Some of you may be acquainted with Tom W. Clark and his excellent naturalism.org web site. In 2007 Mr. Clark published an article titled When Worldviews Collide: Root Differences Between Theism and Naturalism which discusses his impression of a debate between theist philosophers Stewart Goetz and Charles Taliaferro and atheist philosopher Andrew Melnyk. I do not have strong opinions on many controversies, particularly regarding questions that focus on technical questions requiring expertise that we non-experts do not possess, but with this debate I am unequivocally in agreement with the atheist side. One of the reasons I am so comfortable taking sides here is that the only expertise that is needed to evaluate this controversy, despite the academic rigor of the discussion, is a decent general education.
I very much agree with Tom Clark's analysis that at the bottom of the disagreement between the theists and atheists (a.k.a. naturalists) is a different approach to evidence, or in the words of Tom Clark "....disagreements about the explanatory potential of dualism, about the epistemic status of intuitions and data, and about what counts as good explanation."
As Tom Clark argues, "A striking methodological difference between the two sides, one that helps explain their differing takes on reality (dualist vs. monist), has to do with the status of what T&G call first person data. They put great stock in the validity of what they believe are widespread and commonsensical intuitions about metaphysical matters, intuitions deriving from personal experience.... But from a philo-scientific perspective, the claim that some intuitions or experiences wear truth on their sleeve and can’t be second-guessed is to let the tail of data wag the dog of theory.... We shouldn’t trust intuitions, however widely they might be shared, as direct apprehensions of what’s real since they are notoriously unreliable: mass delusion is possible. Instead, we must test intuitions against objective evidence."
Mr Clark continues: "Transparency and reliability come from having specified and verified the existence of all entities, mechanisms, and events that participate in the explanation, such that there’s nothing mysterious or ad hoc involved.... A transparent explanation, at least for naturalists, can’t have gaping holes, filled with unexplained, ad hoc explainers. If some things currently escape explanation, so be it; that simply makes life more interesting." In other words, theists have an unfortunate tendency to mistake weak arguments for strong arguments as a result of applying a too lax standard regarding what qualifies as reliable supporting evidence. Or, as Mr Clark says "Yet their departures from good philo-scientific practice can best be explained, I think, as a function of putting the desideratum of a purposive reality above the desiderata of explanatory transparency, evidential reliability and cognitive coherence."