Lunch time rolled around, and Ron offered to take me out to eat. We walked over to a small (really excellent) German restaurant on Washington Road. What followed was a conversation I'll never forget. I recounted it recently to Ron, who told me that the only detail I got wrong was remembering him as ordering wienerschnitzel. "I've never had a bite of meat in my life," he said.

The Possibility of Design

The philosophy of science program at Pitt is one of the best in the world. I had struggled through difficult but deeply rewarding courses with Carl Hempel, Adolf Grünbaum, Jim Lennox, and others, where the question of the definition of science often came up. I observed to Ron that the philosophers (and scientists) I knew best did not agree about whether design by a non-human intelligence qualified as a scientific explanation.

On the one hand, Charles Darwin had refuted the theories of special creation of the early 19th century -- and thus such theories were testable, not least because they had been tested and falsified. On the other hand, however, the strong positivism that permeated the atmosphere of the 10th floor of the Cathedral of Learning, the home of the history and philosophy of science program at Pitt, often held that "supernatural" explanations were untestable in principle.

But if such theories were untestable in principle, why did so many of my professors, from both philosophy and biology, talk at length about data that did or did not support Duane Gish's creationism, or "scientific creationism" generally (au courant at the time because of the various "balanced treatment" cases in US federal courts). If Gish's arguments could be countered by evidence, then the dialectic of science was already fully engaged. Whatever evidence can challenge, evidence can support. Right?

Intelligent causation, I said to Ron, seemed to me to have been unjustifiably excluded from the roster of candidate hypotheses for the origin of life. Life could have been designed. That might have happened, as an empirical possibility, and whatever is possible ought not to be excluded from science a priori. (Some possible states of affairs might turn out not to be the case, of course, but that is a matter for empirical inquiry, not definitions.)

Of course design is possible and could have happened, Ron said to me, tucking into his meal. That's not the problem.

This answer stunned me, and today, almost 23 years later, I can still experience the sense of amazement and shock. One grows accustomed to positivism after a while, and the familiar "science" and "religion" categories had been well-buttressed by multiple lines of argument from very bright people indeed on the 10th floor (albeit with the glaring inconsistencies mentioned above, e.g., 'Wait until Duane Gish sees this new transitional fossil!' -- and with a long historical record of shifting definitions and practices of science shoved to one side). I fumbled out a reply to Ron: But that's not fair, I protested. Where was the justification?

Ron shrugged. You're right, he continued, it isn't fair.

Science as a Game with Rules

But think about it this way, he went on. Why is it that when a batter in baseball hits a foul ball, he has to stay at home plate (assuming no one catches the ball)? Why can't he run to first base?

If you're going to have a game, he continued, you've got to have some rules. For a long time now -- really from the middle of the 19th century -- one of the rules in science has been that the hypothesis of supernatural design is excluded from scientific discourse as a candidate explanation. Just as in baseball, where the first and third base lines define the field of play, in science one of the defining rules has been that the hypothesis of design, although quite possible, falls wholly outside the lines of admissible discourse.

Ron then referred me to Neil Gillespie's classic treatment, Charles Darwin and the Problem of Creation, where this problem is much discussed. The exclusion of design is long-standing, Ron concluded, and unlikely to change. That's just the way the game is played nowadays.

It's not fair, he said, but those are the rules.

I couldn't think of any reply to this -- after all, a rule is a rule is a rule -- and so our conversation moved on.

Ever Watch A Cricket Match?

There are no foul lines. Sure, there's an outer boundary to the field, but it's an oval encompassing the entire field of play. From the batman's position the ball may legally be hit in any direction (360 degrees). Does that mean the game of cricket is thereby impossible -- with no foul lines?

Of course not. It's just not the same game as baseball. That should alert you to a defect -- and not a minor one -- in the view of methodological naturalism as an historically contingent convention (a rule) that by dint of its necessity now defines science in principle. Rules change, depending on the game being played, and even within the same game, participants may suspend or adopt a rule, given new circumstances. (I'm old enough to remember the NBA before the 3 point shot and the American League before the designated hitter.)

Fast forward to 1986, and graduate school. I'm taking a course on Newton with Howard Stein, which means actually having to read as much as one could understand of Newton's Principia. The General Scholium of the Principia is boldly creationist in its arguments, so I worked up the nerve to write a term paper for Stein, saying that the a priori exclusion of intelligent design from science was justified neither historically (there the idea is in the Principia, arguably the greatest single text in the history of Western science) nor on logical grounds. To reinforce my point, I cited the atheist Bertrand Russell's favorable gloss of the design argument (the only argument for God's existence Russell thought had any merit):

This argument has no formal logical defect; its premisses are empirical, and its conclusion professes to be reached in accordance with the usual canons of empirical inference. The question whether it is to be accepted or not turns, therefore, not on general metaphysical questions, but on comparatively detailed considerations.

(B. Russell, History of Western Philosophy [NY: Simon & Schuster, 1945], p. 589)

When I got my paper back, it was clear that Stein hated it. He gave me a B, a "pass but barely" grade in the program. But his reasons for hating the paper were surprising.

I had struggled for pages to show that design was a genuine empirical possibility. Stein was quite impatient with this approach. In the tiny and precise hand that filled the margins of any paper one gave him, Stein wrote that I had merely labored to establish what Russell (and anyone else thinking clearly) had already conceded, namely, that design was indeed possible. Get on with it, Stein urged. What's the evidence?

Not a word from Howard Stein about the necessity of methodological naturalism. Russell's "comparatively detailed considerations" were evidential, not in-principle, and naturalism as a provisional epistemology would work only if evidence, not metaphysics, were running the show.

Because naturalism -- the ultimate causal sufficiency of autonomous physical laws -- might be false. The best way to discover its actual strength, therefore, is not to assume naturalism's truth without challenge, but to let other contenders into the field of play. Ordinary testability is more than enough as a ground rule for science.

Unless one wants to win, no matter what the evidence. In that case, make methodological naturalism an in-principle stricture on scientific reasoning.

Rig the game, in other words.