First appeared in print in The Rhinoceros Times, Greensboro, NC
Poindexter's Ounce-of-Prevention Project
It's so frustrating watching Defense Department press conferences on C-SPAN.
The reporters are woefully unprepared. Very few of them show any understanding of what the Department of Defense does. They ask questions that are clearly designed, not to find anything out, but to get the DOD spokesperson to "admit" something that can be trumpeted by the media to show the awful things that the military is doing.
Often the reporters' "questions" become downright accusatory.
For instance, at a DOD press conference on Wednesday, 20 November, a spokesman tried to answer questions about Admiral John Poindexter's DARPA research program.
It was obvious that the press already knew exactly the story they intended to tell. All the questions were accusatory. "How is this program within the scope of the Department of Defense?" "Doesn't this amount to spying on American citizens?"
Even after these questions were answered, the reporters, too uneducated to understand the answers, simply asked the questions again.
There was even the offensive and outrageous question, "Since John Poindexter was convicted of lying to Congress -- though his conviction was later overturned on appeal -- why should this program be under the leadership of a man like that?"
This from the same media establishment that thought that Perjuring Bill was a dandy President, and Hillary could be a legitimate senator despite all her lies, obstructions, defiance of congressional subpoenas, and cattle-futures crimes.
Never mind that Poindexter's conviction was overturned, which makes him not guilty after all.
Never mind that all the illegitimate, politically motivated, nasty, and meaningless things that Kenneth Starr was falsely accused of, Iran-Contra investigator Lawrence Walsh actually did, which should fatally taint all the accusations he made -- including those against Admiral Poindexter.
I guess the media's rule is, when Democrats like the Clintons are accused of things that they clearly did, it can never be brought up again in polite company; but when officials in a Republican administration are cleared of politically-motivated charges, those charges stain them forever.
And Tom Daschle dares to complain that America's worst problem is Rush Limbaugh?
Let's get back to the important thing -- Poindexter's research project.
Let me offer what those reporters will never be able to offer -- because they're never going to try: A clear explanation of what Poindexter is doing.
Poindexter's team is going to conduct a study to see if there is a practical way to sort through the trillions of items of "transaction data" that is generated every year throughout the world, in order to find patterns of behavior that might reveal, in advance, a pending terrorist event.
Remember that what made 9/11 -- or the Oklahoma City bombing -- possible was anonymity. Nobody noticed or cared that Timothy McVeigh rented a truck or bought fertilizer -- why shouldn't he? And even though somebody noticed some of the 9/11 mass murderers in flight school, nobody was tracking the fact that they all bought tickets to fly on the same day.
Yet every one of these transactions generated records that became part of some company's or agency's database. It was all there to be discovered -- after the fact.
But what if we could have spotted them before they acted -- in time to prevent their monstrous crimes?
The problem with prevention is twofold.
First, before the crime, they aren't guilty of anything.
Second, out of trillions of transactions all around the world in many languages and in many incompatible computer databases, how in the world is anybody supposed to spot these guys?
It's relatively easy after we know who they are to find their footprints. But how can we set up tripwires to bring them down before they act?
Here's a clue: It won't be a whole bunch of guys sitting in rooms in the Pentagon spying on Americans or even foreigners, listening in on international phone conversations or reading everybody's email.
To do that, it would take more spies than there are people on earth to be spied upon. In other words, to completely monitor all the transactions of six billion people, it would take about twelve billion full-time employees.
That's just a teensy bit more expensive than Poindexter's $10 million budget.
There is no way that Poindexter's research project will violate the privacy of a single American. And if his research shows that transaction-analysis is a potentially helpful tool in the war on terrorism, all the same legal safeguards that are already in place to protect the privacy of citizens will continue in force.
In fact, if you actually understand what transaction-analysis software does, you'll realize that it isn't "spying" at all.
For instance, when you buy a plane ticket, you reasonably expect that it's none of my business when and where you're going to fly.
But it's not as if you expect it to be a secret. The airline has to know. You're going to show your I.D. to get the ticket. The passenger standing at the counter next to you might overhear your destination. The clerks will know.
That doesn't mean the government has a right to track you and monitor your movements all the time. That would be creepy. Like having a newly-reelected Republican Sheriff try to organize a boycott of everybody who contributed, politically, to his opponent -- police-state stuff.
But Poindexter's project doesn't do that. They are conducting their research using dummy data -- made-up airplane reservations and passport applications and credit card purchase records. They'll test their ideas on fictional data. I think everybody will agree with me that the government has no obligation to protect the privacy of people who don't exist.
Even if the research is fantastically successful, the government still won't be tracking you. For instance, a computer system might be tracking Parcel #F8ABE4, but nobody will be able to find out that that person is you unless a judge makes the legal determination that there is sufficient pattern evidence to justify discovering the identity of Parcel #F8ABE4.
Even that is pie-in-the-sky stuff. Personally, I don't believe computers will ever be able, by themselves, to find such patterns, unless they're grossly obvious ones -- like a sudden jump in the number of Saudi nationals entering the United States. And that information the government can already get without a warrant.
