainee claimed to have been buying rare collectibles-mummies, to be precise. In the vast majority of cases, they are not easily assessed one way or another in the declassified material. A great many detainees tell more or less the same small number of stories: That they came to Afghanistan to do relief or charity work, to study, to look for jobs, or to check out the supposedly "pure" Islamic Taliban regime. My guess is that the overwhelming majority of these detainees are lying--for the simple reason that it beggars belief that the roundup of foreigners in Afghanistan and Pakistan could have nailed such a concentration of relief workers and students. But the stories are probably true in a percentage of cases, and the government's evidence of their falsehood may in quite a few instances be weak--particularly if one considers only the evidence that would be admissible in court.

The more I studied the CSRT and ARB records, in fact, the more convinced I became that the government is another victim of the inadequate process that it set up for reviewing the detainees--which has lost in the public arena no matter what those reviews found. The CSRTs found 38 out of nearly 600 detainees to be "no longer enemy combatants," a bizarre euphemism for erroneously held. The ARBs quickly freed another 14 of those the CSRTs found to be enemy combatants. Yet where the review processes have freed people, they never received any credit for separating the wheat from the chaff. When detainees openly admitted their affiliations, validating the government's claims, the secrecy associated with the hearings and records meant that the public never learned about it. And when the reviews resolved contested issues of fact in the government's favor, their ad hoc status, stacked rules, and lack of congressional authorization denied their judgments credibility in the public eye. Almost no matter what they did, people saw them as unjust.

This country desperately needs a new adjudicatory framework for these detainees--one that includes an expanded judicial review, fairer rules, and clearer, less permissive standards for detainability. On this point, my only disagreement with Waxman is on the question of which branch of government should set it up. He urges the courts to do it through the habeas process; I want to see Congress design and enact a statutory scheme. That said, I don't believe that the result of such a process would be freedom for a lot of innocent people. The more likely outcome would be continued detention with the judiciary's stamp of approval for the majority of detainees, and freedom for two groups: A small group of true innocents and a larger group consisting of dangerous folks against whom the government has only weak evidence.

The scary question we need to address as a society is how large we are willing to let that second group grow in order to make sure the first group becomes as small as possible.