Epistemic extremism

On the most recent Atheist Experience (here), I think the hosts (Matt and Jamie) make some comments which go too far. I’m not sure if I have missed this before, or if it is a change in tone that has come up recently. Anyway, it seems to me to be a case of epistemic extremism.

In the show, they were talking to a recently de-converted ex-Christian, Ethan, about his newfound atheological engagement with missionaries and apologists. Ethan was explaining how he is now of the belief that there is no god. The caller had been trying to argue for this point in his conversations with Christians, rather than merely arguing that there is no good reason to believe in god. At about 28:17, the caller (wrongly in my opinion) suggests that “at some point eventually you have to make the decision about whether there is a god or there isn’t”. The hosts object to this comment quite strongly, both repeating “No” several times and shaking their heads.

It is at this point that Matt makes the following remark:

“I can be not convinced that there is a god and not convinced that there isn’t a god for my entire life”.

Now, what struck me about this is the use of the word ‘convinced’. It is a pretty strong epistemic modifier. To me, the modifier marks out the extreme ends of epistemic positioning. For instance, if p is the proposition that “some god exists”, and if epistemic confidence runs from 0 through to 10, then being convinced that p is true means something like being a 10. It really means you couldn’t be more confident of p. Similarly, being convinced that p is not true, means having a confidence level of 0 in p being true. Being convinced that p, or convinced that not-p, means being at the extremes of the epistemic scale for p.

When put like that, Matt is saying that you could be some number between 1 and 9 through your whole life. And that is true. I don’t think that for each and every proposition, you should take up either a 0 or a 10 degree of belief for it. Almost every belief I have is somewhere between 1 and 9, so I am not convinced of very much really. So, as far as that goes, I agree with Matt’s statement.

But it also seems clear to me that this notion of being convinced of p is different (and should not be muddled together with) the notion of believing that p. For instance, last night put my bike in the bike shed in my garden, and I have not yet gone to look at it this morning. I certainly do believe that the bike is in the shed. But, I am not convinced that the bike is in the shed. Unfortunately, bike thieves do operate in this area from time to time. I think it is unlikely, though possible, that my bike has been stolen. I believe it is in the shed, but I am less than convinced about that.

In a similar manner, I am of the belief that god (the Christian God anyway) does not exist, but I would not say that I was convinced that he doesn’t exist. It seems very unlikely to me (less likely than my bike being stolen, even), but it is possible.

I believe without being convinced that my bike is in the shed; and I believe without being convinced that god does not exist.

Just like with my bike, I am pretty confident about whether god exists, but I am not at the extreme far end of the scale. Maybe my degree of belief is like a 1.5 out of 10, or something like that.

I think that Matt probably has a very similar position to me on this question. He is probably of the belief that the christian god doesn’t exist, without being convinced. We’ve talked several times about this sort of thing, so I feel like I know where he is coming from.

Richard Dawkins put himself as a six out of seven on his scale of disbelief. Dawkins own self-description is:

“Very low probability, but short of zero. De facto atheist. ‘I cannot know for certain but I think God is very improbable, and I live my life on the assumption that he is not there.'” (Richard Dawkins, God Delusion, p. 50-51).

It seems to me that Dawkins is not convinced that god does not exist, but he is pretty solidly believing that God does not exist. He has a very strong belief, but he is not an extremist.

On the other side of the fence, we have William Lane Craig (about as far away from Dawkins as we can reasonably expect on the scale). Yet, even Craig declines to go all the way to the extreme of the scale. In this clip, he clearly states that he is not certain that god exists. Obviously, Craig thinks he has very good reasons to believe that god exists, and he does believe pretty strongly that god exists. Yet, it is wrong to put him at the extreme end of the scale either.

In some very important sense then, neither Dawkins nor Craig are convinced (in either direction) about whether god exists. While there may be people who do land on the epistemic scale at a point which is more extreme that either Dawkins or Craig, I think it is safe to assume that the vast majority of people are somewhere in between these two. Hardly anyone is more convinced than these guys, and even they are not convinced.

Matt’s criteria, of not being convinced either way, is so weak that it ends up covering people with such diverse opinions on the same topic as Dawkins and Craig, both of whom come under the description of being “not convinced that there is a god and not convinced that there isn’t a god”.

[As an aside, the definition of an atheist as someone who is “not convinced” that there is a god, is kind of absurd if it ends up classifying William Lane Craig as an atheist.]

Matt goes on to make the courtroom analogy:

“This person has been accused of a crime. Do you think he is guilty? No. Do you think he is innocent? No. Do you ever have to make up your mind? No.“

Now, if we consider the standard of evidence in a court of law for very serious crimes, like murder, the standard used is ‘reasonable doubt’. It is true that given this standard, I would not be able to ‘convict’ most propositions as either being true or of being false. There is a ‘reasonable doubt’ about whether my bike is in the shed. Until I go and look to see if it is there, I am not able to make such a strong claim. Yet, I still believe (quite strongly) that my bike is in the shed. So I might not convict someone of being guilty of murder, yet still believe that they are guilty.

This makes me think that the courtroom analogy, and the notion of being ‘not convinced’ about the truth of a proposition, just obviously don’t track with the everyday sense of believing in things. We often believe things that we are not convinced in, and of which we wouldn’t be able to use the reasonable doubt standard to overcome. And I don’t see that this alone is irrational in any way. Am I being irrational for believing that my bike is in the bike shed, even though I am not convinced of it? I don’t think that is irrational at all.

