Red Button, Blue Button

You find yourself teleported in a room. There are two tables, one with a red button and one with a blue button. Above the red button is a sign that says, in whatever language you are most fluent in, MURDER BUTTON. The same for the blue button, except that sign says SUICIDE BUTTON.

 

In whatever language you are most fluent in, a sign between the buttons reads the following:

 

Hello! I am a Djinn who has teleported you into this dream world. Don't worry, time outside has been stopped. All other able-minded adults in the entire world aged 18 and up have also been teleported into this dream world at this time.

 

If 50% or more of participants press the red button, all participants who press the blue button will be killed. So, if you press the red button, you are guaranteed to survive, but you might be guilty of murdering however many people press blue.

 

If 50% or more of participants press the blue button, all participants will live. You are guaranteed to be blameless of participating in anyone's death, but you might die.

 

Which button will you press? 

 

For those who are thinking of pressing red, please consider: It's statistically guaranteed that at least a small number of people will choose blue. So if red wins, you will be guilty of killing these people.

 

When pressing the red button, you might disagree that this is the MURDER BUTTON. After all, those who choose blue are willing to die; they are killing themselves. The rational choice is to guarantee your survival, and if everyone chooses red, everyone survives. Who cares if a few irrational people die? If irrational people take themselves out of the gene pool, isn't that a win for humanity? "I'm not killing anyone by pressing red; the Djinn is the real killer here! I didn't choose to be here. I'm just guaranteeing my survival, which is the rational thing to do, and everyone else has just as easy access to their own survival as I do, so it's their own fault for choosing to risk their lives for no reason." So, you might figure that pressing the red button does not constitute murder, or even homicide.

 

For those who are thinking of pressing blue, please consider: You may trust your own culture to choose blue with you. But how confident are you that most humans in the world will choose blue? Won't most people follow the above logic about self-preservation and rationality? Maybe only a small percentage will press blue, and your vote never would have made a difference, and you will have died for nothing.

 

When pressing the blue button, you might disagree that this is the SUICIDE BUTTON. After all, if only 50% or more choose blue, everyone survives; no suicide here. Most importantly, you know that some people will press blue to ensure their own blamelessness, guaranteeing some blue votes. You are risking your life to save these people by pressing blue. Isn't that the noble thing to do? Who could live with themselves knowing that they helped cause the deaths of those noble enough to press blue? If red barely wins, nearly half of everyone will die. Are you willing to risk this?

 

I will give all of you one hour to decide which button you will press. At the end of the hour all of your fates will be decided. During this hour, if you press one button, you can press the other button to change your vote. If neither button is depressed within the hour, you die, and no vote will be cast on your behalf. Any deaths will be instant and painless.

 

***

 

It's interesting to see this thought experiment articulated in different ways, and to see on Twitter how different articulations can yield different results. With this articulation, I'd press blue, but not necessarily because I am a "good person", whatever that means, but because doing the blameless thing makes me happy. Participating in the deaths of blue voters would make me lose self-respect, and while red voters would call this irrational, because you shouldn't feel bad for those who take unnecessary risks, their arguments are weak enough that I don't see my mind changing soon. For example, voting blue is a necessary risk if feeling blameless is very important to you. Arguably, voting red is the unnecessary risk – of becoming an accomplice to mass murder.

 

Additionally, voting red incentivizes you to take on a callous attitude towards dead blue voters to cope with your choice, exposing either your own moral insecurity or indifference. I prefer to avoid that.

 

One final note: Some folks want to apply expected value calculations to show that pressing red is the correct choice. But if these calculations fail to incorporate the value of being happy with oneself, the cost of self-hatred, and the cost of lives lost, then they are incomplete.

Florence Bacus on grounding value in conscious states and on what morality is

https://www.youtube.com/watch?v=0zjflulzeAU

 

What morality is:

 

28m: ". . . without God, what is morality? So I think morality is just rational action independent of desire. . . . There are times where the reason I ought to do something is because it achieves some further end, but then there are cases, at least if morality is real, where you ought to do something regardless of whether you antecedently desire that end."

 

29m: "If you understand morality as something else, then okay that's fine, you're answering a different question from me. But I find myself in cases where I do various things just because I want to get what I want, and then I wonder are there cases where I should act a certain way even if it doesn't get me what I want in any respect. So if you're gonna tell me that I ought to do something, either you're telling me I can get what I want—and then I would say yeah, that's an important question, but it's not morality—or you might say I ought to do something regardless of what I want, and I would say okay yeah, I want to figure out what that is, that's what I call morality."

