In the years since September 11, 2001, amateur plane spotters around the world have tracked the movements of what started out as an unidentified flying object, a phantom jet. The plane, a Gulfstream V, the kind of small, sleek private aircraft favored by movie stars and business executives, was first spotted in October 2001 in a remote stretch of the Karachi airport. Eight weeks later it appeared at Stockholm’s Bromma Airport, then at a military airfield in Jakarta. Last fall the London Times reported that the plane had flown to some fifty destinations outside the US, from Guantánamo Bay to Egypt, Jordan, Iraq, Morocco, Afghan-istan, Libya, and Uzbekistan.

In December The Washington Post revealed that the plane is registered to several persons who do not exist—fictitious names, post office boxes, and Social Security numbers. The plane is used, the Post explained, for “rendition”—the relatively new tactic employed by the CIA and the Bush administration to move detainees from one country to another for interrogation. The Gulfstream facilitates a kind of race to the bottom, whereby suspects are whisked out of countries that are reluctant to use torture as an interrogation technique, and into countries that have no such inhibitions.1

The chance sightings of this mystery plane are an apt illustration of the occasional and fragmentary glimpses that civilians get of the vast realms of American military and intelligence activity that are secret. According to the Information Security Oversight Office of the National Archives, which monitors American classification policy, the American government created over 15 million secrets in 2004—or some 40,000 individual classification decisions every day. When occasional slips occur—the sighting of the Gulfstream V, the leaked photographs from Abu Ghraib—we come to appreciate how much of what our government does we simply cannot fathom.

The linchpin of the classification system and the device that allows government secrets to stay secret is the code name. Every program, exercise, and operation has a code name. Individuals, offices, and enemies are all granted code names. Only those who are cleared to know certain information will know a code name, and the higher the clearance, the more extensive one’s code vocabulary. Indeed, so Byzantine is the classification hierarchy that even the different levels of clearance to know secret names have secret names of their own. Thus, William Arkin points out in his new book Code Names, in the months after September 11, 2001, those at the Pentagon who had a security clearance for information that was marked Secret could not gain access to information that was Top Secret. But those cleared merely for Top Secret often did not realize that there was an even higher level above that, known by the code name Polo Step. “Polo Step was used to confine highly sensitive Iraq and counter-terrorism war planning to a small circle,” Arkin explains.

In June 2002, Arkin published the fact that there was a code name called “Polo Step” for Iraq war planning in the Los Angeles Times. Hearing of the leak, General Tommy Franks, head of Central Command, was outraged and immediately contacted Donald Rumsfeld, who proclaimed the leak “egregious” and launched an investigation to uncover the identity of the person who divulged the name. The Air Force Office of Investigations proceeded to interview over one thousand people, many of whom had never had clearance to know what Polo Step was in the first place. The hunt ended up costing $1.5 million, and they never caught their man.

If the effort to chase down the source of that one leak was any indication, the publication of Code Names may bankrupt the Pentagon, for the book is a compendium of some three thousand secret names that Arkin, a longtime intelligence journalist and military analyst, has compiled. Arkin believes that the classification system has developed a momentum of its own, uncoupled from the legitimate demands of operational secrecy, and become a life support system for policies and activities—like the invasion of Iraq—which would not hold up under the light of public or congressional scrutiny. Over the course of a decade he jotted down and collated code words he picked up in his reporting, and assembled a most unusual book—a dictionary for a secret language.

“Code names seemed a good way to try to organize and understand the range and breadth of American military activity in the world,” Arkin writes. “Each code name tells a story of some expenditure of tax dollars and is tangibly connected to some activity with a complex institutional history and significance.” He suggests that in sum the code names will represent “an anatomy, a sort of DNA map of American national security.”

The reader would not grasp that anatomy by reading Code Names straight through, and Arkin has deliberately arranged the book as a reference work: the items are listed alphabetically, and each name is followed by a brief description of the program or entity described. Reading them through one is struck by the curious poetry of code words. For the many chest-thumping entries like Mighty Thunder (an air force–sponsored exercise series), or Resolute Strike (a coalition operation against Taliban forces in southern Afghanistan in 2003), there are entries like Calypso Wind, Optic Windmill, Platypus Moon, Sorbet Royale, and, delightfully, Zodiac Beauchamp. Some of the entries are followed by quite detailed descriptions, whereas in other cases little more than the name itself is available. Acer Gable, for instance, is described as “Air Force unknown project or weapon.” In this respect Code Names is not so much a taxonomy as a kind of archaeology. Arkin must draw together scattered pieces of evidence; whenever possible he deduces from that evidence something larger, and when he cannot he simply presents the artifact as is. For instance, he explains that the term “Talent Keyhole” is used to designate intelligence gathered from communications intercepted by satellite. He knows that the word “Gamma” is a more specific designation for certain kinds of Talent Keyhole intelligence—but he doesn’t know what precisely that subcategory represents.


