Question: Mr. President, under the law, how would you justify the practice of renditioning, where U.S. agents . . . [send] terror suspects abroad, taking them to a third country for interrogation? . . .
Answer: . . . We operate within the law and we send people to countries where they say they're not going to torture the people. But let me say something: The United States government has an obligation to protect the American people. It's in our country's interests to find those who would do harm to us. . . . We still [are] at war.

SO SPOKE PRESIDENT Bush at a press conference on April 28, 2005. There is, however, a contradiction in the president's stated objective to use torture-free rendition as an effective counterterrorist weapon, to wit: Some of the countries to which prisoners have been rendered do use torture. Morally, operationally, and politically, then, the United States has got itself into a muddle through its embrace of rendition--that is, the transfer of American-held terrorists and terrorist suspects into the hands of foreign security and intelligence services.

Though the Central Intelligence Agency doesn't comment officially on the policy, it is one the Bush administration inherited from its predecessor, which used it principally against Osama bin Laden's al Qaeda. According to a former CIA officer knowledgeable about rendition in the 1990s, the Clinton administration first briefed the congressional intelligence oversight committees on transfers in the summer of 1995. Even then, the practice had antecedents in the cooperative "liaison" intelligence efforts Langley has run with a variety of Middle Eastern and Asian countries, notably Egypt and Jordan, the two Arab states with which the Clandestine Service has probably worked most closely over the last 25 years. In the mid-1980s, when Duane Clarridge, the contra-supporting, covert-action-loving favorite of CIA director William Casey, transformed the agency's Counterterrorism Center into a serious bureaucratic force at Langley, counterterrorist liaison work with friendly Arab states increased significantly. Joint-counterterrorist and joint-espionage operations have occasionally cast foreign security services as punishers of deceitful agents controlled by CIA case officers. Depending on the sin, retribution could be severe.

A byproduct of Langley's decades-old inability to recruit or secrete agents inside many Middle Eastern organizations, the increasing use of friendly Arab liaison services is actually part of a global pattern, where liaison work has gradually taken priority over "unilateral" (CIA-only) clandestine operations. In the Middle East post-9/11, this liaison dispensation, according to active-duty agency officers, has skyrocketed. Since the late 1990s, the United States has rendered at least 100 men, according to both congressional and CIA officials, into foreign hands. Most of these transfers have occurred since 9/11, which changed a low-profile counterterrorist tool into standard operating procedure, critical to the way the CIA wages its battle against Islamic extremism.

The moral issues surrounding rendition are what has caught the attention of the press and both Democratic and Republican members of Congress. One regularly hears in Washington a conjecture that rendition may have been the top of a slippery slope that led to very rough tactics at Guantanamo Bay and to Abu Ghraib's abusive antics. This view may have merit. If the CIA is rendering detainees to foreigners because of the efficacy of their aggressive interrogations, then the CIA is acknowledging the utility of heavy-handedness in counterterrorist debriefings. And the agency, by all accounts, does use harsh methods at its own facilities against members of al Qaeda. (Whether these methods constitute torture is another question.) Such a professional disposition could have seeped into intelligence units in the Pentagon that work with the agency. Military intelligence often follows Langley's custom and practice, though it hates to admit it. And once professional ethics start to crack, an implosion can occur. Yet it is entirely possible, if not probable, that military outfits came up with the Guantanamo and Abu Ghraib techniques strictly on their own.

The interrogation methods used at Guantanamo appear to have been derived overwhelmingly from those U.S. Special Forces are trained to resist. And what Americans did at Abu Ghraib could easily have sprung from their modern MTV imaginations, or from a more educated though juvenile understanding of Arab sexual ethics discussed in the taboo-exploring 1973 book The Arab Mind by the late scholar Raphael Patai. Patai seems to have a following in certain quarters of the military; the foreword to the paperback edition of his book was written by a respected former military officer. But the central importance that Seymour Hersh gave to Patai in his essay on Abu Ghraib in the New Yorker is likely overstated. Intellectual influences and traditions in the military, the most hierarchical of organizations, always leave a paper trail, even on subjects that might embarrass non-military personnel. Given the anti-Iraq war sentiment among many Army officers, we would have heard much more detailed commentary about Patai's influence were it actually as great as Hersh's anonymous informants suggest. (In particular, the assertion that the book was "the bible of the neocons on Arab behavior," and thus had a wide prewar impact, is odd given how few neoconservatives have read it--I did an informal survey--and how seldom neoconservatives have Freudian conversations with military officers.)

