Abstract
By Dan Stober and Ian Hoffman Simon & Schuster, 2001 384 pages; $26.00
By Wen Ho Lee with Helen Zia Hyperion, 2001 332 pages; $23.95
A little more than three years ago, Wen Ho Lee's name was first reported in the press in connection with the alleged theft by China of U.S. nuclear weapons secrets from Los Alamos National Laboratory. For much of the next year and a half, Lee, a diminutive code developer and naturalized citizen specializing in hydrodynamics who worked in the lab's X Division, became synonymous with nuclear espionage and lax nuclear security. Arrested and indicted in December 1999 on 59 felony counts of mishandling national security information “with the intent to injure the United States, and with the intent to secure an advantage to a foreign nation,” Lee served 278 days in solitary confinement. Then, the government agreed to release him in exchange for a guilty plea to just one count and a promise to cooperate in an investigation to clear up the many unanswered questions about his actions.
Anyone interested in knowing how Lee became a suspect and why the case against him ultimately collapsed can do no better than to read A Convenient Spy, written by San Jose Mercury News reporter Dan Stober and Albuquerque Journal reporter Ian Hoffman. Stober and Hoffman, both of whom have for years covered the activities of the U.S. nuclear weapons laboratories, have created a lucid and compelling narrative about the case, weaving together many diverse strands to form a devastating picture of an espionage investigation gone horribly wrong. For that reason alone, it should be required reading at the FBI, the Department of Justice, the Energy Department, and in Congress.
Although Lee was initially suspected by the government of providing key design details of the W88 warhead to China, the case against him eventually centered on his unusual and unauthorized copying of weapons codes and associated files from the classified to the unclassified network at Los Alamos and from there to portable tape drives, a transgression discovered after he was fired from Los Alamos. (For background on the Lee case, see my earlier Bulletin articles “A Very Convenient Scandal,” May/June 1999, and “Scientist, Fisherman, Gardener, Spy?” November/December 2000.)
Much has been written about the codes, but until A Convenient Spy, no one has described so clearly how they work. To predict the performance of a nuclear bomb, the codes break down the explosive process, using a grid of more manageable “cells,” with the entire grid known as a “mesh.” As the authors explain, “Each cell was represented by dense mathematical equations that calculated the pressure, temperature, density, and movement of the material within the cell. The answers formed a snapshot of the bomb at one instant in time. Then the code would take a ‘time step’ of a small fraction of a second and begin calculating the next snapshot. A full run of a weapons code on a Cray supercomputer could last overnight or even days.” Lee and his colleagues worked with two types of codes. “In Eulerian codes, the imaginary mesh was stationary. As the bomb exploded, the material in the bomb—plutonium, high explosives, neutrons—moved from one cell to the next. Lagrangian codes were different. The cells moved as the material moved, so that a single cell would track the same bit of material throughout the explosion.”
As Stober and Hoffman write, “Writing code is a cerebral, not a physical, activity. Lee … was often alone in his office, quietly employing numbers to describe the most violent mechanisms that humans have ever invented.”
The authors also do an excellent job describing the special characteristics of the W88 warhead, the most advanced in the U.S. arsenal—its football-shaped “primary” or fission bomb, code-named “Komodo,” its spherical “secondary” or fusion bomb, code-named “Cursa,” and the 3-foot-long, peanut-shaped depleted uranium radiation case enclosing both components, with a diameter of about 9 inches at its narrow end, where the primary was placed. The appearance of some of this information, along with the design details of other U.S. nuclear warheads, in a 1988 Chinese document obtained by the United States, triggered the investigation that eventually led to Lee.
How did Lee become a suspect? When the Chinese document surfaced in 1995, Notra Trulock, head of intelligence and counterintelli-gence for the Energy Department, assumed China had stolen the W88 information from an Energy Department facility. Early on, Trulock, who was known by his colleagues to have designs on becoming director of the CIA, was overhead saying, “We need one good espionage case to make this program grow. There's one spy out there and we're going to find him.” (Trulock has denied saying this.)
