5626463547 | San Jose, CA 95126

Inquiring about Tipping Point Evidence

Two colleagues on a team are debating with each other over whether to use a certain type of database in a current project. One team member insists that a certain query can be performed more quickly if data were stored in a NoSQL repository. Another disagrees. After being tired from not being able to convince each other, one member asks the other: “What would it take for you to change your mind?” The other responds, “You would have to show that the average time over 100,000 queries is faster than x amount.”

A mother and daughter are arguing over whether the husband/father had an affair. The mother is financially dependent on the husband and who has been the victim of her husband’s other past affairs. She fears that she cannot make it through one more experience of this. Her daughter, however, is financially independent and believes that the mother’s staying with her father is going to ruin her mother. The mother insists she has no reason to leave because her husband did not have an affair. The daughter insists that she has given plenty of evidence and still her mother does not believe her. So she asks, “What would it take for you to change your mind?”

Call the question, “What would it take for you to change your mind?”, the tipping point question. (TPQ)

Conversations that include the TPQ are common. It is not the first thing asked in a conversation, at least not usually. Why not? My take is that agents who engage in a conversation with each other see each other as rational agents. Even in scenarios where one agent aims to embarrass or shame the other intellectually, one at least has to see the other as sufficiently rational to understand when one has been intellectually pwned. Of course, anyone who has the slightest degree of social sense recognizes that not all of us come away believing the same thing about just any propositions. We all know about bias, disbelief, epistemic stubbornness, intellectual laziness, wishful thinking, and so on. Nevertheless, we know that despite that we each have capacities for rational belief, hard though it is to realize those capacities in every scenario. If that’s correct – that in our conversations we see each other as rational agents – then there is the presumption that one is generally responsive to evidence in the way that leads to believing the truth. But if two persons reach the point where another has not been swayed by any of one’s arguments, then arises the question of what it would take to change one’s mind. Someone might suspect that one disagrees due to wishful thinking, say; but unless one knows that about the other, there is still the presumption that one is conversing with a rational agent. The question may be a nod to finding a way to move the conversation forward; it may be a way of testing whether the other agent is in fact not being rational because of wishful thinking; it is hard to tell. But the question can have both uses and that’s why it arises when it does, that is, when it is asked at that point in the conversation.  I do not think this is the only use of the question.

In a 2012 Reasonable Faith podcast, William Lane Craig responds to a TPQ concerning evidence and atheism. The transcript is available, which is my source of the quotes. The TPQ comes from a philosophy student who wrote,

Dear Dr. Craig, my question is simply put: what would it take for you to become an atheist? Do you have a limit on how much evidence it would take for you to abandon belief in God? I ask this question because I often find both atheists and theists claim that if there was good evidence they would convert to the other set of beliefs, but never clarify what it would take for them to convert, which seems to be a key issue if one is to persuade another to their point of view. Thus I think answering this question would be helpful to budding apologists out there.

I might be wrong about this, but the very last sentence may an indication of this student’s interest in apologetics. Given that interest, it is more likely that the person is a theist. Nothing else in the quote indicates one way or another what the student’s own view is about atheism or theism. It would be forgivable to think that Craig would then have something positive to say here. But it were not meant to be. He crushes this question and goes on to explain how confused those persons are who ask the TPQ.

Craig writes,

That [TPQ concerning atheism] is a question about a person’s personal psychology. And that’s just irrelevant to the worth of the arguments and the evidence that one is offering in favor of one’s point of view. The question about what would it take for you to become an atheist is an autobiographical fact that is of no philosophical significance, or evidential significance.

The question has no significance, philosophically or evidentially; facts about personal psychology do not provide data relevant to answering questions about theism or atheism. Why not? The answer is not that contingent truths cannot be evidence of necessary truths. That has obvious counterexamples. Craig’s initial answer, rather, is that psychological autobiographical facts provide no guidance about whether, say, the universe was designed. Autobiography doesn’t render an argument invalid or a premise more likely to be false (setting aside disputes over a person’s psychology). There seems to be no clear connection between first person reports of what one would believe under certain circumstances and the truth about the proposition believed. Someone might have no idea how to answer the first but still be able to identify evidence for the latter. At the point where Craig directly answers the question, he says “I have no idea about what sort of things would lead me to become an atheist.” In fact, it is not just that Craig happens to not know what the tipping point is. In addition, “I don’t think they’re answerable.” My most charitable interpretation of Craig here is that he is restricting the scope to himself; he cannot answer those questions, or that specific question regarding atheism, about himself.

