Wednesday, March 7, 2012

On "Where the Conflict Really Lies" (Pt. 9)

[Series explanation and index here.]

Chapter Six


Supposing some results of scientific method really do conflict with a person's religious beliefs, is that person required to resolve the conflict by giving up those religious beliefs? Plantinga's short answer is: no. His longer answer requires an introduction to the concept of defeaters.

A defeater is a belief that has a detrimental effect on another belief. There are two types of defeaters: rebutting defeaters and undercutting defeaters. A rebutting defeater directly conflicts with another belief (and wins the conflict). If I believe my dog is outside, then I see her dash across the living room, my belief that she's outside is defeated by the belief that she's inside. An undercutting defeater doesn't conflict so much as it takes away the reasons for holding a belief. Suppose I believe my dog is outside because I hear scratching at the door. If I open the door and see that a stray dog was doing the scratching, I no longer have a reason to believe my own dog is outside.

A belief that works as a defeater for me, might not act as a defeater for you. Scratching noises might have been my only reason for thinking my dog is outside, but you might have seen my dog through a window two minutes ago. When I open the door and see the stray dog, I have no reasons left to think my dog is outside, but you still do. Ours sets of preexisting beliefs are different.

Can you see where this is going? A religious person typically holds beliefs which aren't in the common store of beliefs from which science proceeds. Just as seeing the stray dog had different consequences for my belief and your belief, scientific discoveries may have different consequences for individuals according to their total store of beliefs.

Plantinga advises theists to admit — when appropriate — that scientific results are reasonable conclusions to draw from the limited viewpoint of scientific inquiry, but not feel compelled to accept scientific results when other conclusions are more reasonable to draw from one's total worldview.

The Reduction Test

Can a theist hold onto any religious belief no matter what scientific inquiry turns up by saying: "Science suggests not-B, but my total store of beliefs includes B. Too bad for not-B." No, because then anyone — not just theists — could do that to avoid ever giving up a belief.

Instead, Plantinga proposes a thought experiment. Take your preexisting store of beliefs and remove B, along with any other beliefs which entail B. This leaves you with a reduced store of beliefs which is as close as possible to your original store of beliefs, except B could possibly be denied. Now, add the scientific suggestion of not-B. What is the best conclusion to draw from:
(Original total beliefs) minus (B and beliefs that entail B) plus (Scientific suggestion of not-B)
Take the example of evolutionary psychology. Plantinga's original belief B is that our minds were designed by God. Science suggests our minds arose by natural processes, without an intelligent designer. Plantinga can reduce his original store of beliefs to leave open the question of whether God designed our minds and still have God creating the world, God intervening in the world, God wanting human beings to have certain mental abilities, etc. Science without these ingredients might conclude: 100% natural origin of human minds! But Plantinga can mix in these extra ingredients and come to a different conclusion without assuming (specifically) that our minds were intelligently designed.

What about a Christian who reads Old Testament poetry and so believes the Earth is rectangular? When she encounters the scientific evidence that the Earth is globe-shaped, she can set aside her belief that the Earth is rectangular and any beliefs that entail it, then see how the scientific evidence interacts with her remaining store of beliefs. Plantinga thinks the globe-shaped Earth belief will win out and replace the beliefs that led her to believe the Earth is rectangular. For example, she might have to drop the belief that poetic sections in the Bible are trustworthy descriptions of the physical world.

I think this is a decent approach. There is a lot of room to argue about how to apply the reduction test, but the exercise of putting beliefs and conclusions into these terms is at least a helpful way to organize a complex issue.


  1. I'd note that this is essentially what Bayesianism does - it changes beliefs minimally to conform with new information. See work by Caticha on Maximum Entropy Inference. Plantinga's model seems to demand even more, though I'm not sure.

    That said, the method appears too vague from your description. Without good criteria to determine how to "rebuild" your beliefs from the reduced state, reverting to it isn't all that grand. 

    I'm also concerned that the method does not take into account the complex relations between different beliefs. Plantinga's belief that god is manipulating nature is both supported by and supporting of his belief that God made us in his image, mentally. The reduction of beliefs only works so-far, in that it takes out one direction but not the other.

  2. There is a good amount of hand-waving to this reduction test as it's
    presented in the book. I don't know if it has been developed further
    elsewhere, since there aren't any footnotes or references that I see.