Faculty research projects take on new meaning after Sept. 11
Jennifer Greenstein Altmann & Steven Schultz
Princeton NJ -- From an effort to build gadgets that detect biological weapons to an in-depth series of interviews with former guerrillas, many research projects at Princeton have taken on new significance since Sept. 11.
Several University offices, including the Office of Government Affairs in Washington, D.C., have fielded questions about research that could be useful in preventing or responding to terrorist attacks. As a result, the Office of Research and Project Administration asked faculty members to identify such projects (see related article).
Here, the Weekly Bulletin focuses on four scholars who responded, each with a very different contribution to a complex and ongoing issue.
Understanding guerrilla movements
Politics professor Paul Sigmund sat in an apartment on the outskirts of Buenos Aires in 1995, face to face with the guerrilla who had assassinated a former president of Argentina, Pedro Aramburu, 25 years earlier. He asked the man, "How could you have done that?"
"He said, 'Everybody in the Peronist movement hated Aramburu because he overthrew Peron, and every year on the anniversary of Peron's overthrow we would have a meeting, and we would all shout: "Death to Aramburu! Death to Aramburu!" So I just carried out their wishes.'"
During the last seven years, Sigmund has interviewed 40 men and women from 10 countries in Latin America who were once major players in guerrilla movements. By gathering these first-person accounts, Sigmund hopes to better understand why people join guerrilla movements and how, years later, they look back on the paths they chose to try to achieve political change.
He hopes to draw some insights that can be used to help understand what drew young men to join al Qaeda. "It's a very different political and religious culture, but I think some lessons may be drawn from looking at how (Latin American guerrillas) got radicalized," he said.
Sigmund is especially interested in former guerrillas who have gone on to become politicians. Of the 40 he has interviewed, about half are or were cabinet ministers or members of parliament. In El Salvador, for example, the former guerrilla group FMLN is now the second largest political party.
"I was talking to Lydia Diaz, a well-known congresswoman in El Salvador who was wounded fighting for the guerrillas," Sigmund recalled. "And she said to me, 'You know, I used to carry a submachine gun. Now I'm carrying a cell phone.'"
In trying to understand how these former guerrillas, who were mostly students in their 20s when they joined the movements, were able to transform themselves into killers, Sigmund found the driving force was usually religion or its functional equivalent. In Latin America that often meant they were drawn to the fight by Marxism or liberation theology.
Finding the answer to why young people are willing to risk their lives for a cause is still, to some degree, relatively simple: "It's part of being young and idealistic and naïve," Sigmund said.
But now that these former guerrillas are in their 40s and 50s, they are, for the most part, critical of what they did. Some of the former guerrillas said that it was the violence employed by the military regimes in many Latin American countries that led them to embrace violence. In addition, the example of Fidel Castro in Cuba persuaded them that armed revolution was the way to achieve social change.
"They look back and say, 'Well, of course, I didn't understand economics. I didn't understand politics. I didn't understand that democracy, however imperfect, is a lot better than dictatorship,'" Sigmund said. "They think the objectives were noble, but the means were wrong. One of them said, 'I'm glad we lost.'"
Seeking new surveillance technology
To many Americans, it is an instantly recognizable image: Norm, Sam, Diane, Cliff, Carla and Coach -- all sitting in the familiar Boston bar. To a computer, however, it is pixels on a page.
Schwartz is leading a project called QUIPP, quick image processing procedures, which is developing devices and programming procedures that automatically and rapidly distinguish people from a background of other objects. The technology could have many uses in surveillance.
"In surveillance at an airport, for instance, you might have some cameras in an area where people are not supposed to be," said Schwartz, professor of electrical engineering. "If all of a sudden you have a face in the scene, you know there's a problem. Once you know there's a face, you can go to a database and say, 'Oh, that's one of our workers' or 'That person is not in our library of faces.'"
The problem is especially hard considering that each person is different and may appear in the image at many sizes and angles. Schwartz's method allows the computer to elimi nate very quickly the great majority of the image and zero in on areas most likely to contain a human face.
In the "Cheers" image, which appears in the doctoral thesis of 1998 Princeton graduate Richard Bang, a computer looked at 38,709 square templates within the picture and selected 886 of them for further analysis. Ultimately, it found all six faces. It did not fare quite so well with the cast of "Dynasty," in which it missed one face that was tilted. The examples show both the power of the technique and the difficulties that remain, said Schwartz.
His research grows out of 30 years of work in an area called detection and estimation theory. Funded in part by the Office of Naval Research, Schwartz looked for ways to discern signals and patterns from a background of noise -- for example, how to identify Soviet submarines from hundreds of miles away by distinguishing the sound they make from the rest of the noise of the ocean.
In all such problems, a key approach is to compare the noisy data with an example of the ideal data. If the desired signal or something like it is buried within the noise, then the comparison should show at least some degree of match, whereas random noise would not.
A sabbatical in Switzerland several years ago introduced Schwartz to the field of image processing, which involves many of the same problems and techniques. He now holds a patent jointly with the Seimans Corp. for a method of performing rapid face recognition in video images.