The reason I think that is the most charitable interpretation of Craig is that it is possible that some people can know about what sort of evidentiary considerations are the most important to them. And so if they can apply those considerations to the topic in which the TPQ arises, they can provide an answer to the TPQ. For example, if the philosophical question is, “what would it take for me to believe that no one is justified in believing anything,” being asked by a radical skeptic to me, I have no idea what the answer is. But if the question is, “what would it take for me to believe that incompatibilism is true,” I have an answer to that question.

It is interesting how cohost Kevin Harris responds. It is unclear whether Harris is persuaded because he notes his own use of that question.

I will say this, quickly,… I have found that it is very personal, especially on the atheist side because that’s usually what I am asking. “Well, what kind of evidence are you looking for? What would serve as a good criteria for you?” It’s usually things that fall very short of what God has already done. It’s things like, well, if God would levitate that object, if God would spell out Yahweh in the stars, and things like that. And all of those things, Bill, could be done by alien life, hallucination, all of those things fall short of things that just can’t be faked, like the Big Bang and the creation of the universe. I find the criteria very silly sometimes, very limited, and not getting a bigger ontological picture.

Craig did not take Harris to be disagreeing because he adds,

Yeah, I mean, let’s be honest; I think that these sorts of criteria, Kevin, are usually self-engineered to protect or insulate one’s own beliefs so that one won’t have to change. One imagines that these conditions will never obtain and therefore one won’t be called upon to change one’s beliefs. But this is all just psychology.[1] It is a way of avoiding the arguments and looking at the premises for the arguments and the evidence for those premises.

Perhaps Craig took Harris to be confessing a confusion. But what Harris disparages are the answers to the question by some atheists, not the question itself. Craig, however, moves from the ridiculousness of the answers to the question being suspect of some underlying motive and back to the irrelevance of psychology. The TPQ is philosophically insignificant, evidentially irrelevant to the the matter about which the TPQ arises, and is likely to be a way of maintaining disbelief. Craig ends, “Even if I am utterly closed-minded and would never change my mind, that gives the unbeliever no grounds whatsoever for ignoring the evidence and arguments that I offer.”

It is unsurprising that I am not at all persuaded given how I set this up. Of course I am not making any claims about whether Craig really knows or does not know what it would take for him to change his view. I do not know the man. He has overstated the epistemic confusion behind this question because his interpretation of what motivates the question is one-dimensional. In that dimension, Craig is of course correct that these autobiographical facts usually provide no evidence for or against the truth of some philosophically substantive claim. But TPQs need not be asked in that way.

I already suggested two ways TPQs can be relevant to the conversation partners.

(1) A TPQ can be a way of moving the conversation forward by identifying considerations perhaps not yet discussed. We often believe substantive philosophical claims because of a collection of beliefs about related matters. Craig’s arguments for theism are many; it is very unclear whether he would remain a theist if he were convinced that the meaning of life did not depend on immortality, that a premise of the Kalam argument were false or unjustified, that the historical evidence for the resurrection of Jesus did not even meet the minimal facts criteria, and all he had was his report about the Spirit’s internal testimony and Plantinga’s ontological argument. An atheist might choose to then start picking away at one of these other topics.

Similarly, consider a conversation between a theist and atheist. Let the atheist be J.L. Mackie. A theist might have asked Mackie what it would take for him to be a theist. If Mackie answered that he would need to accept moral values exist, the theist might agree that the objectivity of moral values is evidence for God, and start finding a way to argue down that path. Doing this does not imply one thinks being a moral objectivist is sufficient for someone, e.g., Mackie, to accept theism. After all, he has a whole book on the problem of evil. But the answer to the TPQ does help guide the conversation.