It is possible that rendition slowly fueled an American willingness to try debriefing methods used by Arab or other Muslim security services against terrorist suspects. One doesn't have to be a fan of Patai to realize that sex is a sensitive issue in the Muslim Middle East, and the security services of modern dictatorships have been known to transgress morality in the pursuit of regime security (Saddam's certainly did). Does anyone think that the internal-security black-clad ninjas of Algeria's military junta--they call themselves les exterminateurs--would scruple to use some form of sexual intimidation against Islamic holy warriors trying to topple the regime if they thought it effective? We definitely have had an expanded military and intelligence liaison relationship with the Algerian military since 9/11. Nor do other, more polite, dictatorial regimes in the Arab world have so rigid a code of ethics as to prevent them from pressuring detainees' ids. Yet this sequencing of cause and effect--the idea that U.S. practice deteriorated under the influence of rendition--runs counter to the CIA's objective in the exercise, namely, to gain distance from the interrogation of suspected terrorists by sending them abroad. What you don't see, you can't teach. The Pentagon and the CIA would have produced reams of paper about the educational impact of rendition, had it existed, and such documents have not so far surfaced. They surely would have if they'd been even tangentially connected to Abu Ghraib. The inspector general's office at the CIA is sufficiently competent, morally attentive, and politically astute to discover and expose any substantial cross-fertilization.

Yet even if we assume that rendition did somehow lead to Abu Ghraib, should we therefore conclude that rendition is wrong? Worthwhile actions can have adverse consequences that don't necessarily negate the value of what was originally done. America killed huge numbers of civilians in World War II, yet only the America-hating fringe believes that those deaths nullify the military and moral triumph of the United States over Nazi Germany and Imperial Japan. Advocates of rendition can argue that the practice saves lives in fighting terrorists who would use weapons of mass destruction if they could. In such circumstances, good intelligence, however obtained, could save tens, even hundreds of thousands of American civilian lives. And Langley's hard-nosed bureaucrats can suggest that it is better to have the "ragheads" do the dirty work than for Americans to sully themselves. It's a decent bet that many Americans, liberal or conservative, would privately agree with that sentiment. It is an ugly world out there, and al Qaeda is to a considerable extent an Arab problem, so why not let Arab allies aggressively interrogate our Arab enemies? They can do this better than we can, so this argument goes, since it is their culture and they've had lots of practice.

SO IS IT wrong to send a suspected or actual member of al Qaeda to Egypt or Jordan for the purposes of aggressive interrogation? One must assume that any Muslim the CIA deposits into the hands of the Egyptian or Jordanian security and intelligence services will be tortured. The services of neither country cherish the Marquess of Queensberry Rules inside their prisons, especially for those guilty of violent political behavior. If rumors are true, bastinado is a type of foreplay among the security organizations in both lands. The "Ottoman pull-up," where the hands are tied low behind the back and the entire body is then lifted from the wrists, is also a common, more pulverizing method to get the mind and body to quiver. And these are primitive techniques in a region of the world where state power is thoroughly modernized. Only democratic Turkey has begun the difficult moral and legal evolution that has led Western democratic societies away from torture as standard police and judicial practice.

If we are to be brutally honest, the compelling reason why Washington has backed rendition is that the Clinton and Bush administrations wanted our Arab allies to do what we can no longer countenance by our own hand (and anyone who thinks Guantanamo and Abu Ghraib disprove this point is ignorant of the history of professionally administered torture). If we understand rendition as do many CIA intelligence officers, we know it has two great advantages. The outsourcing of torture is, especially for the Americans who must otherwise administer the pain, easily the more attractive.

The other big plus is that rendition eliminates the Guantanamo detention problem (or, in the case of the Clinton administration, criminal court cases that national security adviser Samuel Berger always feared losing). Proving guilt in a U.S. civilian or military court through the use of even rock-solid, politely obtained intelligence can be extremely difficult. Proving guilt in such a court with the use of similarly hard intelligence gained through physical abuse is impossible.

