The Johns Hopkins Gazette: April 7, 2003
April 7, 2003
VOL. 32, NO. 29


What Counts As Scientific Evidence?

Philosophers, historians and scientists will ponder the answer this weekend

By Michael Purdy

Johns Hopkins Gazette Online Edition

Johns Hopkins' Center for History and Philosophy of Science will sponsor a first-of-its-kind seminar this week on the nature of scientific evidence. The interdisciplinary seminar, which will be held on the Homewood campus Friday through Sunday, April 11 to 13, has free registration and, in addition to featuring some current Johns Hopkins faculty members and scholars from other universities, will bring back a number of doctoral degree recipients in philosophy of science who have gone on to become prominent in their fields.

Philosophers, as well as scientists, have disagreed and debated for centuries on what counts as scientific evidence, developing a number of different theories, said Peter Achinstein, professor of philosophy in the Krieger School of Arts and Sciences.

"The concern isn't so much how the scientist got the idea for his hypothesis but how he decides what will be the most convincing evidence he can use on skeptics," said Achinstein, organizer of the conference and author of the recent book The Book of Evidence. "Fundamentally, it's a problem of, Why should other people believe what I'm saying?"

Scientists' insights regularly reach beyond their ability to test new ideas directly through experimentation, and that can make the question of proof a particularly tricky one. Achinstein cited Isaac Newton, one of history's most famous scientists, as an example. Newton's groundbreaking description of gravity asserted that it was a force exerted by every mass on every other mass throughout the universe.

The provocative idea was impossible to directly test or verify, so Newton established four rules that emphasized the scientist's role as someone who draws reasoned conclusions from observations of natural phenomena. If a theory could be inferred from the observed phenomena using Newton's rules, then the theory was acceptable science.

In the 19th century, some scientists began to assert that basing all theories on observable phenomena in the manner suggested by Newton was too constrictive and didn't allow enough room for imagination and novelty. Achinstein said English scientist William Whewell was one of the most important proponents of the new method, which emphasized the idea that a hypothesis must organize and unify observational knowledge.

"Whewell argued that a hypothesis has to give you back more than what you already knew," Achinstein explained. "He said it had to give you predictions that you could test, and this became known as the hypothetico-deductive model. You start with a hypothesis, you use formal mathematics to deduce the consequences of the hypothesis, and then you go out and test to see if its predictions are true or not."

Whewell asserted that the best test of a theory created via this route was its ability to account for new observations as science's abilities expanded and improved over time.

"In Whewell's view, as more and more observations were acquired over time, the theory should become simpler and more coherent, neater rather than more complex," Achinstein said. "Further observations should only bring out the beauty of the theory."

An active debate over what counts as scientific evidence continues today. One key issue is whether a universal theory of scientific evidence can be developed. Achinstein noted that Thomas Kuhn, author of a well-known book titled The Structure of Scientific Revolutions, has argued that every scientific revolution brings with it a new theory of scientific evidence.

Among those developed more recently include probability-based theories, which became useful for discussing evidence in contemporary physics.

"One of our speakers at the conference, Kent Staley, a former student of mine now at St. Louis University, has written a forthcoming book on the search for evidence for the existence of the top quark," Achinstein said. "The signal-to-noise ratio in the data is very bad, and the logic of the reasoning is very complicated."

High-energy physics also will be represented at the seminar by Adam Falk, vice dean of faculty in the Krieger School of Arts and Sciences and a professor of physics and astronomy.

"I work in high-energy physics, and quantum mechanics is an inescapable part of my scientific world," Falk said. "Quantum mechanics is very strange, counterintuitive and poses interesting issues for philosophers that go to the heart of what it is to exist, what it is to cause and what it is to have evidence for something."

Achinstein highlighted the fact that Falk will speak on Sunday on issues of evidence in physics. He was eager to have a working scientist on the schedule, and Falk was glad to oblige.

"Conversations between scientists and historians and philosophers of science are not as rich as they could be, and coming as I do from a family with both philosophers and scientists makes me more sympathetic to exploring ways of asking questions in both fields," Falk said.

Additional faculty speaking at the seminar are Sharon Kingsland, Lawrence Principe and Steven vanden Broecke, all from the Department of History of Science, Medicine and Technology. Returning students speaking at the conference are Staley, Steven Gimble of Gettysburg College, Frederick Kronz of the University of Texas-Austin, Richard Richards of the University of Alabama and Alex Rosenberg of Duke University.

Achinstein plans to publish the proceedings of the conference and hopes to make the event a regular Hopkins tradition.

The conference begins at 7:30 p.m. on Friday with a general symposium on theories of evidence featuring Achinstein, Kronz and John Norton of the University of Pittsburgh. It will be held in the Sherwood Room of Levering Hall. For a complete schedule, see Calendar, this issue.