(2) A TPQ might reveal an answer that is evidence of another’s bias or irrationality. This is suggestive of the second example with which I opened, and it is suggested in Harris’s description of answers to the question. Why should any of this matter to the discussion over philosophical topics like whether God exists? That is, why should we be interested in TPQs over such discussions? I think we can see the answer by considering the question asked to Craig by the student. There is the concern that theists and atheists might be arguing in bad faith, that one might engage in an argument with a person who will hold a belief no matter what the evidence is. Such persons are unlikely to have epistemic policies that arrive at truth, especially where it counts. Normally we disqualify listening to people like this or wasting time trying to convince someone by presentation of rational arguments. But Craig says that even if he were he like this, we should still listen to him. That is fundamentally confused. The value of the TPQ as it pertains to finding evidence for or against some philosophical question does not settle all the value the TPQ has as it pertains who we choose to engage with in intellectual disputes. Moreover, if we are to take expressions of religious experience seriously (e.g.,. a personal confession about one’s experience with one’s Lord and Savior Jesus Christ, at the end of a debate about God’s existence, say), the TPQ is relevant to deciding how much credence to put into another’s confession. 

I’d like to end with an illustration of a contrast. The contrast is between two separate conversations, where Robert Lawrence Kuhn asks two different theists why he should believe in God. These conversations were recorded for the show, Closer to Truth, which has many excellent interviews and much interesting content. Pay careful attention to how Kuhn responds in each conversation.

The first is his conversation with Craig. Actually, Craig does not get asked exactly the question of why Kuhn should believe in God. He is asked about whether God’s existence is demonstrable. But how Craig answers the question is so similar to how he addresses the question of whether God exists (e.g. in his debates) that I think the comparison is appropriate.

Now watch this interview with Sarah Coakley.

I started by saying the TPQ is asked at the end of a conversation, at least usually. In apologetics these days, the context of desire is largely ignored. It is about memorizing these arguments. Notice, however, that Coakley does not dismiss them. She insists that she would defend them against the trend in theology (departments) to dismiss their epistemic importance. But she starts with the TPQ. Does it make a difference to how Kuhn responds? It seems to me at the very least, Kuhn is at least intrigued and listening in this second conversation. What Coakley is doing is difficult; it involves being able to read other people and have the wisdom to know what to say, when to say it, and how to say it. You cannot learn this by memorizing arguments from an apologetics program.

Are questions about psychology philosophically or evidentially irrelevant? And worse, are they indicative of some epistemic-defense mechanism? To the latter: surely not. Coakley is not asking that to maintain her theism in the absence of arguments or more powerful arguments against theism. To the former: it depends on what is salient. Is the question asked in a context where another’s rationality is open to question? Perhaps the person is being asked about what motivates certain beliefs?  I find it hard to believe that anyone thinks that God’s existence depends upon one’s psychological autobiographical makeup, as if that is what is motivating the TPQ. The student’s question, or tipping point questions, can be significant as it pertains to who we engage and to our own reflection about how we evaluate arguments. Sometimes we don’t know answers to TPQs. But we should not be dismissive of them.

2018-06-10T14:15:08+00:00 June 8th, 2018|6 Comments


  1. Oscar Klein June 9, 2018 at 9:24 pm

    Hi James. Thanks for your posts, I really enjoy and benefit much from them. Since you addressed the topic of TPQ’s, where do you stand on apologetic methodology? Vantillian Presuppositionalism or Evidentialism?

  2. James A. Gibson June 9, 2018 at 10:48 pm

    Hi Oscar. I would say that I identify with neither camp, although I used to identify as a presuppositionalist (of the CVT variety). I certainly do not think I am presupposing Christianity when I discuss with, say, an atheist. Nor do I think I have the obligation to start thinking of it that way in those sorts of discussions. But neither am I prepared to agree with the evidentialist that belief in God must depend on evidence in order to be rational. For a more positive description, like Coakley I think that issues of how we evaluate evidence and arguments are shaped in part by the motives of the heart; I’d also nod to influence of William Wainwright on me. Given that, you might be able to see why I look to coherence or larger worldview considerations as well as think that discussion of values matters. But also I think there are arguments that do make it more probable than not that God exists or that Christianity is true or whatever. In my view there is nothing intrinsically wrong with giving an argument that only renders God’s existence more probable than not on some evidence. On reflection, it looks like my view is quite a lot like Craig’s! Maybe that is right. But if that is right, I don’t think that the way in which I present the arguments is stylistically the same as his. We can’t all be masterful debaters and polished. And the Kalam is not my goto – although it is entirely appropriate for Craig given how much work he spent on that argument. He and I also have too many other theological and philosophical differences that the exact moves we make at various points (e.g., on the problem of evil) are different. To my mind, I do not come across at all writing like an apologist out of Biola, although I did attend there, fwiw.