 

Florence goes on to accept a form of Kantianism. I reject Kantianism quite thoroughly, so it's no surprise that my value theory would differ from Florence's value theory as discussed in the next section.

 

I don't think there can be a categorical imperative, so the core questions of morality are different for me. People, including myself, tend to live their lives believing certain things about themselves, like that they are a good person, that their beliefs are true, and so on. Part of one's self-conception is that these beliefs that one has about the world and oneself are not totally, radically incorrect. Here's the problem: Your beliefs about the world and yourself are totally, radically incorrect. At least, they might be. Mormons, Protestants, Catholics, Muslims, and naturalists cannot all be right. If naturalism is true, then the beliefs of Mormons, Protestants, Catholics, and Muslims are profoundly false. People don't want to hold profoundly false beliefs about the world and themselves. They want to believe what's true. That will include truths about the nature of morality and what it means to be a good person.

 

Many people, myself included, and arguably all people capable of basic rationality, hold the following desire within themselves: "I want to believe what's true about morality. I want to believe what's true about right and wrong, so that I can do what's right and not do what's wrong. I want to desire what I ought to desire, and not desire what I ought not. I want to know what it means to be a good person so that I can become that kind of person."

 

This changes the framework: Morality is not about being rationally obligated to do something irrespective of what I want, but morality is, just like epistemology, about reconciling my beliefs to reality. If I believe rightly, I will desire and act rightly. Morality is not about acting against my desires for the sake of the greater good, but about adjusting my desires to be aligned with what the greater good actually is. To suggest that someone could have all true beliefs but still act wrongly is to suggest that truth and morality come apart, and I don't believe that. If I behave within the confines of truth, then I cannot be blamed, lest truth itself is somehow evil. (I have encountered philosophers who think truth and morality can come apart. Someone with all true beliefs could still lack the virtue to act according to those beliefs, or by akrasia do what's wrong despite knowing what's right. I accept the Socratic rejection of akrasia.)

 

(Feel free to replace 'greater good' for something more deontic sounding, like: for the sake of acting from a good will, or promoting the kingdom of ends, or treating people as ends unto themselves, or acting freely, or fulfilling obligations, or doing what I am rationally obligated to do, or doing what's right because it is right, or for the sake of moral goodness – basically, for the sake of right action broadly speaking: Morality is not about acting against my desires for the sake of right action, but about adjusting my desires to be aligned with what right action actually is.)

 

A potential challenge to my view falls along these lines: If morality is all about maximizing goodness, then doesn't morality fall apart at the point of survival? We don't blame predators for killing prey because their lives depend on it. Can we say the same for us? When Nazis murdered Jewish people and defended themselves by saying "We were just following orders", they apparently have in mind something like: "If I didn't follow orders, I would have been killed, and you can never blame a creature for doing whatever it takes to survive." You can tie this into desire in this way: Everyone has a desire to live; no amount of "reconciling your beliefs to reality" will change that. So if morality, to be a meaningful authority over us, requires us at some point to give up our lives for the greater good, then we will be acting against our desires.

 

I reject this and say yes, even the desire to live falls away when greater goods are recognized. Parents don't think twice in risking their lives for their children. In movies we often see characters sacrifice themselves to buy time for the rest of their group. If I realize the truth that my happiness is no more intrinsically valuable than the happiness of others, and thus come to have the goal of maximizing happiness period (not just my own), then at some point my own survival can become an obstacle to my goal if my survival requires the suffering of others. So we see that my consequentialist, "true belief first" approach to morality can accommodate the truth that for morality to be "authoritative" in some sense, morality cannot break down at the point of survival; at some point choosing your own survival over the happiness of others is evil.

 

Grounding value in conscious states: 

 

22m: "If we're going to assess the realist view that grounds [value] in the intrinsic goodness of states of consciousness, and I want to say no, [things are] only good because people prefer them, we'd want to look at a case where the two come apart. So, a case where a state of consciousness is good, but a person doesn't prefer it, or a state of consciousness is bad but a person doesn't disprefer it."