As such, Code Names may be more effective as a provocation—an act of protest against a bloated classification system—than as a Rosetta Stone for scholars, practitioners, or intelligence hobbyists. Arkin suggests that “an article that includes a never-before-published code name adds a rare authenticity, because it suggests not only sourcing close to real secrets” but also a source “who is courageous (or reckless) enough to make an unauthorized revelation to the media.”

This is true enough. But Arkin may overlook a broader irony of the cat-and-mouse game played by the defense establishment and the press. Reporters will often struggle to reveal the name of a particular program, and be lauded and reviled in equal measure when they succeed in doing so. But the moment the name is published the military and intelligence powers in question simply change the code name. Thus by the time the Air Force Office of Investigations began polygraphing civil servants to determine who had compromised Polo Step, the name Polo Step had almost certainly been abandoned, relegated to the formidable scrap heap of discarded names and replaced with some new—and for the time being secret—designation.

Like cherry blossoms or shooting stars, secret names are evanescent, gone the moment they are apprehended by those without the need to know. This perishability is something of a problem for Arkin’s book, since, in the very act of presenting the revelations that it does, Code Names runs the risk of rendering itself obsolete. While those who oversee the classification system will not be able to chase down the leakers of the thousands of code names cataloged in the book, it does seem certain that any entries that are still in active use will promptly be changed. For all of their value in providing a sense of institutional hierarchies and national security secrecy, code names themselves are intrinsically ephemeral, interchangeable, and disposable.

This in no way diminishes Arkin’s broader argument against overclassification, however. Traditionally the tension between secrecy and openness when it came to the intelligence and defense communities was rendered as a contest between the demands of national security on the one hand, and liberal democratic principles of transparency on the other. In that tension, secrecy won more often than not, because the exigencies of national security enjoy one of the larger loopholes in constitutional and statutory law. Certainly some degree of secrecy is necessary in order to maintain the security and effectiveness of American operations. In 1998, the Washington Times revealed that the National Security Agency (NSA) could eavesdrop on Osama bin Laden’s satellite telephone. No sooner had the story appeared than bin Laden stopped using the telephone, effectively disappearing from the radar screens of US intelligence. This kind of leak has disastrous effects; Michael Scheuer, a twenty-two-year veteran of the CIA who ran the bin Laden desk at that time, recently told a gathering of intelligence officials in Washington that he believes you can draw a direct causal line from the publication of that story to the attacks of September 11.2

The problem is that cautionary tales of this sort have led to a pronounced overreaction. The NSA, the largest intelligence agency in America (and the world), recently announced its intention to summarily reject requests brought by civilians under the Freedom of Information Act (FOIA), arguing that since the overwhelming majority of requests would ultimately be rejected after consideration anyway, it would be inefficient to assign employees the task of actually considering individual applications. This blanket of secrecy applies not only to information on current budgets and operations, but also to historical information. When petitioned under the FOIA for the aggregate intelligence budget for the year 1947, a matter of considerable interest to historians and none whatsoever to terrorists or other foreign adversaries, the CIA recently replied that it would not divulge the figure, on the absurd but insurmountable grounds that to do so would jeopardize national security.


But since September 11, and particularly since the publication of the 9/11 Commission’s report this past summer, the traditional tension between openness and security has been inverted. The sorry image that has lately emerged from the protracted post-mortems of September 11 is that of an intelligence community hopelessly undone by its internal rivalries, alienated from the business and academic communities, and paralyzed by an inability to communicate—an inability that was symptomatic of arteries of communication clogged with classified data. In this sense, secrecy can actually be bad for national security. “Secrecy stifles oversight, accountability, and information sharing,” the 9/11 commissioners concluded. “Unfortunately, all the current organizational incentives encourage overclassification. This balance should change.”3

Humbled by a sequence of disastrous failures, defense and intelligence agencies have been forced to concede that operating without the scrutiny of Congress or the press may in fact be deeply hazardous. The 9/11 Commission suggested that public disclosure is “democracy’s best oversight mechanism,” and that the press, long regarded as an adversary by defense and intelligence officials, could actually serve as an effective check on policymaking. Polo Step was a very exclusive club, after all: a limited number of officials were privy to the early plans for an invasion of Iraq. And perhaps on reflection that was the problem. Bad policies and ill-fated expeditions are the inevitable outcome when skeptics are turned away at the door.