Trulock later signed off on an investigation that focused on ethnic Chinese U.S. citizens involved in the design of nuclear weapons who had visited China, based on the assumption that, for espionage purposes, China would target ethnic Chinese only. Before the investigation had been completed, Wen Ho and Sylvia Lee (who also worked at Los Alamos) were named as suspects.
When Dan Bruno, Trulock's principal investigator, set out to find the spy, he failed to examine closely the Chinese document to determine what, exactly, had been compromised, or where the information might be obtained in the United States. (Officials at Los Alamos later determined, after Lee was fired, that a key document containing the W88 data in question had been sent to 548 addresses inside the Energy and Defense Departments and the military contractor community. They also traced erroneous information on one of the other warheads, the W87, to a publication by Jane's, a purveyor of magazines and reference books on military technology.)
Compounding matters, Bruno refused to look for clues at the Defense Department or any of the contractors who manufactured the W88 reentry vehicle, arguing, “That's outside my sandbox.”
The FBI, which eventually took over the case, never objected. Bruno also declined to explore the possibility that the source was inside Energy headquarters, out of concern that such an investigation could not be kept secret. He failed to visit either the Rocky Flats Plant (where the plutonium components for the W88 were manufactured) or the Pantex Plant (where all U.S. nuclear weapons are assembled) because no one at either site had any contact with China.
That left the three design laboratories. Although Sandia National Laboratories maintained all of the design documents for the W88, it was not required to keep records of foreign travel and so was eliminated from the investigation. Because Bruno believed the entire design of the W88 had been compromised, he assumed that the spy would have had access to the warhead's blueprints. Upon discovering that Lawrence Livermore National Laboratory did not receive the blueprints until 1990, at least two years after China apparently obtained the data, he chose to ignore Livermore, too, and concentrate exclusively on Los Alamos, where the W88 was designed. And so it went, with each questionable assumption, misstep, and oversight leading investigators further and further away from solving the case.
A different and very personal perspective on this affair can be found in My Country Versus Me, written by Lee with Helen Zia. The book is at its best in describing Lee's reaction to the maelstrom in which he found himself in the spring of 1999. Although there is little here that is newsworthy, it is more than appropriate that Lee have the opportunity to rebut the accusations against him and explain his actions. The prose is simple and direct, although I wish that Lee and his journalist co-author, in reaching to connect with their readers, had avoided describing events in terms that Lee, on his own, would be very unlikely to use, such as this description of the FBI's repeated interrogations of him: “The whole thing was turning into a Keystone Kops slapstick routine. The joke was on me, and I didn't even know it.”
Lee's anger and dismay about the government's leaks to the media about him and his wife, and about the FBI's handling of the case, is palpable. So is his disgust with the members of Congress who played fast and loose with the facts. Lee's descriptions of his 278 days in solitary confinement—the total lack of privacy, insufficient and inadequate food, the requirement (mandated by the Justice Department) that he wear leg and wrist shackles whenever outside his small cell—are compelling.
Troubling too are his claims that jail officials misled him or made things especially hard for him in order to force him to confess.
Upon learning that another inmate has committed suicide, Lee wonders whether the government wants him to suffer the same fate. He resolves to persevere in the knowledge that “I would have to get out sometime, because the indictment was all lies, exaggerations, and fabrications.”
The Lee portrayed in My Country Versus Me is straightforward and hardworking. Lee traces his interest in mathematics to his childhood in Taiwan. When he was growing up, “getting involved in politics was so dangerous [that] I stuck to science and math, which are much more fascinating to me and have a clear benefit to humanity.” Back when Taiwan was ruled by a military government, “Subjects like history or geography were boring, full of propaganda. Because of that I didn't much care for them.” As an adult, Lee had only a few passions: his work, his family, classical music, gardening, and fishing. Lee was so apolitical that even after he became a citizen in 1974, he never chose to exercise his right to vote.