 

When Florence uses the term 'good' or 'bad', I don't know what these are supposed to mean. If these mean 'preferred' and 'dispreferred', then it's a tautology to say that good experiences are preferred experiences. Maybe the question is: Is there any analysis of good/bad on which a good experience can be dispreferred?

 

Yes. In the phenomenal analysis of good/bad where goodness/badness just are conscious states, there are cases where a state of consciousness is pro tanto good (good to a degree), but a person doesn't prefer that state because obtaining that state entails a cost that a different state doesn't.

 

There's always pro tanto desire for intrinsically good experiences, but there's also always pro tanto desire to avoid intrinsically bad experiences. There's also a pro tanto desire to obtain intrinsically better experiences; do I stay home and do something safe and comfortable, or do I go to the party and potentially have greater fun at the risk of greater discomfort?

 

So all of these calculations are warring within us, and this is why it's easy to make sense of avoidance behavior in response to something we know will feel good, and affective behavior in response to something we know will feel bad.

 

Avoidance behavior + intrinsic goodness: There are things we know will feel good but will make our lives worse in the long run. We see the intrinsic goodness and thus have a pro tanto desire for it, but we also see the cost and have a pro tanto desire to avoid that cost. Maybe this is eating junk food, procrastinating, smoking a cigarette, cheating on a partner, or getting back with an ex. Despite whatever intrinsic goods these things will provide in the moment, we avoid them when we do because of the greater costs down the road.

 

Affective behavior + intrinsic badness: There are difficult conversations to be had with ourselves or with a loved one, and that pain motivates us to avoid the conversation, but we see how things will be better over all if we suck it up and push through it. Applies also to chores, self-improvement, working towards difficult and personally meaningful achievements, and undergoing medical procedures – these things can be greatly painful but we do them anyway because not doing them is even more painful in the long run.

 

Paradoxically then, I might prefer to have a difficult conversation than to eat junk food, because I know one is worth the pain while the other isn't worth the happiness. And yet the pain of having a difficult conversation is still intrinsically bad despite preferring it to eating junk food, and the happiness of eating junk food is still intrinsically good despite dispreferring it to having a difficult conversation.

 

Indeed, I don't see how to make sense of the concept of preference without first going through the intrinsic goodness / badness of experience. To prefer X over Y is to see how X is better than Y, and to see that X is better than Y is to see that X is more good than Y (however good Y happens to be). If goodness just is preference, then this is circular. But this definition of preference appears non-circular to me, which tells me that goodness isn't just preference.

 

If goodness is defined in terms of desirability, and to prefer X over Y is to have greater desire for X than Y, then I don't see how to make sense of the concept of desire without first going through the intrinsic goodness / badness of experience. To desire X is not to see X as desirable, as that's a circular definition, but to see X as good. Why do we desire? Because we experience intrinsic goodness and intrinsic badness, and we desire intrinsic goodness because of what it is, and we desire to avoid intrinsic badness because of what it is. It is only because intrinsic goodness / badness have the essences they do that explains how it could be that desire exists at all.

 

Returning to preference, I don't see how to explain why my preference would be what it is without intrinsic goodness and badness. Experiences happen first, then preferences later. So preference depends on experience. But preference also depends on what I see as good or bad. So preferences depend on my good and bad experiences.

 

You might try to say that my preferences depend on what I take to be my good or bad experiences, but this just kicks the can down the road a step. Why would I take something to be good or bad? Because of the intrinsic goodness or badness that this thing constitutes or causes. There is a difference between experiencing badness and judging something to be bad. Experiences come first, and preferences and judgments come later based on experiences. Desires qua judgments of goodness come later, but desire as affective intending is simultaneous with intrinsically good experience; you can't have a good feeling without at the same time having an affective intentionality towards that feeling, meaning having a pro tanto desire for, like, conation, motivation for, or intending toward that feeling. As discussed here (https://benstowell.blogspot.com/2026/03/there-is-conceptual-connection-between.html), while affective behavior and intrinsically good experience can come apart, affective intentionality or intending toward is conceptually connected to intrinsic goodness. We intend to pursue happiness. Why? Because of what happiness is, and that intention begins the same moment the happiness is experienced.

 

P.S. - Florence mentions becoming an ex-Christian as of a year ago (so roughly March 2025). Being an ex-Christian myself, starting sometime in early-ish 2022, I understand what it's like to be a Christian for a long time and to feel the internal violence of being gutted of your entire worldview, being left to pick up the pieces. It's not easy, and it's traumatic, and I feel great sympathy for anyone who goes through that.

