There’s a desperate shortage of foreign language speakers at our intelligence agencies. Not that they’re doing anything about it.
December 3, 2001, Vol. 7, No. 12
A RUMOR HAS BEEN CIRCULATING in intelligence circles that communications intercepted prior to September 11 referred in Arabic to a “Christmas gift” for the United States. What no one listening to these messages realized was that the same expression can mean “an unpleasant exploding surprise.”
This anecdote may or may not be true. But the lack of trained linguists in our intelligence services is no rumor. Directly after the September 11 attack, FBI Director Robert Mueller issued an urgent appeal for Arabic and Farsi translators, posting a toll-free number for applicants on the FBI’s website. But this is too little, too late: A critical shortage of linguists with security clearances has crippled American intelligence efforts for decades, and will take decades to remedy fully.
One intelligence failure after another has been linked to the lack of translators and interpreters in the U.S. intelligence community. Following the 1990 murder of Rabbi Meir Kahane in Manhattan, the FBI confiscated handwritten materials in Arabic from the assassin’s apartment. No one translated them. The FBI also seized Arabic videotapes and bomb-making manuals from Ahmad Ajaj, a Palestinian serving time in federal prison for passport fraud. No one translated them. Prison officials made tapes of Ajaj as he described bomb-making techniques over the phone. No one translated them. After the 1993 World Trade Center bombing, all of these materials were at last reviewed. They pointed clearly to the impending attack.
An inability to translate evidence impeded the investigation of the bombing of the American embassies in Kenya and Tanzania in 1998. The lack of translators hampered the investigation of the October 1999 downing off Nantucket of EgyptAir Flight 990. Policymakers were not warned of impending nuclear detonations in India and Pakistan, intelligence sources say, not because the evidence was unavailable, but because analysts could not understand it. According to a recent House Intelligence Committee study, countless data are never analyzed by the NSA and CIA because too few analysts possess language skills: “Written materials can sit for months, and sometimes years, before a linguist with proper security clearances and skills can begin a translation,” the authors note. A mountain of similar testimony has been presented before the House and Senate intelligence oversight committees in the past decades; nothing has been done.
The Armed Forces Medical Intelligence Center in Fort Detrick, Maryland, has no cleared linguists on its staff: The center is charged with tracking foreign medical capabilities, infectious diseases, and biomedical subjects of military importance. Journalists have been told that the government suspects domestic extremists of mailing anthrax to members of Congress and American news organizations. This is reminiscent of the joke in which a man looks for his missing wallet underneath a street lamp, because that’s where the light is.
CIA sources with knowledge of the agency’s current language capabilities say that there are perhaps four or five truly competent Arabic speakers in the entire Central Intelligence Agency. There is, according to one recently retired CIA official, only one speaker of Farsi at the agency with an intimate knowledge of the language. The Farsi expert enjoys a “fluent and melodious” command of the language; he is a connoisseur of Persian lyric poetry. Unfortunately, he is also “about 9,000 years old.”
True, there are analysts in the CIA’s Directorate of Intelligence who read some Arabic, but the language they read is classical Arabic, not colloquial, and they can speak neither. Asked to confirm this assessment, another officer familiar with the agency’s language capabilities snorts: “That’s generous. Most of the analysts don’t know squat about Arabic.” Says another senior agency official: “There’s probably not a single analyst in the DI who’s totally proficient in modern, colloquial, spoken Arabic. There’s one guy who reads Uzbek, but he doesn’t speak it.”
Case officers who study Arabic in the United States are often sent for a single overseas tour (usually two to three years), then rotated elsewhere, where their language abilities atrophy. “There were no officers in Germany in the 1990s handling the Middle Eastern terrorism problem,” says an official. Why not? “They couldn’t speak Arabic.”
The intelligence community has almost no knowledge of the languages spoken in or around Afghanistan–Pashto, Farsi, Dari, Tajik, Azgari, Uzbek, Turkmen, Aimaq, Baluchi, Ossete, and Yaghnobi. Dari and Pashto were taught at the CIA in the 1980s, but the people who speak them are now retired. Until recently, the Defense Language Institute in Monterey did not teach the Dari variant of Farsi, the primary language of Afghanistan. U.S. Customs employs one Pashto speaker. The INS has almost no Arabic speakers. For now, intelligence officers and military personnel in Afghanistan are relying for translation on the Pakistani intelligence service–which created, supplied, and funded the Taliban.
The CIA’s open-source translation arm, the Foreign Broadcast Information Service, was gutted following the fall of the Berlin Wall. FBIS used to translate a wealth of otherwise unobtainable primary documentation. Now it limits its efforts to wire service copy and articles from foreign news websites. Its translators are overtaxed: Last November, puzzled analysts endeavored to limn a translation from a Palestinian newspaper in which the authors appeared to charge Israeli defense officials with the use of “phlebotomized uranium.” After some consideration, readers realized that the uranium in question was, in fact, depleted rather than skillfully drained of its blood.
