Alex Hyun

The Limits of Evidentialism

— Originally published in print // Permalink

It is not uncommon to hear a religious skeptic denounce belief in God on the grounds that there is little or no evidence for His existence.Alvin Platinga, “Justification and the Classical Picture,” Warranted Christian Belief. Being accused of lacking good evidence for one’s beliefs certainly sounds insulting; but what, exactly, is supposed to be so objectionable? I suspect that it’s not merely an allegedSee Jerome Gellman’s Experiencing God and the Rationality of Theistic Belief and Alvin Plantinga’s “Naturalism Defeated” (PDF). lack of evidence that draws the skeptic’s rebuke. Rather, it is the further thought that, if the Christian lacks evidence for God’s existence, she must therefore be irrational or unjustified in believing that God exists. The implicit principle governing this inference is what I will call the Evidentialist Principle (EP):

EP: A person’s belief is rationally justified only if she has good evidence for that belief.

Let me clarify my use of the term “good evidence.” If someone has evidence for one of her beliefs, she has some kind of argument for that belief. If somebody has good evidence for her belief, then the argument she has in support of that belief is a good one: that is, it seems valid, its premises are prima facie plausible, and it lends considerable evidential support to its conclusion.

If EP is correct, then the skeptic is surely right in thinking that the Christian who lacks good evidence for her belief in God is in dire epistemic straits. For she would then be open to the following Evidential Objection:

The Christian does not have good evidence for her belief in God. A person’s belief is rationally justified only if she has good evidence for that belief. Therefore, the Christian is not rationally justified in believing in God.

Most Christians have at least once heard a skeptic critique religious belief using the Evidential Objection or something very similar to it. One way to respond to this argument is to refute the premise: “The Christian does not have good evidence for her belief in God.” This is the strategy used in much of today’s Christian apologetics.See Lee Strobel, The Case for a Creator; Norman Geisler and Frank Turek, I Don’t Have Enough Faith to be an Atheist; J.P. Moreland, Scaling the Secular City; and Richard Swinburne, The Existence of God. But this is not the only way to undermine the Evidential Objection. Another way would be to attack ep itself, which is the other premise of the argument. It is this latter strategy that I will pursue. I am not saying that having evidence is useless when it comes to justifying our beliefs. I do not wish to challenge the view that evidence is useful, but, rather, that having evidence is always necessary for rational justification. I am not arguing that one’s belief can be rationally justified when there is good evidence against that belief. I am only arguing that it is possible that one’s belief be rationally justified when there is no evi­dence for that belief. If my argument succeeds, then it still might be the case that a theist’s justification for her belief in God is defeated by good evidence for God’s non-existence. I contend, however, that ep, which is an essential premise of the Evidential Objection, is plainly false.

Critique of EP

I will offer a counterexamples to EP: belief that one’s memory is at least sometimes reliable.

First, consider the belief that one’s memory is at least sometimes reliable. A person’s memory is at least sometimes reliable if and only if her memory sometimes gives her true beliefs.Tom Morris, Philosophy for Dummies (New York: Wiley Publishing, Inc., 1999), 58. It certainly seems like we are rationally justified in believing that the faculty of memory is at least sometimes reliable. We reveal our trust in the reliability of memory when we form beliefs about what we ate for breakfast this morning, what party we went to last Friday night, and which elementary school we attended. And normally, we take this trust and belief in the reliability of the memory to be rationally justified. Imagine the poor soul who failed to have this trust in his memory. He would have to be agnostic about everything that happened prior to the present moment, and surely such agnosticism is irrational. This suggests that it is rationally justifiable to believe that one’s memory is sometimes reliable. Indeed, such a belief might even be rationally obligatory.

But this is a problem for the proponent of EP: while we are rational in believing that our memory is at least sometimes reliable, it seems that we have no evidence for this belief. Now, it is naturally tempting to think that there is a certain sort evidence for the memory’s reliability. For surely I have often parked my car in a large parking lot and, upon finishing my grocery shopping (or whatever), been able to find my car quickly and easily by consulting my memory about where I parked my car. So there you go: I have reason to believe that my memory is at least sometimes reliable.

The fatal problem with this argument is that it is clearly a case of circular reasoning: it assumes precisely what it seeks to prove. After all, I can know that I’ve found my parked car many times in the past only if my memory is working reliably. So the above argument that the memory is sometimes reliable can work only if we assume that our memory is sometimes reliable. Not very impressive. Can the proponent of EP cite any other evidence for the reliability of memory? If so, the burden is surely his to produce such evasive evidence. I don’t know what this evidence would look like since appeals to past experience are, in this case, illegitimate. Given the failure of the only natural way to argue for the memory’s reliability, we are prima facie justified in believing that there is no successful argument. So it seems we have a counterexample to EP: belief that one’s memory is sometimes reliable is rationally justified, yet we lack evidence for it.

A Revised Evidentialist Principle

We have seen that EP is false. As a result, the Evidential Objection to Christian belief is unsound, for this argument uses EP as a premise. But surely EP captures, even if in a distorted way, something of our commonsense beliefs about rational justification. After all, there are a great many beliefs that do seem to require supporting evidence before they can be justified. The beliefs that some bachelors are tall and that Professor MacFarlane has an even number of hairs seem to be beliefs of this kind. Intuitively, these and many other beliefs do require evidence in order to be justified, which I think explains whatever plausibility that EP seemed to have.See section two of Keith DeRose’s essay “Ought We to Follow Our Evidence?” for a contextualist response. Since some of our beliefs must be supported by evidence while others do not need to be, it seems that the following Revised Evidentialist Principle captures our intuitions better than EP:

REP: Some, but not all, beliefs are such that a person is rationally justified in having them only if she has good evidence for them.

It is important to see that rep speaks only of what is necessary, rather than sufficient, for rational justification. By itself, rep does not commit us to believing that any particular belief is rationally justified.
Now, let an “A-type belief” be the type of belief that requires evidence in order to be justified; let a “B-type belief” be the type of belief that does not. The all-important question then is this: is belief in God an A-type belief or a B-type belief? In light of the existence of B-type beliefs, any successful evidentialist objection to theism must give us good reason to think that belief in God is an A-type belief. I think this task is quite difficult. I doubt that it can be done. Space constraints prevent me from defending the B-type status of theistic belief.See chapters five and six of Plantinga’s Warranted Christian Belief for an account of the nature of rational justification (or “warrant,” as he calls it) under which theistic belief is a B-type belief. It is worth pointing out that many externalist accounts of justification would deem belief in God (as well as all other types of belief) B-type beliefs. But at the very least, it has been shown that an evidential objection to belief in God is much more difficult to pull off than most evidential objectors realize.

Practical Application of our Findings

Christians sometimes worry that they have little evidence for their beliefs. Some people even leave the faith over this sort of consideration. But given our findings, surely the Christian ought not to be too hasty in jettisoning her faith. After all, many of her most fundamental beliefs are such that she is fully rational in having them even when she lacks evidence for them. Her belief in God may well be like these beliefs in this respect. Perhaps the Christian believes in God because of the Holy Spirit’s work in her heart, a process that may not furnish the believer with a cogent argument, but can be compelling for the recipient of this grace nonetheless.