Matt Duncan on certainty

https://www.youtube.com/watch?v=P_Odt7Mzu-4

 

32m: Duncan makes a distinction between a psychological sense of doubt and an epistemic sense of doubt. The more epistemic justification you have (whatever that is), then the less psychological doubt you should have.

 

"And I distinguish that psychological sense of doubt with this kind of epistemic sense of doubt, where it's something like you've got a very strong form of justification or something – lot of different ways to spell that out . . . 

 

Duncan goes on to discuss when we can be certain:   

 

. . . the way I prefer is that it's immune from skeptical doubt, where that means that there's no skeptical scenario consistent with the way things seems to you where the relevant thing doesn't exist."

 

34m: "[I] argue ultimately that the phenomenal properties of our experiences pass the doubt test. I mean, this is a familiar idea, you know, Descartes, I can't doubt that I exist . . . I spend a lot of Chapter 3 unpacking the epistemology there. . . there are two routes to doubt here. One is lack of acquaintance . . . But there's this other route, phenomenal indistinguishability, that still leaves room for some kind of doubt, and that's why I think acquaintance itself doesn't foreclose doubt. Now in the case of phenomenal properties, I think both routes are foreclosed. There's the acquaintance route – foreclosed – and then there's the phenomenal indistinguishability, and the reason why that route is foreclosed here is because phenomenal properties are defined and individuated by the way they seem, the way they appear. And so there couldn't be any skeptical scenario in which things appear like this and yet your different phenomenal properties are instantiated in your experience because it would just follow that things wouldn't seem like this, they would seem at least a little bit different. . . . for phenomenal properties, there really is no skeptical scenario consistent with the way things seem in which those aren't instantiated in your experience. It's really phenomenal acquaintance where we get full certainty, because both of these routes to doubt are foreclosed."

Peter Unger, "An Argument for Skepticism" – is metacertainty necessarily dogmatic?

In a previous post on certainty, I discussed Peter Unger's challenge that certainty is dogmatic. I offered a view that allows me to capture both certainty and open-mindedness. But here's a point in Unger's favor: 

 

You might argue that to assert something as being certainly true, or to assert that something certainly exists, you must also therefore assert that your analysis of what it means for something to be true is correct, or that your analysis of what it means for something to exist is correct. But you probably don't have, and shouldn't have, this further certainty, given the widespread disagreement among professionals over what it means for something to be true or to exist.

 

My solution before applies here too: by saying "I can't imagine that being false" or "I can't imagine that not existing", I am open to the possibility that it's either possible or impossible that someone could successfully imagine the truth or existence of the thing in question. If it turns out that someone can imagine the truth or existence of the thing in question, then one explanation for why they can imagine this when I can't is exactly because they have the right analysis of truth or existence when I don't.

 

But this works both ways: it may be impossible for anyone to successfully imagine the truth or existence of the thing question, and part of the explanation for this could be because the correct analysis of truth or existence entails the truth or existence of the thing in question.

Liz Jackson on Romans 1

Liz Jackson writes ("Permissivism About Religious Belief", https://philpapers.org/rec/JACPAR-11):

 

In Romans 1:20, St. Paul writes, “For since the creation of the world God’s invisible qualities—God’s eternal power and divine nature—have been clearly seen, being understood from what has been made, so that people are without excuse.” Given passages like this one, those in the Christian tradition might argue that theism is not a permissive case. Paul seems to be arguing God is clearly seen from creation (and in a later passage, the moral law). From this, some conclude that theistic belief is rationally required of most or all of us, given the evidence.

 

In reply, my arguments in this chapter are consistent with a relatively straightforward reading of Paul in this passage (and similar passages). In Paul’s time, people had drastically different evidence that more clearly pointed toward theism. When Paul was writing, belief in God and/or the supernatural was extremely common and widespread (perhaps the divine hiddenness problem wouldn’t have gained much traction in the first century!) People spent much more time in nature, staring at starry skies and breathtaking sunsets; today, much more time is spent walking around concrete jungles and staring at computer screens. Furthermore, as Charles Taylor (2017) argues, the Reformation, Enlightenment, and Romanticism involved deep cultural and intellectual transformations that gave rise to a much more secular society. My arguments in this chapter regard our current evidential situation, and while I maintain that permissivism about theism is supported by the widespread historical disagreement about God’s existence, I don’t claim that permissivism about theistic belief extends to all historical periods, places, and times.

