Closing Guantánamo will not be easy. As David Cole points out, some of the detainees held there cannot be sent back to their home countries, and others are too dangerous to release. But while he insists that outsourcing the problem is not an option, he seems all too willing to bring the problem home. Establishing a long-term detention-without-trial regime on U.S. soil for Taliban and Al Qaeda suspects-Cole's proposed solution to the mess at Guantánamo-would import the Guantánamo approach rather than eliminate it.

A prominent critic of Bush administration counterterrorism policies, Cole condemns Guantánamo as a symbol of lawlessness. Change is needed, he asserts; the substantive and procedural unfairness of the current system must be remedied.

What is surprising is that the changes Cole proposes are so meager. His recommendations would modify the current system by narrowing the category of suspects who can be held and providing fairer procedures, but they would not challenge the underlying detention paradigm. While they would certainly be improvements, they would not be meaningful reforms.

* * *

At Human Rights Watch, our experience with long-term preventive detention is not limited to Guantánamo. In a range of countries where we work-including Egypt, Syria, Libya, Malaysia, and Saudi Arabia-al Qaeda suspects, other alleged terrorists and extremists, and even government opponents are detained without trial based on security rationales. Notorious past examples include apartheid-era South Africa, which held tens of thousands of government opponents in preventive detention as security threats during the last decades of white rule.

While democratic governments have experimented with preventive detention on security grounds-the British internment of Irish nationalists comes to mind-reliance on preventive detention is typically a hallmark of repressive regimes. Human Rights Watch has found, moreover, that the use of preventive detention is nearly always part of a larger package of abuses, often involving arbitrary arrest, secret and incommunicado detention, and the infliction of torture and other ill-treatment during the initial weeks or months of confinement. The Bush administration's detention-related abuses are no exception.

Cole thinks that his proposal avoids the dangers of these broad, security-based approaches by adhering to the laws and customs of war. He advises setting up a preventive detention regime under the rubric of military detention: a concept, he explains, "with fairly well-established parameters." His proposal, however, would stretch these traditional parameters beyond the breaking point. While Cole does, in an understatement, acknowledge that the fight against al Qaeda is not a "traditional" armed conflict, the real question is whether it should be deemed an armed conflict at all. The fact that NATO and the United Nations authorized a military response in Afghanistan should not be conflated with international recognition of a global war against al Qaeda: a conflict that is untethered both temporally and geographically, and that has few of the customary attributes of war.

* * *

Even judged by his own proposed criteria, Cole's arguments fail. Cole states that preventive detention should only be deemed permissible "upon a strong showing that the criminal justice system cannot address a serious danger to the community." Yet he cites no empirical evidence to suggest that the criminal justice system is unable to prosecute al Qaeda operatives, and in fact experience to date shows the contrary. Although Cole speculates that a jury might acquit an active and dangerous al Qaeda operative, I know of no cautionary tales-nor, apparently, does Cole-involving terrorists who have actually been set free. If anything, the cautionary tales I know tell the opposite story: one of over-zealous prosecutors and convictions based on thin evidence.

Lacking factual examples to support his claims, Cole relies on logical suppositions. He lists three reasons why the criminal justice system might be incapable of prosecuting enemy combatants, only one of which is arguably relevant to trials of al Qaeda operatives. (The first two-that soldiers are generally privileged to engage in warfare and are legally compelled to do so under their countries' domestic laws-are clearly inapplicable to terrorist groups.) Citing difficulties associated with "battlefield captures" and "the need to maintain military secrets," Cole contends that the criminal justice process will sometimes prove inadequate. Later in his piece, in discussing a recent ruling in the case of Ali Saleh Kalah al-Marri, Cole refers to an additional objection: that al Qaeda operatives might use the criminal process as a forum for passing information to their compatriots.

Some of these concerns are misplaced; others are overemphasized. The first concern-that evidence and testimony collected on the battlefield would be inadmissible in court-is simply a red herring. Alleged al Qaeda operatives are nearly always detained during law enforcement operations, not military confrontations. They are arrested by police in Pakistan, stopped by immigration authorities in Egypt, or-like Ali Saleh al-Marri-picked up at their homes in the United States by the FBI. Moreover, in those cases where evidence or testimony is obtained in an actual battlefield setting, the courts have never rigidly or formalistically refused to admit it. They have not insisted that Miranda rights be read on a battlefield, for example, but instead have taken a practical and flexible approach to ensuring that the right against self-incrimination is respected.

The need for secrecy, while compelling in some instances, is hardly unique to cases involving al Qaeda. Were the names of confidential informants divulged in criminal prosecutions of Hamas, Lashkar-e-Taiba, or Abu Sayyaf members-or members of Mexican drug cartels-similarly harmful results could ensue. The federal courts, at any rate, are fully competent to grapple with these problems. They have relied on the procedures outlined in the Classified Information Procedures Act (CIPA) in countless terrorism prosecutions, including those involving al Qaeda operatives, to ensure that evidence obtained by intelligence sources is admitted without compromising those sources.

Cole is also insufficiently attentive to the negative impact of relying on "confidential information" (called "secret evidence" in his other writings) in hearings relating to preventive detention. At Guantánamo over the past seven years, we have seen what happens when detention decisions are made on the basis of classified information. While detainees may still be able to respond to the allegations against them, they cannot effectively refute those allegations when key evidence is kept secret. To upset CIPA's carefully calibrated balance by allowing decisions based on secret evidence would guarantee unjust detentions.

The remaining concern-about the need to prevent terrorists from passing information-is, once again, hardly unique to al Qaeda prosecutions, nor is it necessarily exacerbated in that context. The federal courts have adjudicated cases of treason, military spying, and passing nuclear secrets. Where necessary, judges have the power to issue protective orders and take other measures to prevent dangerous communications.

* * *

Near the beginning of his article, Cole sounds a warning. If the United States were to establish a system of long-term preventive detention applicable to a broader category of suspected terrorists than he believes is appropriate, then "what would stop its extension to other serious crimes?"

Cole's question is, of course, relevant to his own proposal. There is nothing unique about al Qaeda operatives to sustain a principled distinction between them and other terrorists, just as there is nothing unique about terrorists that would preclude rules designed for them from being extended to drug traffickers, mafiosi, and members of other violent criminal gangs. Indeed, given al Qaeda's amorphous composition, even distinguishing al Qaeda operatives from al Qaeda associates and members of related groups might not be feasible. Were Congress to pass legislation establishing the detention regime that Cole recommends, it would inevitably be stretched to cover other dangerous and threatening groups.

There is a final issue to bear in mind. As the United States tries to regain the moral authority that it has lost in recent years, policymakers should consider this question, too: on what principled basis would this country criticize the use of preventive detention abroad, if it were to be employed here? How could U.S. officials counsel India-a country that repealed a broad preventive detention law in 2004-to deal with the serious threat of terrorism using fair criminal justice norms?

The United States is near the end of an extended experiment in preventive detention, one that the world has watched closely. It has been setting a negative example for seven years; it should aim to take some positive steps now.

This article is a response to David Cole's Closing Guantánamo. Read all responses here.