Michigan Today - September 2007

Text Size

September 2007 | Home

Talking about science

When do we know that we know?

by Tim McKay

Arthur Thurnau Professor of Physics and Astronomy

email | print | respond

On February 2, 2007, the New York Times published a remarkable headline: "Science Panel Calls Global Warming 'Unequivocal.'" The language is pretty strong. It's clear this panel intends to convince, to assert that this global warming is real. That it's a scientific fact.

Scientists talk a lot about facts, about things that we know are true. And it's true that most of what we know is so well established as to be beyond doubt. Dinosaurs once walked the Earth. Lightning is an electrical discharge. Continents drift. These things are true, part of reality.

It hasn't always been so. Each of these facts, like every bit of scientific knowledge, began as a mystery. Early explanations often seemed preposterous and faced ridicule. But in each case the gradual accumulation of tested predictions grew until the collected weight of evidence became overwhelming. At some point, a key moment, almost all individual scientists personally endorsed these ideas, and new scientific facts were born.

How does this happen? When do we know that we know something?

The first thing to understand is that this process isn't sudden or completely clean. Sometimes it happens fast, especially when new kinds of data become available. But usually it takes time and involves a lot of steps, both forward and back. This intermediate state, suspended between mystery and understanding, is where scientists spend their lives. We don't work on what we know. That's the stuff of textbooks. Most of us don't work on absolute mysteries either, questions we don't know how to address. Scientists work between these extremes, climbing toward understanding along a slope which continually recedes.

One sign of this gradual shift in confidence from meager to complete is the long standing tradition of scientific wagers. During the early days of a new idea, when it's clear no one can prove their case convincingly, the bets between contending scientists almost always involve wine; a nice bottle, if the stakes are high, even a case.

But some facts are so well established, I'd literally bet my life on them, and you would too. Most of us think nothing of climbing aboard a plane, knowing that flight is well enough understood to risk being lifted 35,000 feet in the air.

This process of establishing scientific knowledge is a fascinating one, involving both technical issues and the sociological issues always present when people try to reach consensus.

I've personally witnessed some extraordinary shifts as part of my own work in cosmology. Way back in the 1930s Edwin Hubble discovered that the Universe is expanding. Thousands of observations made during the ensuing 70 years have all confirmed this remarkable fact. During the 1990s, it became possible to measure the cosmic expansion history, to find out whether the rate of cosmic expansion was different in the past. Just about everyone expected to find that the expansion used to be faster, that gravity was gradually slowing down this flying apart. Confidence in this was high, so much so that the new experiments set out to measure the "deceleration parameter."

But it didn't happen that way. In 1998, two competing experiments both came out with results suggesting that the cosmic expansion is actually accelerating, that galaxies are receding from one another more and more quickly as time goes on. This was a big surprise, and initial reactions were mostly skeptical. Most of us thought something was wrong with the measurements. Despite the skepticism, this was an exciting result, so many people went to work on it. Some ran tests, made new observations, and cross-checked data. Others thought about what it might mean if it were true, and how could we confirm it.

In less than a year, reality had begun to emerge. New observations provided experimental confirmation, while theoretical considerations revealed ways in which this surprising result actually simplified the picture. In May 1999, I attended a conference in Paris which brought together much of the cosmology community. Several experimental teams presented new expansion history results, all in rough agreement. At the end of the session everyone looked around the room and thought about what they'd learned over the last year. An unspoken chorus seemed to pass through the room, saying, "Well, I guess that's it." Things had changed. We knew something new.

Of course this isn't really the end. No scientific conclusion is ever completely final; everything is open to revision. But over the past decade the evidence for the accelerated expansion has become much more solid. Remarkably, we still don’t really understand this cosmic acceleration. It is widely considered one of the greatest mysteries of physics. But it's well on its way to being established fact.

 That change mainly occurred among a select group of cosmologists. Sometimes we're all a part of it, as with global warming.

The global warming debate is the stuff of headlines and the topic of daily conversation. Michigan students returning from winter break last year were all abuzz with tales of barren ski slopes and snowmobiles stranded forlornly in garages. Evidence of warming has been accumulating for decades, gradually growing in weight and certainty. Theoretical understanding of how climate works has advanced apace. This is important, because it allows us to learn not only what is happening in the climate, but perhaps also why it's happening. Just now, the tide seems to be turning on this topic.

Why? What's changed? Nothing sudden, really. On the technical side, among the scientists, there's the gradual accumulation of more and more data, each bit more and more carefully checked. This wave of detail has become, for most who work in the field, irresistible. For the rest of us, the decision comes more from impression. We all compare what we're hearing with our personal experience, weighing our limited exposure to the technical work alongside. Something about the last few years (the insistence of the experts, a couple of the warmest years on record, conversations with friends and family) has pushed many people over the edge. The general mood about this is clearly different. We have found new facts.

Science exists to establish facts. Doing so is a process; a technical, human, social process. Allowed to play out fully, it works great, and has established most of what we know. Unfortunately, the natural human desire for quick and certain answers often conflicts with this. The news media magnifies the trouble, focusing on the first announcement of the newest result, and rarely returning to follow the whole story.

Remember this when you read about something truly new in science, just announced and not yet fully digested. Think it over, weigh the evidence, make a bet if you’re excited about it. Then give it a while to settle. In six months or a year the situation will be much clearer.

Back to top

Tim McKay grew up in Michigan, in love with the natural world. He studied physics at Temple University and the University of Chicago, and now teaches it to anyone who will listen.

McKay joined the University of Michigan faculty in 1995, where he is currently Arthur Thurnau Professor of Physics and Astronomy.