 

I see a few issues with this response to Romans 1. First, Christians usually believe in some form of inerrancy and inspiration: Because the Bible is inspired by the Holy Spirit, anything it teaches to be true is true. (I acknowledge Randal Rauser's thesis of providential errantism and set that aside.) Combine this with the interpretation that Paul is teaching us that those with non-belief are without excuse and you arrive at the interpretation that this is an inspired and therefore inerrant teaching of the Bible.

 

Even if Paul is saying something from the point of view of his time and place, the Holy Spirit can inspire teachings that are applicable well into the future, and modern Christians usually see themselves as reading something that was in a sense meant for them. Very few Christians, in fact this is probably a heresy, see the Bible as a fossil or snapshot of a particular time and place—this would basically be a rejection of inspiration.

 

There's two aspects of the Holy Spirit view of scripture: The Holy Spirit guides the authors of the Bible to write from their own perspective in a way that includes layers of meaning and timeless truths applicable to future generations, and the Holy Spirit guides the modern Christian to the knowledge that a particular interpretation of a passage applies to the modern Christian and their life despite the gap in time and culture. Viewing the Bible as a fossil would mean to jettison both aspects of the Holy Spirit view of scripture. So a very natural and Christian way of interpreting Romans 1 is that its teaching applies for all time.

 

Second, Paul says since the creation of the world . . . from what has been made . . . This kind of language has a universality to it. We look at the time interval from the creation of the world to Paul's time; Paul says God's attributes have been obvious in that interval. It would be strange for God's attributes to be obvious for such a long interval, but suddenly change in the comparatively negligible interval of Paul's time to ours. (And sure, Paul's phrasing of "since creation" need not be literal, but this applies even if we think of it as "since human civilization" or "since humans began believing in gods".)

 

Plus, if God's attributes are made obvious by "what has been made", then there is no relevant difference from Paul's time to ours. Paul is surely citing objects of creation like the starry sky, the planets, the trees, the animals, the sun and moon, and humanity. We have all these today, but it's not at all obvious that any of this were made by a perfectly loving God. For some of us, it's painfully obvious that a perfectly loving God almost certainly does not exist because of what has been made: death, war, disease, murder, poverty, and so on. While Christians view these as accidents of sin, for some of us it's apparent that the systems of our world systematically produce these evils. This is perhaps most apparent with the evils of evolution, and the predation, starvation, parasitism, and mass amounts of death it took to create us.

 

If anything, because we have more rigorous philosophical arguments for God from fine-tuning, the inference from "what has been made" to God's existence should be more powerful than ever, and so all the more are modern humans without excuse in their non-belief in God. And yet we also have more rigorous philosophical arguments from evil, and for some of us the problem of evil trumps the argument from fine-tuning, and it's not a remotely close contest.

 

Third, what Paul says is arguably false even in his own time. For much of human history, humans believed in multiple gods, and didn't necessarily believe in a singular all-powerful, all-knowing, all-good creator. So it's just not true that most groups of people have clearly seen the workings of a singular God; many, maybe most, interpret creation as the workings of multiple gods.

 

Plus, even for Jewish people and Christians, who both believe in a singular God, they strongly disagree who this God is. Christians believe Jesus is God incarnate, while Jewish people strongly disagree. God's invisible qualities have been clearly seen – including the invisible qualities of begetting the Second Person of the Trinity or "spirating" the Third Person of the Trinity? Not at all.

 

Even Christians disagree among themselves, with some viewing the Holy Spirit as a person in some sense separate from the Father and Son, and others viewing the Holy Spirit as something more like a force or essence that is not separate from the Father or Son; or some viewing Jesus as purely man while others view Jesus as both man and divine. How many heresies were there in the first few centuries leading up to the Councils and Creeds?

 

Even if you extend the idea of God to the Greeks like Plato and Aristotle, it's obvious from the Greek texts that this conception of God is radically different from the God that Christians believe. So it's very misleading to say everyone in antiquity believed in God or gods as if there were anything resembling a consensus as to what that means.