Ethan replies to these comments, by referring to Jordan Peterson, who he claims just ‘tap dances around’ an issue, instead of laying out reasons to think it is true. At this point, Jamie jumps in and says:

“But if he is going to tap dance around, can’t he tap dance around and show weaknesses in the way that you have presented evidence for your claim? Wouldn’t it be better if you made him play defence on a battleground that very clearly he can’t hold?”

Jamie is clearly suggesting that Ethan shouldn’t make the claim that god doesn’t exist, but instead try to make his interlocutor ‘play defence’ for their claims. Then Matt joins in by saying:

“Ethan, by making the claim [that god does not exist], you have put yourself on a battleground that you can’t win”.

Here is where I think the main disconnect really kicks in. Matt’s claim clearly presupposes the idea that being convinced, or being beyond reasonable doubt, is the standard we should be using. But, if the claim is merely a belief claim, then this just seems wrong.

If I make a claim, like “I believe that god does not exist”, I do have a rational requirement to be able to justify that claim if someone challenges me on it. If I have no reason whatsoever, then (perhaps) that means that I cannot be rational in holding the belief. I could also have something which is a bad reason for having the belief. I need to be able to say something better than “Because I flipped the coin and got heads-up” for why I believe the proposition. So something is required (not nothing), and it needs to be a ‘good reason’ (not just flipping a coin, etc). But does it have to be enough to convince me? It seems obvious to me that the answer to that is: no. I can rationally believe something without being convinced of it. Think of the bike example. These make up the vast majority of our beliefs. Do we want to say that the vast majority of our beliefs are irrational, just because we are not convinced that they are true? I don’t think we do. Saying that we do sounds like epistemic extremism to me. It sounds antithetical to the sceptical, scientific, rational outlook the hosts usually try to defend.

But now Matt goes on to make some even more bizarre comments. He says to Ethan:

“Prove to me that you are not a mass murderer”

Ethan falters and confesses that he could not provide a proof of this that would convince Matt. But hold on a minute. What is the standard supposed to be here? Is Ethan only supposed to make claims that he has good enough reason to be convinced of, or to only make claims that he has good enough reason to convince Matt of? Which one is it?

Presumably, Ethan can be very confident, and have excellent reasons, to hold the belief that he is not a mass murderer. It seems almost unimaginable that you could forget such a thing. It’s logically possible, but it is way less likely than my bike being stolen, and there is nothing wrong with believing that the bike is in the shed. Ethan clearly has good enough evidence to rationally justify his own belief that he is not a mass murderer.

Can he justify it to Matt to the same extent? Well, possibly, but not over the phone in 2 minutes. What sort of significance are we supposed to derive from the fact that Ethan cannot summon up evidence over the phone to a complete stranger that he is not a mass murderer? Should we use this as a standard which means that Ethan shouldn’t claim to not be a mass murderer? This seems wrong to me. Ethan can certainly have the belief that he is not a mass murderer, and should be able to say outloud that he has the belief.

If I claim to be thinking about the number 7 right now, there is nothing (in principle) which could convince you beyond all doubt that I really am thinking of it. The same goes for all claims about the contents of our consciousness, such as that I am cold, or hungry, or like jazz music, or have a headache etc. I can be immediately aware of it, and to that extent I really am convinced of it, but I can give you no evidence beyond telling you. If you don’t believe me, there is nothing I can do to persuade you. But does that mean that we are not allowed (rationally) to report to others what we feel like, or what we are thinking about from time to time? Am I breaking a rule of sceptical discourse if I do so? I think not. Yet this does not meet Matt’s demand. I cannot prove to him that I am thinking of the number 7.

The standard for making a claim (most of the time) is not that you have enough evidence to convince your interlocutor. You do not have to be able to persuade them beyond a reasonable doubt. If you merely believe a proposition, without being convicted of it, then you have some justificatory burden if you make the claim, but it is not the same burden as it would be if your belief was at the extreme end of the epistemic scale.

If I said I was convinced that my bike was in the shed, it would be reasonable to expect that I have very good reason for the belief, such as that I was watching a live-feed camera showing the bike in the shed, etc. But if I merely claim to believe that it is in the shed, I need something less than that. I don’t need to convince a jury beyond a reasonable doubt to justify a belief in a proposition. I need some justification, but it needs to fit my confidence in the claim. So I believe the bike is in the shed because I remember putting it in there last night, I haven’t heard any noises that sound like bike thieves, I know that the crime level is low, etc. These are good reasons for having the belief. They justify the belief, even though they do not convince me. I have not “put myself on a battlefield I cannot win” by making the claim. Winning means having a reason that is proportionate to your degree of belief, not ‘being able to convince my interlocutor beyond a reasonable doubt’. So I just think Matt is wrong here. It isn’t a battlefield we cannot win on. If you believe god doesn’t exist, you can make that claim. You need something to justify it (to be rational), but you don’t need to convince me, and you don’t need to convince yourself.

Part of the reason for the seeming slide towards epistemic extremism may be simply sloppy presentation on the live show. We have all misspoken before, of course. But part of it seems to me like it might be caused by the apologetical atmosphere in general. An apologist has a very ambitious goal in mind, most of the time. They are not just defending the rationality of their beliefs, but actively trying to persuade non-believers. If your goal is to persuade me to change my mind about the truth of p, you need to have a very good justification for thinking that p is true. You need to have a better justification than you do to justify the claim that you merely believe that p.

Yet, we can see the epistemic extremism on display in the Atheist Experience in this episode as a conflation of these two different standards. Ethan can claim to believe that god does not exist, and he needs to have something to say about why he has this belief for it to be rational. But he does not have to produce the level of evidence that would be required to convince someone else to believe. These things are distinct.