Uncertain Science . . . Uncertain World

By Henry N. Pollack ’63 PhD, Cambridge University Press, 2003, $28 hardcover.

U-M geophysics professor Henry N. Pollack ’63 PhD explores uncertainty in various contexts—from the laboratory to the stock market to the battlefield—and shows readers how to use their everyday experiences to evaluate and understand uncertain science. In this interview with U-M science writer Nancy Ross-Flanigan for Michigan Today, Pollack explains why he wrote the book and touches on some of its key arguments.

Ross-Flanigan: What compelled you to write a book about uncertainty?

Pollack: I lecture frequently to nonscientific audiences, including alumni groups, civic clubs and ecotourists. I’ve found that the subject of uncertainty has really bypassed them. They’re so confused about it, even though there’s a lot of talk about uncertainty all the time. In most daily activities, uncertainty is just accepted as part of life: Is the market going to go up or down? Will there be a war next month when I have to travel? But when it comes to uncertainty about science, people seem to have a higher standard. They want science to be certain.

Why do you suppose that is? People are accustomed to science giving them very good, very reliable, very certain answers. We’re able to control robots on other planets, we’re able to predict the time of ocean tides, we know exactly when sunrise and sunset will be at any time of year any place on the globe. So when science doesn’t produce that certainty, people say, “What’s the matter, guys? Why are you on thin ice now?” In a sense, it’s a measure of our success that the public believes that science should be successful and certain in every endeavor. But scientists don’t have all the answers. If they did, they’d be wasting their time on something that’s already known. Interesting science is conducted in areas where we don’t know the answers. The uncertainties stimulate scientists to have a look.

You maintain that uncertainty, rather than impeding progress in science, actually fosters creativity, and also that certainty can be a bigger impediment to science than uncertainty. Why is that? Uncertainty says that we don’t know the answers, so it leads to a competition of ideas. Conversely, certainty says that we do know the answer, so we can stop thinking as hard about the question; it leads to closed-mindedness. There’s a saying that nothing is so dangerous as a person with an idea—if it’s the only one they have. When faced with uncertain situations, people come up with different assessments and different ways of approaching the problem. That leads to an exploration of many pathways. Some will be dead ends but, nevertheless, when you have an array of people fanning out in different directions, some are bound to make progress. Not knowing everything is not the same as not knowing anything. Most scientific endeavors are works in progress. We learn something, we have some new ideas, we discard a few old ideas, but we’re making headway.

Why is there still so much uncertainty in certain areas of science, such as climate change research? I like to point out that the uncertainties in projections about future climate are not so much uncertainties about climate science; they’re uncertainties about social science. For example, the range of outcomes depends on projections of how many people will be living on Earth in the 21st century, and demographic projections are tough business. In the 20th century, we went through a period of exponential growth, with populations doubling every 30 to 40 years. By the end of the 20th century, that rate of growth had slowed somewhat. With projections for the 21st century, the United Nations now has a high projection, a medium projection and a low projection, depending upon certain social factors playing out. One of them has to do with education—birth rates are very clearly correlated with the educational level of women. The question is, how much education will we have for women worldwide in the 21st century? These are social science issues, not climate science issues. Another question that affects these climate scenarios is, “What kind of energy will we be using?” Will we have continued dependence on fossil fuels, with the production of greenhouse gases, or will we move to other forms of energy that will have less climatological impact? That’s a good question, but it’s not a climate science question. It’s a question of economics, a question of politics.

In such uncertain situations, how can scientists—and society—decide what to do and when to do it? If you wait for uncertainty to be resolved, you’ll wait forever. You have to go ahead and act, but be ready to change course if necessary. A question that arises from time to time is, “When will the world run out of oil?” The wide array of answers is bewilderingly diverse. Optimists say, “Never.” Meanwhile, we hear from other quarters that this century will be the last in which oil powers the global economy. One thing about oil is certain, however: The rate of creation of oil in the Earth proceeds on the slow geological time scale, whereas its consumption is taking place on the fast human time scale. In other words, nature will not rescue us by making oil as fast as we are extracting it. Whatever nature has created over the long history of the Earth is the resource that we have to work with. Worldwide production of liquid oil will peak between 2010 and 2020. When that occurs, the world will be embarking on a new economic pathway.

What are the choices for our future sources of energy? There are large reserves of solid oil—more than the remaining liquid fraction—in the form of oil shale or in gooey bitumen locked in the pores in sandstone, but the highly energy-intensive and expensive technologies aren’t in place to extract it. Reliance may shift initially to natural gas, but that resource is being depleted, too. Alternative energy scenarios could, in principle, ensure a smooth transition away from oil, but the pace of development of the alternatives must accelerate if they are to take up the slack when conventional oil production begins to decline globally in 10 years or so.

Uncertainty comes into play not only in predicting the future, but also in trying to reconstruct the past. How is that situation different? Reconstructing the past is complicated by incomplete, sometimes inaccurate and sometimes conflicting information. Think of how difficult it has been for investigators to figure out what made the Space Shuttle disintegrate or TWA Flight 800 crash. Another common arena for this kind of uncertainty is the courtroom, where witnesses present varying accounts of what happened, and some details never make it into testimony. In the trial of Timothy McVeigh for the bombing of the Murrah Federal Building in Oklahoma City, for example, the FBI inadvertently neglected to present thousands of pages of evidentiary documents to the defense. In spite of all this untidiness, juries have to listen carefully, weigh and evaluate the evidence and reach conclusions. They aren’t given the luxury of saying, “We need more research.” Sometimes they make mistakes, but we value our constitutional right to speedy judgment by a jury of our peers enough to accept the decisions they make, even though they’re based on incomplete, inaccurate and conflicting evidence.

How do the tools and techniques of science help reduce uncertainty? Experimentation is a hallmark of science, and experiments are basically exercises in asking questions: “What happens if we do this to that?” You pose a question, and you propose a procedure that you hope will answer the question unambiguously. Occasionally you’ll have something that turns out to be extraordinarily unexpected—the kind of experimental result that makes people say, “I don’t believe it!” And that leads to another group testing it. Science has a way of self-correcting. If the experiment can’t be repeated, and different groups can never get the same results, then slowly, the original idea is set aside as unrepeatable. Repeatability is a central tenet of science. If something is really true, then more than one person can discover or demonstrate it.

Pollack was born into a Nebraska farm family and became interested in earth sciences in college at Cornell University, where he became fascinated by the fossils in the dramatic gorges of the Finger Lakes region.

  • U-M Facts

  • Maps

  • U-M Events

Subscribe  |  Unsubscribe