According to a senior CIA official, the agency, fearing embarrassment, exaggerates the linguistic strengths of its officers to Congress. “They list a bunch of people who they say speak Urdu, but they’re talking about people who haven’t used the language since 1985. So they say, ‘we have 18 speakers of Urdu,’ but none of these people are in any way capable of working on the street.”
Does this skew intelligence collection efforts? “Absolutely. It’s terrible,” says the official. “It’s a huge error. It has an enormous effect–if you can’t speak the language, it’s easy to deal with liaison” instead. Those foreign intelligence officers “probably speak English. But you’re cut off from more than 90 percent of society. You can’t spend time on the streets, in the marketplace. You’re a prisoner in the embassy. So you spend your time with the elites, with westernized business people who are in denial about what’s really going on in their countries. And you have no idea what’s really going on.”
WHAT CAN BE DONE? There are only two solutions: Take intelligence officers and teach them languages, or take linguists and teach them to be intelligence officers. Unfortunately, the former is virtually impossible, and the intelligence community, still fighting the Cold War, is stubbornly unwilling to do the latter.
Consider the average newly hired intelligence officer, a monolingual in his late twenties, from, say, Alabama. How long does it take to train him to speak Arabic to the level of proficiency required to conduct intelligence work in the Middle East? Arabic poses unusual problems for the American student. It is diglossic: The Classical Arabic of books, newspapers, formal speeches, and broadcasting, with its Koranic derivation, varies greatly from the colloquial Arabic spoken in homes and in the street. Each Arab country has its own dialect, and many are mutually incomprehensible: A student of Egyptian Arabic has little hope of understanding Moroccan Arabic. The phonology and morphology of Modern Standard Arabic are completely different from those of Germanic and Romance languages; sounds are formed with parts of the palate unused by Americans. The monolingual from Alabama may be able to order tea and see to it that his shirts are starched after ten months of language instruction, but it is the rare American who, after ten months of study, can understand a fuzzy exchange between two mumbling native speakers discussing potentiometers, circuits, and the limitations of various explosive devices at high altitudes.
Efforts to teach Americans to speak difficult languages, either in universities or government institutions, have generally failed. The National Security and International Affairs Division of the General Accounting Office found that one-third of the graduates from the Defense Language Institute, the nation’s finest language training center, had not attained the minimum proficiency of level two, on a scale of one to five, with five being the highest. American universities are of no help either. Beyond the two years of classes usually required for college entry or graduation, American students are not obliged to study languages. Few Americans undertake the serious study of a language at the appropriate age (the younger the better), fewer still study the rare or difficult tongues. According to the most recent government figures, only 4,800 American undergraduates are now studying Arabic. Of these, perhaps 10 percent will attain proficiency. Only 600 are studying Farsi. Fewer than 500 are studying Urdu. Fewer than 10 are studying Uzbek, Turkmen, and Tajik. None of the minor languages of Afghanistan and its environs are studied in the U.S. educational system at all.
During the 1960s and 1970s, the Department of Defense attempted large-scale contracting of language instruction to civilian universities, including Syracuse, Cornell, Indiana, and Yale, as well as several commercial language schools. In all cases the programs were judged failures; graduates could not satisfy DoD requirements. These programs failed for the simple reason that to learn a language requires total immersion over a period of five to ten years; the effort must include living abroad amid the people who speak and use the language. To achieve true proficiency in another language, the mother tongue must, in effect, be abandoned or subjugated to the adoptive tongue. No American university now trains students to the high professional level needed for intelligence work.
Surely, then, the United States must take native speakers of foreign languages, and train them to be intelligence officials? After all, the United States is nothing if not rich in immigrant resources, and many of these immigrants are devoted patriots–all the more so for having recently arrived in America from one or another repressive, barbarous hellhole. Estimates vary, but it is safe to say that the number of U.S. citizens from Arab nations is more than one and a half million, and the number of highly competent Arabic speakers is more than a million. There are also more than a million Iranian immigrants in the United States, several hundred thousand of whom doubtless speak excellent Farsi. Estimates of the size of the Afghan population in the United States vary from 41,000 to 180,000. Whatever the precise number, their ranks are filled with speakers of Dari, Pashto, and other useful languages.
Many of these immigrants are painfully underemployed, as anyone who comes into contact with them can attest. In a cab from Union Station a few days ago, my driver was a fully trained aeronautical engineer from Lahore and a fluent speaker of Urdu. He was desperate to find work in his field or, failing that, any work more stimulating than driving a cab. He pressed his card into my hand and begged me to help him find a better job. It would seem that the simple solution for him would be to call the FBI hotline. In fact, with his heavy accent and a personal history of some kind of unpleasantness with the Musharraf government, he would be wasting his time. And this is sadly true for the thousands of foreign-born Americans who could supply the linguistic expertise their new country lacks.