Rendition also solves the problem of how to deal with minor-league would-be Islamic terrorists or guerrillas who may or may not have had the United States in their sights. These are individuals who are guilty by association with al Qaeda and its allied groups but would never, ever be prosecuted in an American civilian or military court for lack of legally admissible evidence.

The large-scale renditions to Uzbekistan that the New York Times has reported on were probably mostly of this petty, Guantanamo-avoidance type. Central Asian Islamic extremists picked up in Afghanistan after the fall of the Taliban have probably not yet fully embraced the overriding America-focused hatred of al Qaeda (though with continued U.S. backing of Uzbek president-for-life Islam Karimov, this might change). Culturally and linguistically, these al Qaeda-allied militants have limited operational range; they are what Ahmed Shah Massoud, the murdered leader of the Afghan Northern Alliance, once called "Islamist cannon-fodder" for the Taliban. Whatever information the Uzbek security service could beat out of these men would be considered a tangential benefit to the CIA--while removing them from its hands, or the hands of the U.S.-backed government in Kabul, would be a blessing. Transfers of such detainees further cement the 14 years of pretty warm relations between Langley and Tashkent's security and intelligence services--easily, according to active-duty CIA officers, the best liaison cooperation the agency has enjoyed with any Central Asian state since the crack-up of the Soviet Union.

Given the legalistic nature of the CIA, Michael Scheuer, head of the agency's Osama bin Laden/al Qaeda desk in the late 1990s, is undoubtedly right about the high level of oversight these practices have had. Scheuer wrote recently in the New York Times, concerning rendition, that he had never seen "a set of operations that was more closely scrutinized by the director of central intelligence, the National Security Council and the Congressional intelligence committees," or "that was more blessed (plagued?) by the expert guidance of lawyers." This is not at all the impression that Republican senator Pat Roberts tried to convey in April at confirmation hearings for the national intelligence director, John Negroponte. "There has been a great deal of discussion, Mr. Ambassador, about the U.S. government's involvement in interrogation, rendition, and detention of terrorists in the global war on terror," the chairman of the Senate Select Committee on Intelligence opened. "[C]an you commit to us that as the DNI you will ensure the intelligence community's activities comply with the Constitution of the United States and all applicable laws and treaties . . . ?"

In times of war and great stress, the capacity for officials to not know what is unpleasant is large and most human. Is it unreasonable to believe that for both Democrats and Republicans the pictures of Abu Ghraib have muted the sounds of human beings falling from the Twin Towers? As memories of 9/11 fade, we have retreated to pre-al Qaeda ethics where all of us are more comfortable.

BUT IF WE leave aside for a moment the question of whether torture can ever be a morally legitimate counterterrorist tool, we can examine a different question; that is, the efficacy of rendition. And this is one place where the practice is really indefensible. Though agency officials will not publicly acknowledge its use, CIA case officers and knowledgeable analysts do so readily in private. In his approval of rendition, Scheuer is probably representative of a CIA consensus, at least within the Near East division of the Directorate of Operations. It is, other agency officers will tell you unofficially, an ugly, but needed tool against Islamic holy warriors. But is it? It shouldn't be that hard to understand that rendition is actually an abdication of intelligence professionalism. It is more evidence--there is so much--that the Directorate of Operations is a broken institution and increasingly the obstacle to developing more effective counterterrorist tactics for the United States overseas.

A cardinal rule of the intelligence business--many case officers would call it the first principle of collecting human intelligence--is to maintain control of the individuals you are debriefing or interrogating. It is for this reason that, historically, the CIA has looked askance at "liaison" intelligence and always has clearly marked, at least for consumers within the intelligence community, the foreign provenance of such information. This intelligence might be good, even excellent, and when the CIA could corroborate it with its own intelligence or verifiable fact, then such liaison intelligence could be labeled, with appropriate warnings, as worthy of serious consideration. Still, it was always seen as somehow tainted, even when it came from the "Anglo-Saxon cousins"--the British, Australians, and, before they went a bit dippy at the close of the Cold War, the New Zealanders and Canadians. Whenever possible, the CIA always strove for greater knowledge of, and often greater access to, the human sources of this information. Joint operations and debriefings could develop: In the process, foreign intelligence and security agencies could earn goodwill and chits with Langley, sometimes redeemable through CIA-provided goodies and services.