 

If the conception of God or gods is radically different between cultures, and Christians are the only ones who got it right, then it's not true that since the creation of the world God's attributes have been known. It's only thanks to the events of the Old Testament and the recent events of the Gospels that we now know the most important attributes of God, namely that God incarnated and died for our sins and you must have faith in Jesus, the incarnation of the Father (if you are a binitarian) or the incarnation of the Second Person of the Trinity (if you are a trinitarian) or a man who is somehow special enough to die for our sins but whose righteousness only imputes to us if we have faith in him? (if you are a unitarian?). God's attributes are related to our salvation, and so the lack of consensus among Christians as to how salvation works means even Christians haven't seen God's attributes. Invisible qualities indeed. Besides, many Christians emphasize God's ineffability; no one has a mind great enough to grasp God to any serious degree, which just is to say that God's invisible qualities have not been clearly seen at all.

 

Because Christians made up a tiny minority in Paul's time, Paul must have had a broader understanding of God when orating what became Romans 1. But if that's right, then Paul is not speaking from this local perspective, but is speaking universally; God's eternal power and divine nature have been seen by everyone, so that everyone is without excuse. Why assume that Paul is making an exception for future generations? And if you're not assuming this, then you grant that Paul could be, for all you know, applying what he's saying to future generations too, and so you should at least be agnostic with respect to whether Romans 1 is committing Christians to impermissivism.

 

Regardless, because of the problem of evil, God's eternal power and divine nature, even in some broad sense, haven't been seen by everyone. If anything, what has been seen by everyone is either God's non-existence or God's permitting of evils that for the life of us we cannot figure out why he'd permit. And when Christians, or even ex-Christians like me, try to come up with plausible theodical goods, we fail, and in my case I'm more convinced than ever that there are evils—unbearable suffering—for which there could not possibly be justifying theodical goods.

 

But putting Paul's mistakes aside, the question is whether Romans 1 commits Christians for all time to an impermissivistic attitude about the rationality of belief in God, and there's an interpretation where yes, it does, and this interpretation appears at least as solid as any other.

A proof against God's existence from evil would explain the sociological data

I heard the recent story of Noelia Castillo, a woman who died in Spain at 25 years old from euthanasia. She was sexually assaulted multiple times, and became paralyzed after a suicide attempt. (https://www.cnn.com/2026/03/26/europe/euthanasia-spain-noelia-castillo-latam-intl)

My argument from unbearable suffering is a suggested proof against God's existence. Proofs either for or against God's existence do not fare well in the history of ideas; they always fail to generate any kind of consensus among experts that God either must or must not exist. So chasing after proofs about God might seem like a wild goose chase.

And yet here is reason to chase after proofs. There is a strong, deep, global intuition that there is a tension between God and the evils we see.

Someone might be tempted to cite global statistics on belief in God and show that this tension must not be that strong, because a significant percentage of humans believe in God. But this statistic does not show how many of these believers are nominal, how many will come to lose faith, how many are struggling with doubts caused by evil, how many are engaging in self-deception or believe because of the comfort these beliefs bring in the face of evil.

Plus, we see religiosity dying in Western countries, and we see an inverse relationship between level of education and religiosity. This is what we expect to see if it were the case that God does not exist and if belief in God were ultimately a superstition sustained by a lack of awareness of facts about our world, and a lack of careful, critical thought – things cured by education.

When we look to the intellectual world, we see much more widespread disbelief in God. Within the intellectual world, but also very much outside of it, we see this deep and encompassing force, like a gravity, that the problem of evil exudes, pushing down belief in God.

The question is: Why does this broad, deep, powerful intuition that God and evil are incompatible exist? There are two general answers: Either this force comes from a place of widespread confusion, or comes from the truth.

I doubt this Zeitgeist comes from a place of widespread confusion. One reason why is because of how intimately familiar we are of pain and happiness. We have vivid experiences of pain and happiness. Goodness and badness are constant, everyday, familiar concepts. Another sociological fact is found in the perceived failure of theodicies and the failure of responses to evil. Even theists often remark about how poor the responses to evil and divine hiddenness are, and acknowledge how there are evils that are resistant to go-to theodicies of free will and soul-making.

Why are things like this? Why is the problem of evil so compelling? Why are responses to the problem of evil so embarrassing?

Here's an explanation: Because the problem of evil proves that God does not exist, and this wide and deep gravity that pervades everything has as its source this proof, whatever and wherever it is.