Yet precisely when it seemed likely that revelations about the dangers of secrecy would lead to better communication between intelligence agencies and the outside world, and more rigorous scrutiny of the defense establishment by Congress and the press, the opposite happened. As the NSA’s blanket exemption from the Freedom of Information Act and the rising tide of classification make clear, defensive officials have pulled the shroud of secrecy ever tighter in the years since September 11, 2001, and, Arkin says, there has been a “secrecy explosion.”

According to the Center for Strategic and Budgetary Assessments, “black,” or secret, programs represented some $23.2 billion, or 17 percent, of the Pentagon’s total budget for 2004, the highest proportion since 1988. When it comes to defense allocations, this kind of secrecy can have perverse effects. The single largest item in the intelligence budget for 2005, for instance, is a highly classified photo-reconnaissance satellite known as Misty, which will ultimately cost $9.5 billion and will be of questionable use in actually keeping America safe.

Several members of the Senate Intelligence Committee argued that the program represented a wasteful misallocation of money that had not been adequately debated by Congress, and that it would at best duplicate technologies which US intelligence already has access to. Because the satellite has such an enormous price tag it would, they said, divert valuable resources from other, more effective intelligence activities. As such, Senator Jay Rockefeller, vice-chairman of the Intelligence Committee, suggested that the expenditure might actually be “dangerous to national security.” Nevertheless, the satellite received its funding, the senators were widely rebuked for having raised the matter at all, and the Justice Department considered investigating them for having done so. In this manner a major decision that will have important effects on the security of the United States was made without aggressive or informed debate, even on the part of congressional committees charged with conducting closed-door oversight of secret intelligence.4

If national security in and of itself is no longer a justification for this degree of classification, one might be inclined to ask what is. Part of the answer is simply that the culture of secrecy is integral to the US defense bureaucracy. Code names and passwords have traditionally been the province of guilds and clubs, a way of keeping others out for the sake of keeping them out. In their modern form, they suit the turf wars characteristic of government: one way to limit scrutiny of some publicly financed plan, from the war in Iraq to the Misty satellite, is to limit the number of initiates who are in on the plan.

This argument is hardly new: it was best expressed by the late Senator Daniel Patrick Moynihan in his book Secrecy.5 Moynihan initiated a bipartisan congressional commission on government secrecy in 1994, which concluded that the self-preservation of petty bureaucrats was at least in part to blame for the bloated classification system that the American government maintains today—a system that costs taxpayers more than $6 billion every year.6 Moynihan traces this idea back to Max Weber, who declared that

The pure interest of the bureaucracy in power…is efficacious far beyond those areas where purely functional interests make for secrecy. The concept of the “official secret” is the specific invention of the bureaucracy, and nothing is so fanatically defended by the bureaucracy as this attitude, which cannot be substantially justified beyond these specifically qualified areas.

Arkin describes the secret-keepers as a “priesthood” and a “cabal,” and he explicitly argues that the kind of institutional secrecy and bureaucratic information hoarding that was at least in part to blame for the attacks of September 11 has not only endured in subsequent years, but has indeed intensified.

In fact, Arkin has recently discovered at first hand just how fanatically some will defend the system of official secrecy. In March the Washington Times journalist Bill Gertz, who has built a career on publishing leaked information from his many sources in the defense community, received a document that seemed to implicate Arkin as an Iraqi spy. According to the document, a classified cable from the Defense Intelligence Agency, “preliminary reporting…indicates possible US citizen William Arkin received monthly stipend for period 1994–1998” to report on “United Nations Special Commission activities.” When Gertz telephoned Arkin for comment, Arkin was baffled, and suspected he was being framed. His suspicion was supported by the Pentagon, which deemed the cable a forgery. It would seem that Arkin’s book has struck a nerve, and that the avid protectors of secret government feel deeply threatened by his disclosures.7


Arkin’s title does some injustice to the book, for in addition to the codex of secret names he has provided an extremely useful survey, again arranged alphabetically, of America’s military and intelligence allies around the world. The phantom Gulfstream V has on a number of occasions picked up prisoners and transported them to Egypt. Several of those “rendered” to the Egyptian authorities in this manner have claimed that they were tortured by their new captors, that the United States was effectively outsourcing torture to its less scrupulous allies. If these claims are true, they represent an exceedingly intimate relationship between American and Egyptian military units, and Arkin provides strong evidence that this kind of back-channel cooperation in the war on terror is no longer an exception but a widespread practice.