Now, what Langley is doing through rendition is the reverse. The United States is voluntarily diminishing, if not ending, its dominion over intelligence sources it controls by transferring them to other countries. And we're not talking Great Britain, France, or Israel, with whom the United States has the ties that are possible only with Western democracies, but Arab and Muslim states with whom we have, especially since 9/11, more complicated relations. If it is true, as it strongly appears to be, that Langley actually transferred individuals to Syria--an officially recognized state sponsor of terrorism--then rendition has taken Langley, the White House, and Congress into a positively surreal "realist" world.

The mind spins thinking how agency officials would phrase the sourcing notes on intelligence collected from Syrian debriefings: Information collected by a foreign intelligence service that the United States now strongly suspects is aiding Iraqi insurgents; this intelligence service also has a long history of operationally aiding Palestinian terrorist organizations and the Lebanese Hezbollah, and colluding with Iran's Revolutionary Guards Corps; this intelligence service may also have arranged the murder of Lebanon's former prime minister, Rafik Hariri, and is one of the great drug-smuggling organizations in the world. However, in this particular instance, we believe we have properly calibrated this service's possible ulterior motives and deem the intelligence collected by this service to be worthy of dissemination to the intelligence community for its consideration.

Now, it may well be true, as Scheuer argues, that rendition sometimes produced intelligence information that was valuable against al Qaeda. But surely the Central Intelligence Agency could have obtained the same information if it had applied similar interrogative techniques. With in-house control, the agency could have had greater confidence in the information collected, since it would have been guiding and monitoring the process 24/7. From the perspective of an intelligence officer, it makes absolutely no operational sense to have someone torture for you if you have the option of doing the dirty work yourself. (And the United States certainly has the capacity to "false flag" an interrogation--make it appear that non-Americans are in charge--if this is deemed advantageous.) Since the highest-profile al Qaeda subjects, Khalid Sheikh Mohammed and Ramzi bin al-Shibh, both arrested in Pakistan, were quickly rendered to the Americans, it seems plain that in the case of the most important assets, we insist on maintaining control ourselves. If rendition were operationally superior to in-house aggressive interrogation, the CIA would have been derelict in not leaving these two masterminds in foreign hands.

Furthermore, rendition is not necessary as a means of drawing on specific national or regional expertise in a given interrogation. If the CIA on occasion desires such aid, Egyptian president Hosni Mubarak or King Abdullah of Jordan is no doubt thrilled to provide the officers required. The more these two leaders can build a clientele within the U.S. government, the better they can counter President George W. Bush's pro-democracy foreign policy. Imagine the National Security Council's democracy promoter, Elliott Abrams, arriving in Cairo to warn Mubarak not to rig national elections or imprison democratic dissidents, while CIA personnel fraternally engage Egyptian security and intelligence officials in questioning al Qaeda suspects on Guam. These are the contradictions in U.S. policy that anxious dictators and kings live for. All the CIA ever had to do was whistle if it needed foreign talent to make up for any American deficiency in its interrogations.

As for those who believe that torture itself is an ineffective intelligence tool--and a wide swath of the intelligence community, after the Abu Ghraib scandal and before President Bush's reelection, seemed to think so--rendition makes no sense whatsoever. It would amount to the United States willfully diminishing the flow of reliable information. John Negroponte appears to be in this camp. He remarked in his confirmation hearing, "Not only is torture illegal and reprehensible, but even if it were not so, I don't think it's an effective way of producing useful information." If Negroponte sincerely believes what he said, then the practice of rendition is over.

The CIA, however, may know something that Ambassador Negroponte will reluctantly grow to appreciate. Anyone who has had serious pain intentionally inflicted upon him or her--and clandestine-service junior officers receive a small but sufficient dose in their training--knows that the truth has an imprint upon the brain that can withstand the distress, confusion, and loneliness of aggressive interrogation. Lies don't have the same tenacity. There is a reason why the former prisoners of war who briefed my junior-officer class on the pitfalls of imprisonment warned that the truth will come out. No professional intelligence officer alive wants to torture people. No moral man isn't repelled by the damage done to the victim and the perpetrator. But sadism and primitive notions of justice and salvation aren't the only reasons why men, particularly men in danger, have had recourse to torture for millennia.