Schwartz believes the work could have many applications beyond recognizing faces. A robot, for example, could move through an industrial plant looking for areas in need of repair. Or a processor aboard a military surveillance plane could instantly identify a mobile missile launcher. A longer-term goal would be to apply the techniques to medicine and try to detect cancer cells from a large amount of tissue.
"It's the same kind of question: How do you eliminate uninteresting data, data you don't want to waste your time on?" he said.
Schwartz also acknowledges that his work raises fears about the misuse of surveillance.
"Yes there's a problem," he said. "There is a certain element of big brother. But I tell you if I worried about this, I would not use EZpass (the auto-mated toll-collecting system). But since I hate waiting on line at a tollbooth, I use it. From EZpass people can easily tell where I've been. It's a trade-off I'm making."
Exploring cooperation and conflict
Gil Rozman had specialized in the study of China, Japan and Russia for more than 30 years when he realized that "the hole in the doughnut," so to speak, was Korea.
"I had worked on all sides of Korea," said Rozman, the Musgrave Professor of Sociology, "and I realized I should be studying Korea too."
So two years ago, Rozman began private tutoring with the director of Princeton's Korean language program, Gwee-sook Kim, to start learning Korean. He also arranged to spend a semester studying Korean intensively at Yonsei University in Seoul. Now he has broadened his field of interest to explore the issues of regional cooperation and the potential for regional conflict among China, Japan, Russia and North and South Korea.
North Korea was thrust into the current war on terrorism when President Bush named the country as one of three nations that he considers, along with Iran and Iraq, part of an "axis of evil."
"That term was not welcomed by the other countries with which we need to coordinate policy," Rozman said. "We don't have the power in this region to decide what's going to happen to North Korea on our own. We need to work with the other countries, and each has its own agenda. Each has a different notion of the problem."
The threat that North Korea poses to the region and the world may surface in the next year or two, Rozman said.
"It's possible that North Korea may test a missile, or export weapons in a way that the U.S. considers provocative, and there will be a U.S. warning and maybe even some kind of showdown," he said.
Rozman's current research entails monitoring each country's reaction to events and the direction of the debates within each country on regional issues. He is co-editing a volume of articles that grew out of a December conference in Washington, D.C., organized by Korea University, which focused on the U.S. war against terrorism and North Korea's weapons of mass destruction.
Rozman is especially interested in studying Russia's posture toward the Koreas. He also is keeping an eye on how Japan regards U.S. actions in the region.
"I'm interested in the degree to which the U.S. balances unilateralism and multilateralism," he said. "Are we able to build a coalition and to reach out to others who are skeptical? How do we engage them in an effort to change North Korea? I think that's a real challenge for an administration that has very strong unilateralist tendencies."
To improve his conversational skills in Korean, Rozman is sitting in on a first-year Korean class a couple of times a week. That will help him with his research on South Korea; he would like to spend some time interviewing South Korean specialists on domestic and foreign relations and to participate in conferences where Korean is spoken. But deciphering what is happening inside North Korea is extremely difficult because the media is heavily censored and travel in the country is restricted.
"I don't know when there will be an opportunity to do research in North Korea, and I don't know when my Korean will be good enough to be serious about the kind of work I'd like to do there," Rozman said. "But this issue is not going away."
Developing bacteria-testing devices
One such device has demonstrated the ability to detect such bacteria as salmonella and E. coli. It has even distinguished a harmless form of cholera from the virulent kind. The detector is contained entirely on a device similar to a computer chip. It works by pumping a microscopic sample of water into the path of a micro-wave beam and reading how the sample absorbs or reflects the microwaves.
The chief benefit of Sohn's device is that it requires no special preparation of the sample and no additional chemicals. It works instantly and can distinguish between many different substances in a single test.
Current tests tend to be slow and involve costly materials. Some, for example, require growing the bacteria in a laboratory, or exposing the sample to carefully prepared chemicals that respond to only one kind of bacteria.
"A soldier could take one of our chips out in the field and do a quick diagnostic," she said. "You avoid all the preparation you might need to give the soldier or commander enough information to say, 'Is it bad? Is it good?'" The quick test would be followed by more thorough assessments.
Another benefit is that the device can test a miniscule amount of liquid. In her lab, Sohn supplies the detector with samples that are about 1/100th the volume of a single drop of water from an eyedropper.
Other devices Sohn is developing can identify a particle by measuring its electrical capacitance -- essentially its ability to hold a charge.
Apart from their usefulness against terrorism or biological warfare, the devices may also be valuable in medicine. The detector might measure very rapidly a patient's white blood cell count, and even distinguish between the five different kinds of cells. Sohn has been developing such applications in collaboration with Dan Notterman of the molecular biology department.
Sohn's interest in biological applications of physics developed only in the last four years. Her original expertise was in the area of nanotechnology -- the creation and manipulation of very small devices. She realized that the techniques she used with nanoscale electronics might work just as well, if not better, with biological materials.
Now with considerable expertise in the field, Sohn recently was appointed to the New Jersey Domestic Security Task Force, which was established after Sept. 11 to examine whether threats such as anthrax could be handled better in terms of prevention, diagnosis and treatment.
"I think it's a very important issue and I'm really
honored to be asked to be part of it," she said. "This is a
totally new experience to me. It is nice to be able to be of
service to society, and it gives me a new perspective on my
research as well."
Editor: Ruth Stevens