  3. Oscar Klein June 10, 2018 at 11:34 am

    Thanks for your reply, James. You make a lot of good points. However, I would say it’s counter-intuitive to see Calvinists embrace the kind of evidentialism espoused by the likes of Craig (even if it’s not the same style). It seems to me that there is a major methodological discontinuity happening when Calvinists make use of that method, and it becomes plain when one starts talking about the role of sin in epistemological systems of thought. Guillaume Bignon makes the same argument as well — that you can be somewhat in the middle of it… take the good of each and leave out the inconsistencies. I do not see any space for neutrality in this area, however, just like there is none in terms of other basic epistemological and metaphysical issues. Take, for instance, the argument of the existence of God being more probable than not. If that is it what it takes to properly appeal to the atheist or anti-theist, then there is absolutely no grounds for saying we actually know God exists. As an example, let’s take the next presidential election: what if Donald Trump’s being president is more probable than not (51%). Would I truly hinge my life on that? Of course not. I am not properly warranted to be able to hinge my life on something that has even a chance of being wrong. As Kevin Scharp has pointed out, though it might be GREAT that God’s existence is more probable than not, I still don’t know God and don’t really believe he exists, epistemologically speaking. My argument should be consistent through and through. But there is a reason Craig refuses to be consistent with his argumentation: He is tethered to an Orthodox understanding of Christianity, where the bible is ultimately preeminent as our epistemological foundation. He WOULD say, “Therefore, God most probably exists”, if he was consistent in his argumentation. But he doesn’t do that. Which makes his public debates all the more insonsistent. It took the atheist Kevin Scharp to bring that out to light. I highly recommend Craig’s debate with Kevin Scharp:


    I also recommend K Scott Oliphint’s essays on (and which are quite sympathetic but critical of) Alvin Plantinga’s Religious Epistemology (below is one of them):


    Again, I know you yourself may have a different style than Craig, but I still see no neutral landscape possible on this issue, at least not in its basic principle and methodology. I of course don’t belive that evidence is wrong (as most Van Tillians would heatily acknowldge), but there are ways that you can use them and still hold to a consistently reformed and biblical epistemology. And that’s evident in other areas. You hold to a consistently reformed epistemology with respect to the problem of evil. You don’t give in to libertarian assumptions. Libertarians acuse you of eschewing freedom and responsibility for exhaustive devine sovereignty. But, as every good Compatiblists, you question the libertarian presuppositions behind their view of freedom. That is exactly what Van Tillians argue with respect to epistemological methodology. An example of a good Van Tillian would be Douglas Wilson, for instance, who did a great job debating Christopher Hitchens. Thanks again for all your great content.

  4. James A. Gibson June 10, 2018 at 1:38 pm

    OK. So when you see me give an argument where either (1) a premise is false; or (2) a premise is unjustified; or (3) a conclusion is not appropriately supported by the premises, then challenge me on that. But if you don’t got (1), (2), or (3) to raise against an argument for which I am persuaded (or at least inclined to think is pretty good), then I see those points as irrelevant to what I should think about those arguments. I guess the exception to this is that if someone were to convince me that there were some performative error in *expressing* the argument, I’d be open to hearing that. Here’s a case: either I refuse to give the Kalam argument to one person or a class of 300 students will get lit up. In that case, the right thing to do is to not give the argument to that one person, even though the premises might all be true and the conclusion supported by the premises. I don’t think presuppositionalists have anything close to a compelling reason like that against expressing an argument (e.g., Kalam). Anyway, that’s all I want to say about this topic. Cheers.

  5. Oscar Klein June 10, 2018 at 2:41 pm

    Cheers! Thanks for the correspondence.

Comments are closed.