IF NEGROPONTE tries to end rendition, Langley is probably going to resist. Most important, rendition appeals to the CIA because it is easy. Having others do your work for you is always bureaucratically commendable. And rendition leaves no counterterrorist debris. In foreign hands, terrorist suspects just disappear into the cells of Middle Eastern prisons or--for the lucky, with time--filter back to their homelands, where, inshallah, they will cause no further harm, at least to the foreign country that incarcerated and tortured them at America's request. Rendition gives the CIA power and clout in Washington. It has become an integral part of America's counterterrorist modus operandi--a thing nobody really wants to talk about but most probably view as valuable. But why exactly should the rest of us take Langley's word for it that rendition is an essential weapon?

If the Bush administration were wise, it would immediately set up a small bipartisan group, chosen by members of the intelligence oversight committees, to examine in fact how uniquely valuable the practice has been. Odds are such a group, if it got into the weeds and compared the intelligence gathered by "them" and by us post-9/11, would have a hard time being effusive about the program. The Central Intelligence Agency's decades-old quest to become a barely covert version of Foggy Bottom has left the Directorate of Operations increasingly captive to the preferences and good will of foreign intelligence and security services. This irresponsible trend can lead to trouble.

It is easy to envision the administration getting publicly raked over the coals in the not-distant future for some rendition mishap. In that event, Congress would surely feign ignorance of its own tacit assent to torturing foreigners. To forestall this, the administration should force itself and Congress to write down what they truly think about rendition, both its performance in the past and its justification as an ongoing practice. They should review the past, enumerating when and where each responsible congressional or executive-branch party was briefed about "Muhammad" being sent to Egypt, or "Ahmad" being sent to Jordan, or "Umar" being sent to Saudi Arabia. Our elected representatives should be obliged to look at the intelligence return on the questioning of such men, and then to reapprove the questioning techniques.

In so doing, good men of conscience would realize that the United States must have room in its foreign and counterterrorist policies for the repatriation of detainees who are not anti-American terrorists and who cannot stand trial in U.S. civilian or military courts. In this war against Islamic extremism, we cannot avoid vacuuming up some individuals who hate us and who may have an ideological sympathy or association with militant groups that have done or would do us harm. Detainees selected for repatriation would not be full-blown members of al Qaeda or allied terrorist groups. They might not have been involved directly in terrorist acts, if only because they lacked the will to become holy warriors. This is the way a thoughtful, humane counterterrorist program must sometimes work. (A non-humane alternative would be to practice Algerian counterterrorism, where the emphasis is on killing, not on incarceration and patient interrogation.)

Without seeing the interrogation and debriefing notes of all those the United States has put into Guantanamo Bay and other U.S. facilities since 9/11, it's impossible to know for sure, but there's a decent chance that many if not most of those Washington has detained fall into the above category--Islamic militants not proved to have committed specific terrorist acts. The United States really has only two options for these men: keep them incarcerated until they're old and broken and then send them back to their homelands, or repatriate them as soon as the U.S. government knows they fit in the above category. It is possible, if not likely, such men would endure some torture upon their return home--some could even be killed, though the CIA probably couldn't figure out beforehand which ones would provoke such wrath. But this is the fate of many in the Muslim Middle East. (Just ask a member of the Muslim Brotherhood in Egypt--now, at this moment, a tolerated fundamentalist group--whether he knows people who have been tortured and freed. He'll probably give you a very long list.)

Repatriation is unpleasant. It is ugly and unjust for the truly innocent who get sucked up into an American or allied dragnet and then are returned home, suspicions having been raised by their foreign incarceration. But this outcome for non-lethal Islamic militants and the innocent is surely better than spending years--decades, perhaps--in American jails waiting for late middle-age to rob them of their capacity to cause harm. No "safe" third country will take in these men. No American president is going to let them, at any age, take up residence in the United States. With all its problems, repatriation is the most humane option Washington can choose for the collaterally damaged in the war against Islamic terrorism.

REMEMBER: Repatriation isn't rendition, which seeks to use foreigners as front-line interrogators of terrorists and terrorist suspects (and only as a welcome side effect reduces the population at Guantanamo Bay). Eventually, the contradictions between the practice of prisoner transfer, if it continues, and the Bush administration's intent to work for the democratic transformation of the Middle East could paralyze the administration's foreign policy.