Simply put, our government does not trust native speakers of foreign languages and makes it nearly impossible for such volunteers to obtain security clearances. Prospective employees of the CIA, for example, are required to list the names and addresses of every foreigner with whom they have a close or continuing relationship. Someone who speaks Dari with native fluency almost certainly will have relatives and friends in Afghanistan, and will probably be acquainted with Islamic fundamentalists, former Communists, and other miscreants. If he knows many of them, it is very unlikely that he will receive a security clearance. If he knows only a few of them, he is probably not from Afghanistan. A qualified candidate’s application can languish for years; he will be given no information about its status; often, he will be interrogated abusively by polygraphers who have never themselves left American soil and who suspect that a knowledge of Dari is evidence, prima facie, of untrustworthiness. The government’s pay rates for contract workers, moreover, run well below the hourly rates offered by private businesses. The rational speaker of an exotic language will give up. He will take a job with Shell Oil.
One CIA official laments that the agency’s suspicion of native speakers of foreign languages runs so deep that the organization is reluctant to hire them even as instructors. “Some of the people teaching [a required language] were pathetic, because there was no way good teachers could get a clearance….They’d been in the U.S. so long that they’d actually forgotten [their native language].” Even native-born Americans will have trouble obtaining security clearances if they have done the one thing most necessary to achieve language proficiency–lived and studied overseas. “The more time a candidate for employment has spent abroad, and the more foreigners he knows, the less likely he is to receive a security clearance,” says a CIA officer.
In truth, says a retired senior CIA officer, “the CIA doesn’t want language speakers. They want to work with people who resemble themselves. . . . Anyone who immerses himself in a foreign culture is suspect….These people [CIA officers] live in malls. You put them in a foreign city and they take taxis to the embassy, not the metro. They never meet foreigners.” Another officer concurs: “There’s no interest in learning languages in the CIA. No matter what they say, learning a language won’t get you promoted. There’s no interest in learning Dari–the prevailing attitude is that such languages, including Arabic and Farsi, are what ragheads speak. They’re still thinking in terms of the Cold War–going on the dip circuit and speaking to people who speak English.” Promotion panels within the CIA’s Directorate of Operations reward case officers who recruit assets (however dubious the assets’ value), not those who learn languages. “There’s actually a disincentive to learning languages, because if you take two years to study a language, you’re out of the running for assignments where you might get promoted. People would rather go to, say, Mexico City than spend two years in class,” says one former officer.
Although the CIA puts prospective officers through an extensive battery of physical, psychological, verbal, and mathematical tests, not one of these tests measures the candidate’s natural ability to learn a foreign language. It is not an aptitude the hiring process takes into account at all.
IF THE UNITED STATES is serious about preventing the next terrorist atrocity, the language crisis must be redressed. In the short term, intelligence agencies will have to be forced to overcome their scruples about hiring and clearing foreign-born linguists. Certainly, the induction of a large cadre of staff agents with connections to hostile countries poses a security threat. But there is no choice: The threat posed by having no speakers of foreign languages in the intelligence community is vastly greater. The congressional oversight committees should impose a one-month deadline for processing clearances. The CIA will howl about this, but that’s just too bad: Its leaders have presided over the most catastrophic intelligence failure in the history of the American polity, and have thereby proved themselves incapable of making these difficult decisions themselves.
Language specialists do not need Q clearances; they need never set foot in CIA or NSA headquarters. They can be placed in a separate facility in the Virginia suburbs, and they can be flown where needed around the world. Information to which they are exposed can be strictly compartmented. Ideally, the recruits will be women: Linguistic research suggests that women learn languages faster, both as children and adults, and are more able translators. Moreover, women are far less likely to be sleeper agents for Islamic radical organizations. The government should pay these officers what they would be able to earn in the private sector. Funding for FBIS should be restored. And the heads of the various intelligence agencies should be put on notice that if so much as one more American death devolves from a failure to understand confiscated materials and intercepts, their careers will be over.
In the long term, the structure of foreign language education in the United States must be imagined anew. The educational system of the world’s only superpower is organized on essentially isolationist principles. Students of the only country capable of enforcing a civilized order in the world cannot remain profoundly ignorant of the languages, politics, and religions of that larger world. There is an extraordinary disjunct in academia between professions of multiculturalism–exhortations to celebrate diversity–and any kind of serious commitment to learning about other cultures. It’s one thing to tell students that Muslims are wonderful peaceful people; it’s something else entirely to teach students Arabic so they can read what is being said about their country in the wonderful peaceful newspapers of the Middle East. But anything less than the latter is insipid, meaningless pablum, and dangerous to boot.
If we are serious about training competent linguists, education in foreign languages should begin before the age of eight. Language instruction should be mandatory from elementary school, and college scholarships from the Defense Department should be given to students who major in obscure languages of value for defense. Far more American students should study abroad–not in France or Italy, but in Tunisia and Pakistan–and they should do so not for a semester, but for several years.
In the 19th century, British soldiers and administrators studied classical Pashto as a matter of routine. The United States is now the leading world power, but its efforts to understand what the rest of the world is saying have by comparison been purely desultory. Now we are paying the price.