Some notes on Loftus' definition of faith

John Loftus offers this definition of faith.

John Loftus: "Faith is an attitude or feeling whereby someone attributes a higher degree of probability to the evidence than what the evidence calls for."

He goes on to say:

I further argue that reasonable faith is an oxymoron because having faith is an irrational leap over the probabilities, which again, is what all Christians do in practice. 

I replied by saying;

This strikes me as a simple error in linguistic philosophy. If Christians develop and use the term "faith" then you have to pay attention to what that linguistic community means by it. Irrationality isn't part of the meaning of faith. It may be that, in believing their religion and in having faith, they commit irrationality, but this is contingent on the evidence being different from what they suppose it to be.

If you assessed the evidence and you thought that Christianity was probably true, you could still have faith. This would be true if you were right about the evidence or wrong about the evidence. You can't presume your assessment of the evidence to define words that Christians use. Link.

And he in turn said: 

Sure we can, and we do.

Definitions of words couch arguments in disguise. In our arguments we demand a different definition of "faith" because our definitions more accurately describe what we see from believers around the world in their respective sects. It's the same phenomena all over the globe.

When it comes to evaluating the strength of the evidence for a claim, any claim, my definition works just fine. After all, we're not just talking about a word but a concept, one that should apply to everyone and not just Christians, otherwise you are special pleading. Agreement among Christians isn't a criteria for how a word is used anyway, especially since the debate over the word has been framed in the western world by Christianity.

All you need to do, Vic, is say that what you have concluded is more probable than other alternatives and that you do not claim a higher probability to the evidence than what the evidence calls for.

Can't do that?

Why not?

I can say that. I can say that I do not hold any beliefs that I consider to be less probably true than their denials. If I thought they were less probable than their denials, I wouldn't believe them. 

Now, I suppose, given lottery paradox considerations, I probably should say that I hold some beliefs as probably true which are actually probably false. But, for every belief I hold, I consider the belief to be more probably true than its denial. Yet, I maintain that I have faith in the existence of God. 


Nor does it seem plausible to me that I can only truly have faith if, in fact, I have mistakenly assessed the evidence. It doesn't seem plausible to me that when someone exhorts me to have faith in God, that what they are exhorting me to do is to make an error in my assessment of the evidence.

Now, I don't think Loftus wants to say that everyone who has religious faith recognizes the irrationality of their beliefs in the act of having faith. That would be implying that all those who have faith are fideist, and understand their position to be fideistic. But, they don't. Thomas Aquinas, C. S. Lewis, and William Lane Craig explicitly do not understand faith in this way. They think that they hold reasonable beliefs, but nevertheless they believe that they exercise faith anyway. 

The trouble is that if you make a characteristic part of a definition, you implicitly say that you couldn't use the term unless it the characteristic were present. Suppose whiteness had been part of the definition of being a swan. If that were the case, then we couldn't possibly have discovered black swans in Western Australia. The black birds we found over these would not have been white, and therefore would not be swans. But since, when we encountered these black birds, we called them swans and rejected our previously accepted idea that all swans were white. 


If you think that no one has a reasonable faith, that doesn't license you to make unreasonableness part of your definition of faith unless you are prepared, in case you do encounter someone who has reasonable faith, to say "OK, that's reasonable, so it can't be faith." In other words, in the (unlikely?) event that Loftus should change his assessment of the evidence concerning Christianity to correspond with that of William Lane Craig or Richard Swinburne, he would then have to say "Fine, Christian belief is reasonable, but what that means is that these people don't have faith. They have something else instead, such as knowledge. The are making a linguistic error by describing themselves as having faith."

What Loftus wants to say that, in fact, religious believers, in the course of having faith, believe things that are less probable than not to be true. I take it he is implying that there are objectively valid probability assessments for beliefs, something I am kind of skeptical about. Probability theory tells us how to go from one probability to another based on evidence, via Bayes' theorem. What it doesn't tell you is where the initial probabilities are supposed to come from in the first place. If we can't use personal probability assessments, as some people have maintained, the where are we supposed to get our probability assessments in the first place? Does Loftus have an epistemic probability theory that he thinks works?


A good definition of faith has to be one that language users of various persuasions can agree on. Loftus' account could succeed as a description of faith (this is what happens every time someone has faith) but it can't be a definition of faith.  


Further, a definition of faith needs to fit with our use of the term outside of religion. For example, a manager might have faith in his ace pitcher after he walks the first two batters. Now, faith does need to be in spite of something, but it doesn't need to be irrational, at least in this context.  The manager could have good reason to retain confidence in his pitcher despite his walking batters in the beginning of the game. 


I am speaking here as a fussy analytic philosopher, and not as a Christian apologist.