He also provides some insight on why someone would seek work at a place like Los Alamos. Lee says he wanted to work at a government laboratory because “it was well known among scientists that the national labs had better computers than private industry.” Of his decision to switch from working on unclassified projects like nuclear reactor cooling systems to nuclear weapons, Lee writes matter-of-factly, “Most of the work at Los Alamos has to do with nuclear weapons, and those jobs are more steady. In the Cold War days, the nuclear weapons projects were well funded, whereas in Q Division it was always a struggle from budget to budget. I figured my future would be more secure if I worked in the weapons area.” That he would be working to help design weapons capable of killing millions of people was apparently of no concern.
Additional insight into Lee's thinking on these matters is found in A Convenient Spy. His problem keeping jobs (he was fired from his first and forced out of his second), his complicated relationship with his wife (she also copied classified information to Los Alamos's unsecured network and once, out of anger at her supervisor, deleted some of the original files), his pack-rat tendencies (not uncommon for code developers), and his habit, beginning in 1980, of providing sensitive but unclassified information to unauthorized persons in Taiwan, all help round out and, to a certain extent, explain his actions.
Each book provides its own answers to the central question in the case: Why did Lee copy and down load onto 17 portable tapes both unclassified and highly classified weapons codes and data? Lee says he began downloading data in 1993 to protect the files from loss should Los Alamos change its computer operating system (a change in operating systems in the late 1980s caused him to lose some important files).
“Wow, 400 words per minute! Is that typing or shredding?”
He also insists that as part of his job, he needed his version of the codes to compare against future changes made by other developers (changes that could affect the functioning of his part of the code). He did all this himself, in violation of laboratory regulations, rather than seek the assistance of the lab's computer support staff because: “I didn't want to spend time explaining which parts of which files I wanted to save or how I wanted them organized, nor did I want to wait for the weeks or months it would take them to make my backup tapes. I wanted my files kept in a certain way—just as I kept my office, my music, my tools, and my kitchen organized just so.”
Lee acknowledges this was a mistake for which he deserved to be disciplined, though not one significant enough to justify being fired, indicted, and jailed. Lee also explains that the codes are of no use to anyone without first being compiled by programs kept only in the computer support division. The users' manuals were also essential, and Lee did not download them (although he kept paper copies in a safe in his office).
Stober and Hoffman do not find this convincing. Lee backed up not just “his” codes but also many other codes unrelated to his work in hydrodynamics. Los Alamos also allowed scientists to easily make backup copies on the secure computer network from their own computers. Lastly, they note that Lee lied to his lawyers, his friends, and even his own family about his activities. He claimed he began making the tapes in the late 1980s after losing important files when the lab changed operating systems. But his first file transfers occurred in April 1988, eight months before changes in the operating system created temporary problems for many scientists in X Division.
If Lee downloaded the codes to ensure his continued access to them should he become a consultant or a teacher focusing on the unclassified aspects of hydrodynamics, as some have suggested, why did he also download the highly classified data files and input decks containing results from U.S. nuclear weapons tests as well as the precise specifications of U.S. warhead components? Furthermore, Lee had access to numerous unclassified codes and data files. In fact, he used those files to copy over the weapons codes on his tapes shortly before he was fired.
If Lee were assisting China, why would he (or his handlers) run the risk of leaving his large and clearly labeled files in plain sight on the unclassified Los Alamos network for as long as 11 years? For that matter, why would he take 70 days over a period of nine years to amass his collection? And why wouldn't he have sought to transfer archived nuclear test data and actual bomb blueprints, a more valuable (and easier to use) commodity than the codes, which only approximate nature and often contain “fudge factors” inserted by developers like Lee to make the results match actual testing data? There is also no evidence that Lee provided his archive to anyone in China or Taiwan (or that anyone in Taiwan or China provided him anything in return).
Ultimately, Stober and Hoffman conclude that “the most reasonable explanation for Lee's collection lies in the gap between his scientific skills and his craving for job stability and professional recognition.” A hard-working but not outstanding scientist, Lee struggled to overcome his poor language skills and throughout the 1990s feared losing his job due to budget cuts. Perhaps his data collection was intended to provide him with a greater sense of job security, something he could fall back on after Los Alamos. If so, then ironically Lee's actions helped create the very outcome he worked so hard to avoid. •