Business science and business religion

By David Weinberger

"Religious beliefs are absolute," said the posting to the archly realistic mailing list, "and as such, they are unfalsifiable." They are, according to the writer, therefore in the same category as beliefs about Santa Claus and preferences for boysenberry yogurt. This raises two questions for business: How many business beliefs are falsifiable? And are the ones that are not falsifiable really of no more value than a belief in the Easter Bunny?

Falsifiability is an important concept. It's also widely misused to discount knowledge that needs much more serious consideration. The concept was invented by Karl Popper (1902-1994) as a way to distinguish science from non-science, but it also explains science's great strength. Popper said that a scientific statement is one that can be empirically tested and proven false. "If observation shows that the predicted effect is definitely absent, then the theory is simply refuted." So, Einstein's theory of relativity leads to particular, testable expectations; for example, a consequence of the theory is that light is bent by gravity. This was tested during the eclipse of 1919 when it was discovered that a star near the blacked-out sun seemed to be in a different spot than when it is observed at night when its rays are not passing near the sun. The displacement was precisely what Einstein's theory predicted. Had it not been displaced, Einstein's theory would have been proved wrong. That Einstein's theory is capable of being proved false indicates that it is a scientific theory. Its falsifiability makes it scientific. Similarly, my belief that there's an invisible Martian using ESP to beam psychic messages into my brain couldn't under any imaginable circumstances be proven false. Thus it isn't a scientific statement at all . . . even if it turns out to be true.

This is as clarifying as a bright light. Of course there are difficulties and ambiguities, every one of which has been worried to death by philosophers who are like Rottweilers pulling at Popper's cuffs. What counts as a disproof? What does "could" mean in "could be disproved?” When does an experiment have enough weight to overturn an entire theory? Valid questions, but Popper's essential insight remains: What makes a statement scientific is that it could be proved false.

Religious statements fail the falsifiability criterion. What are the deductions from "Jesus was the son of God" that could be empirically tested? Under what circumstances could we disprove the Jews' contention that we were chosen by God to fulfill His commandments? But so what? There's more to believe than what science tells us. The problem comes, of course, when something unscientific, such as astrology, makes scientific-sounding claims. Religion usually doesn't make a pretense to being scientific, so the fact that it's not falsifiable doesn't discredit it.

Falsifiability has important uses in business. But, frequently, the only experiment that counts bets the company itself.

For example, the marketing department may be making claims about the market demand for a proposed product direction. Those are falsifiable claims: Run some studies and find out if the demand is there. Of course there will be arguments over the interpretation of the results. Perhaps the wrong people were surveyed. Perhaps the questions were misleading. Perhaps a marketing campaign would lay the ground for successful adoption. (Similarly, perhaps the instruments used to measure the position of the star during the eclipse were miscalibrated.) But, if the marketeers insist in the face of well gathered evidence, they are not operating on the basis of science. They may have good reasons--hunches count—but that fact should be acknowledged. Without that acknowledgment, the discussion becomes as hopeless as an argument between the Rev. Jerry Falwell and Prof. Stephen Jay Gould over the nature of evolution.

Nevertheless, even if all the studies indicate the company should take the new product direction, the ultimate experiment that proves that the decision is correct is the movement of the company itself. Marketing, engineering and sales may have tested every proposition, but to falsify a business strategy, you have to adopt the strategy. Until the company does it, no one can be sure that it will work. That is the nature of complex organizations in even more complex environments.

Here's the kicker. For a strategy to work, there has to be a ton of unfalsifiable propositions floating around. That is, people need to be able to stop talking like scientists in clipped phrases that can be tested. They need to move into and out of scientific mode, brainstorming, emoting, yelling, denying and laughing. Otherwise, there will be nothing interesting to falsify. The key is to know when you're talking business science and when you're talking business religion.

(NOTE: Most of the statements in this column are non-falsifiable.)

David Weinberger is the author of the upcoming book “Small Pieces Loosely Joined." He can be contacted through www.evident.com