Arkin tells us that King Abdullah II of Jordan permits American special forces to train and operate on Jordanian soil, American eavesdroppers to establish covert positions near the Iraqi and Syrian borders, American CIA officers to investigate Jordan’s Iraqi population, and a secret American Army intelligence team (code name: Gray Fox) to operate inside the country. Before the war, the “overt side of US–Jordanian relations…provided cover for a quiet American buildup toward war against Iraq,” Arkin observes. “The public problem associated with all of this was that like many of its Arab neighbors, the Amman government also had to cater to public sentiment that strongly opposed a US war against Iraq.” Despite the obvious presence of American troops in Jordan the government continued to officially deny the existence of any alliance or agreement with the United States. So strenuous was the effort to keep everything under wraps that the two American air bases established in Jordan were code-named West Wing (as in west of Baghdad) in order to avoid referring to Jordan by name in any American planning or correspondence. Thus, leading up to the invasion of Iraq,

Covert intelligence operatives and commandos from the US, Britain, and later Australia arrived in country [i.e., Jordan]. “Black” special operations helicopters and gunships deployed to the Jordanian desert. Patriot surface-to-air missile batteries showed up in Amman…. The very same deployments were being publicly announced in less sensitive Kuwait to the south. But the deployments in Jordan were officially secret.

And it almost goes without saying that one of the places where the CIA’s rendition jet has regularly been spotted over the last three years is a military airport in Amman.

Arkin suggests that there are “dozens of Jordans around the globe.” He cites more than twenty countries that provide bases or other facilities to the US military; seventy-six that have granted permission for US military planes to land on their soil; and eighty-nine that grant overflight rights to those planes.

This second section of Code Names lists some 206 countries and territories that cooperate with the United States on military and intelligence matters. Arkin lists not only the many American bases on foreign soil, but the bilateral and multilateral agreements that define the relations between these defense services. Not all of these are secret, and the extensive reach of the US military throughout the world is widely known, but in collating all of the available information on these agreements, Arkin’s book provides an important service as a guide to the American presence around the planet. It is a companion in this respect to Arkin’s earlier book, Nuclear Battlefields (1985), which lists the locations of the known nuclear installations throughout the world.8

The close ties of some countries, like the United Kingdom, Poland, or Australia, to the US are well known. Others are less so. In Code Names Arkin tells us that Kyrgyzstan allows American combat and support planes to make use of a base adjacent to Manas International Airport, and that despite the absence of diplomatic relations between the US and Bhutan, Bhutanese military officers have attended the Pentagon’s Asia-Pacific Center for Security Studies.

Like Jeffrey Richelson’s classic reference work The US Intelligence Community,9 and The Ties That Bind10 by Richelson and Desmond Ball, Arkin’s survey of America’s allies is full of useful information not only for scholars and practitioners of intelligence, but for any serious newspaper reader. And while in Code Names Arkin is often at a loss to detail the significance of a name or explain the function of a program, his lacunae nevertheless tell us much about the vast and invisible realms of our government’s secret activity. Like the gaps in Sappho’s fragments, the absences in Code Names are intriguing for what they cannot tell us. Any study of government secrecy is bound to be incomplete; indeed, the incompleteness is part of the point. The American public has become so inured at this point to revelations of ill-considered decisions up and down the chain of command, and the White House and the Pentagon so skillful at downplaying or dismissing these revelations, that whether Code Names will be effective as a provocation is unclear. But while the rendition flights continue for the time being, Arkin’s book enables us to follow the vapor trail: when we hear that the plane has refueled or taken on detainees at some foreign airstrip under cover of night, we can learn about the secret agreements that authorized the landing, and guess at where the jet is coming from, and where it might be headed.

This Issue

May 26, 2005