There is a philosophical thesis (attributed jointly to Pierre Duhem and Willard Quine) that, when simplified, explains how a given set of facts can produce more than one apparently true conclusion: essentially, different background assumptions lead to different conclusions. A related concept is known as underdetermination: that a given set of evidence can be explained by more than one–potentially conflicting–theory.
One pertinent example: most biologists look at the diversity of species and say that evolution by natural selection (with at least a hint of randomness) is the best explanation, whereas believers in Intelligent Design see God’s hand at work. Given a certain view of available evidence, both explanations might be possible (especially if an all-powerful God simply creates everything–including fossils–in situ).
So how can we resolve this problem whereby a set of facts can justifiably be argued to support multiple potential theories?
One approach is to limit ourselves to certain kinds of theories as potential explanations: science tends to allow for only theories that are potentially testable, verifiable, falsifiable, etc. Most scientists say–despite arguments to the contrary–that the existence of a divine presence guiding evolution is simply out of bounds for scientific inquiry. It’s a matter for faith, not empirical inquiry; it’s religion, not science.
Another approach, favored by Owen Gingerich, astronomer and author of God’s Universe, turns to Aristotle to help differentiate these two kinds of explanation. Put in Aristotelean terms, faith could be seen as a search for “final” causes, while traditional science could be said to stick instead to “efficient” causes. There is thus no conflict between science and religion–and no worries about underdetermination traceable to this conflict–since each explains different things.
Gingerich looks to Blaise Pascal‘s notion that “some things only the heart knows” to explain this idea and justify his belief in (small case) “intelligent design.” Since science cannot know or determine certain truths (final causes, in Aristotelian terms), we can freely posit a (distant) intelligent designer without worrying about stepping on scientific concepts of proof. In essence, two truths become simultaneously possible, because they occupy different domains of truth.
The law is somewhat similarly concerned with causality. But the law is not concerned with science’s version of “efficient” causes, nor is the law looking for “final” causes in the metaphysical sense. Instead, legal analyses look to “but-for” causation and speak of “proximate” cause in a search for limited, but specific, legal culpability. The point, though, is similar to that advocated by Gingerich: to limit the scope and breadth of various theories of causation. In other words, the idea is to restrict potential problems of underdetermination.
Intelligent Design (not Creationism, and not the lower-case “intelligent design” of Gingerich), on the other hand, believes that science can be used to access the truth of an intelligent creator, and that this search is scientific.
Creationism, on the other hand, tends to reject science more firmly. It inherits from a tradition of the literal exegesis of scripture used, for example, in the 16th century. 16th-century exegesis is related to but not identical to today’s Biblical literalism. After all, bringing in a passage of scripture today is no longer a means of shutting down debate.
So how did Copernicans in the 16th century deal with the issue, given the power of literalism at the time? They argued that perhaps scripture itself underdetermines potential explanations–even if it can shut down blatantly conflicting theories.
Thus, for example, Johannes Kepler tried an accommodation approach with literalism. He maintained that God–in order to be understood by normal people–caused the Bible to be written in ordinary language. This is why there are no discussions of epicycles (or DNA, for that matter) in the Bible. The Bible thus accomodates ordinary folk with a different, non-scientific vocabulary that, if read correctly, does not conflict with science.
Of course, many–most?–of today’s scientists simply step outside of the argument, and simply point to materialist, naturalistic explanations as being all that is necessary, and certainly as the only valid scientific theories. Why? Because they work.
If all of these approaches to dealing with underdetermination are dissatisfying you, and you can’t accept naturalism, then there is always the choice to go to absolute knowledge, as David Bloor reminds us: if the Pope says it’s true, then no doubt exists, and we escape the problem of underdetermination and uncertainty.
This post is based on discussion in a graduate seminar on science and religion on Monday, April 27, 2010.