For the "realists" in the government, there is perhaps no aspect of America's counterterrorist strategy that is more appealing than rendition, since it binds the CIA and foreign services together in a very hard-nosed and (the realists believe) effective way, and it obliges our allies to do most of the heavy lifting. It's possible to imagine President Bush reining in the intelligence and counterterrorist establishment in Washington that has grown accustomed to rendition and is determined to maintain established liaison relationships and practices. The president could easily say to the concerned bureaucracies that any counterterrorist practice that works against the expansion of democracy in the Middle East is unacceptably counterproductive. The growth of democracy among Muslims is the only sure way of breaking bin Ladenism. Now, not later, the United States must wean itself from compromising alliances with autocratic regimes, no matter how "pro-American" specific joint-intelligence operations may be.

Of course, it is also possible to imagine the administration backing off its determined promotion of democracy in the Middle East, as it realizes that elections in some countries, particularly Egypt, could lead to fundamentalists' gaining significant power. There is an odd alignment brewing in Washington between sincere advocates of liberal democracy in the Middle East and realist types who don't at all see the spread of democracy as a necessary or urgent antidote to bin Ladenism. The liberal-democracy advocates live in fear of Islamic activist gains through electoral politics. They don't see this as part of a beneficial and inevitable evolutionary process within Islam. These liberal democrats think you can encourage and empower progressive Muslims over the heads of illiberal dictators and kings, so when democracy finally dawns in the Arab Middle East there will be little to fear. The progressives will have triumphed over those who too thoroughly mix politics and faith. This utopian mindset, combined with the realist critique and understandable feminist anxiety about an electorally powerful conservative Islam, could gain the high ground in this administration. And whatever else this alliance does, it will powerfully reinforce the rendition ethic in Washington. When you're taking such a long view of President Bush's "forward strategy of freedom," short-term compromises become easier to bear.

But fate may be kind to the president. The number of captured holy warriors appears to be in decline. Both Afghanistan and Iraq appear to be working. The "golden age" of rendition may well be over. The administration may not have to worry seriously about tradeoffs between democracy and counterterrorism in the Arab world. Then again, in the not-too-distant future the CIA or the Pentagon could discover a new cesspool of bomb-happy anti-American jihadists in a foreign land, and have no alternative but to kill and capture as many as possible. The tension between democratic values and counterterrorist imperatives could become acute. It would behoove the administration to know for sure whether that tension should even exist--does rendition do anything for us that we cannot do better ourselves?--before such a scenario develops.

ALL OF THIS, of course, jumps over the great debate the United States has been avoiding since 9/11. Are there circumstances where the American government ought to countenance the use of torture? Moral people are often forced to weigh ideals in urgent and unforgiving circumstances. In extremis, we do certain things. Even uncompromising ethical codes based on divine authority usually recognize, however reluctantly, the doctrine of "the lesser of two evils." But the extreme case--the terrorist cell with a weapon of mass destruction in, say, New York or London--is the relatively easy one. How many lives must be threatened before "lesser evils" may legitimately be chosen? What if a terrorist suspect doesn't have information relevant to an imminent threat? Is it ever legitimate to torture a man or woman to extract information that may possibly save tens of thousands of lives in the future?

Professor Alan Dershowitz of Harvard tried to provoke us to think about this issue after 9/11 with his discussion of "writs of torture." If the administration had tackled the issue then and forced the country to have an open debate about the ethics of combating mass terrorism, it's likely there would have been greater clarity at the Pentagon and the CIA about what to do against whom, when. We would have less need of lawyers and of the moral camouflage, obfuscation, and cowardice that often come with top-secret classification. Our present muddle over rendition might not have arisen.

Even now, having this debate could stop us from doing--or not doing--something that our collective national conscience would later regret. The most extreme reactions usually occur when people fail to prepare themselves thoughtfully for easily foreseeable situations. One thing is certain: Our avoidance of this necessary debate is a disservice to the men and women of the CIA, the Pentagon, and the FBI who may one day be called upon to do the unspeakable if once again we feel imminently threatened.

Reuel Marc Gerecht is a resident fellow at the American Enterprise Institute and a contributing editor to The Weekly Standard. He is a former case officer in the clandestine service of the CIA.

Next Page