The real value of this transaction-analysis software comes when law-enforcement or border-control or defense-intelligence agencies provide a "seed." That is, they already know that Scooby Doe is a known associate of known terrorists or has just entered the U.S. from a nation that sponsors terrorism.
These agencies get court permission to "track" the particular person. Now the transaction-analysis software is not "looking for patterns," it's tracking the transactions of a particular person. It tracks backward -- where has he been before? What has he bought? What phone numbers has he called? -- and it tracks forward, following his transactions through the U.S.
The software does what humans could never do. Every plane ticket that Scooby Doe buys, the software looks at every other ticket bought for the same place within three days. Then it looks to see if any of those other people are also on the list of "watched persons."
Since the only way you can get on that "watched list" is through court approval, this system is not picking out people to spy on. It's simply doing a far more thorough job of noticing connections between suspects than any number of human workers could possibly do.
But what if, by sheer coincidence, you -- Ms. Innocent, we'll call you -- happen to be on a series of business trips and end up flying to the same cities as a suspected terrorist on three different occasions?
It might trip an alarm. Someone may take notice of you and ask you questions -- enough to ascertain that it was just a coincidence.
But that sort of thing already happens every day in the process of police investigations. Innocent people have to be questioned when they are inadvertently in circumstances that make them potential suspects.
Ms. Innocent is going to be able to explain her travels without difficulty, and because she has not been training in terrorist camps in Libya or Syria or Iraq, and does not speak enough Arabic to find a restroom in Amman, the anti-terrorist investigators will happily cross her off their list -- as investigators already do whenever crimes are investigated.
I have actually seen a demonstration -- after the fact, alas -- of how even a "fuzzy-logic" transaction-analysis software program with access to only a few key databases could have found every single one of the 9/11 terrorists in advance.
The few of them that were known terrorism suspects, if they had been used as the "seed" for the system, would have led to all the others, by examining public-record databases, and would have noticed that they all bought tickets for the same day -- in time to stop them.
Could have. Would have. But didn't.
Why not? Because the different agencies that knew who the "seeds" were had no knowledge that such transaction analysis was even possible.
Why didn't they know? Because the people using that kind of software right now are businesses, who use it to try to keep known felons or corporate spies from getting jobs with their company.
That's right. It's being done already, all the time, by corporations that don't have to operate under the same constraints as government. So it's not as though the government would be finding out stuff that nobody knows.
On the contrary, the government would be limited to finding out way less about you than the phone company already knows.
If the system might help, why not give it a try, as long as the legal safeguards remain in place? Of course, that's always a big "if," but our government has a pretty good record of not becoming a police state, because we rein in those who forget the rules.
Transaction analysis wouldn't stop every act of terror -- nothing can guarantee that -- but it will make it far harder for possible terrorists to carry out their mass murders undetected.
Just imagine how far you'd get if you tried to do any serious travel without leaving any electronic footprints anywhere. Now imagine Arab-national foreigners trying to do the weird things one has to do in order to not use credit cards or show I.D. Get on an airplane? Rent a car? Buy a gun? Not impossible, but far more expensive and dangerous.
If properly used and properly supervised, transaction-analysis software with access to a lot of databases might be an important part of our national-defense toolkit -- without compromising our privacy one whit more than it is today.
And contrary to the suspicions of the Media (who are themselves the worst violators of privacy in the world, by profession) Admiral Poindexter is the opposite of what they accuse him of being.
He's a man who values privacy at least as much as you or anyone else, and who loves the Constitution and is sworn to defend it. I've heard him say, passionately, that if this can't be done without violating our constitutional protections, then it's not worth doing.
I think his research project is a valuable one. I hope it pans out. I hope an effective, safeguarded system can be deployed.
Here's a little thought-experiment to perform. Think of all the innocent passengers on those planes on 9/11. Imagine if you could ask them, now, whether they would have minded having the government have access to the database that contained a record of their plane-ticket purchase, if, by having that access, the terrorists could have been prevented from getting on those planes?
Dead people have plenty of privacy.
Living people don't. The best we can hope for is to be left alone most of the time. But as long as we live in human communities, somebody's going to see at least some of the things we do.
That's OK. Because one of the reasons we live in communities is so that, collectively, we can be safer than we would be if we lived alone in a wilderness where we had to fight off all dangers alone.
We always sacrifice some freedom and privacy in order to have safety, and we sacrifice some safety in order to have privacy and freedom.
America has struck a pretty good balance, and when it leans so far one way that we lose too much of the other, we adapt a little and swing back.
Our sacred freedoms were in far more danger from the Florida supreme court in November and December, 2000, than they could ever face from Admiral Poindexter's research project.
But just as the U.S. Supreme Court gets vilified for having prevented the theft of a presidential election by a law-flouting court, so also Admiral Poindexter is being attacked as a menace to privacy by the worst privacy-violators in our society.
So it goes.
Copyright © 2002 by Orson Scott Card.
Reproduction in whole or in part without permission is prohibited.