Peter Unger, "An Argument for Skepticism"

Unger's general strategy in this paper is the following:

 

Knowledge > Certainty > Severe characterization > Unreasonable.

 

That is, knowledge claims are claims of certainty, and under the severe characterization of certainty, certainty is dogmatic and unreasonable. So knowledge claims are always unreasonable.

 

I might be inclined to accept that knowledge claims are claims of certainty, and to accept that certainty under the severe characterization is unreasonable. And yet I think knowledge claims are fine. So I must give something up, and one place to look is whether the severe characterization is accurate. Unger states:

 

". . . one's being absolutely certain of something involves one in having a certain severely negative attitude in the matter of whether that thing is so: the attitude that no new information, evidence or experience which one might ever have will be seriously considered by one to be at all relevant to any possible change in one's thinking in the matter."

 

There are two separate questions here: what certainty is and when it's okay to be certain. To the latter, we might say that if the denial of p entails p, then it's okay to be certain that p. And we might say that if p is necessarily true, then it's okay to be certain that p. But this doesn't answer the first question of what does it look like to be certain of something.

 

(I've heard the distinction made between 'epistemic certainty' and 'psychological certainty', but I'll set that aside. I'll also set aside questions about what it means to believe or to have credence.) 

 

Here's a less severe characterization of certainty: I am certain that p just in case I cannot imagine p being false. This doesn't entail a dogmatic attitude, because I am open to the idea that you could change my imagination regarding p if it's really the case that you can imagine p being false. In other words, I am open to my imagination becoming like yours.

 

Now, Unger has something to say in response to this: That I basically just admitted that I'm not absolutely certain of p. After all, if I were absolutely certain of p, then I should say that there's no chance of my imagination being changed such that I come to believe differently about p. If I say there is a possibility that my imagination could be transformed such that I now see that p is false, then I admit that there is some possibility of p being false, which is to admit that I'm not absolutely certain that p is true.

 

And this exposes the commonsense definition of certainty: to be certain that p is to believe that there is a zero percent chance of p being false, or P(~p) = 0.00.

 

But the idea of psychological probability feels vague in my mind. "There is a zero percent chance of p being false to me" – What does this mean? Maybe it means, among other things: I cannot imagine p being false.

 

My response is that there are many beliefs where apparently not only can I not imagine them being false, but no one can. So if my imagination were to be changed to be like that of anyone else's, I still would not be able to imagine p being false. So this gives me the best of both worlds: I keep both my certainty and my open-mindedness. By being certain that p, I admit nothing more than that I cannot imagine p being false, but I can be open to my imagination being changed if you can imagine p being false. If you can't, then it turns out that it's not possible for my imagination to change such that I now see that p is false, at least with respect to your imagination. If no one can imagine p being false, then it turns out that it's not possible for my imagination to change such that I now see that p is false with respect to anyone's imagination.

 

The key is that I'm not assuming that there really is a chance that p is false, and that someone out there sees that p is possibly false; for all I know, no one can, has, or ever will see that p is false exactly because p cannot be false. You can't know what's not true. So if it's not true that p is possibly false, then you can't know that p is possibly false. And if you can't know that p is possibly false, then you can't see that p is possibly false. And if you can't see that p is possibly false, then you can't imagine p being false. So if p cannot be false, then you cannot imagine p being false.

 

Side note:

 

***

 

I would apply this to belief in God. We might say, initially, that we can imagine both God existing and God not existing. But if we buy the logic of ontological arguments, it will turn out either that God must exist or cannot exist. So it will turn out that p cannot be false where p is either "God exists" or "God does not exist".

 

My response is that we cannot imagine God existing if God cannot exist, or we cannot imagine God not existing if God must exist. This is not strange; many people claim to know God on a personal level, or claim to know that God exists. If it turns out that God cannot exist, then these claims will be retrospectively mistaken, just like claims about the imaginability of God will be retrospectively mistaken if God turns out to be an impossible being.

 

To give an illustration of this, let's say there's a race of beings who think they can imagine a square circle existing or not existing, and without reason to believe one side over the other, they are agnostic about square circles. One day they realize that square circles are impossible, and so they realize that they actually weren't successfully imagining square circles existing; they just couldn't yet see the impossibility of them.

 

This is where the distinction between 'epistemic possibility' and 'metaphysical possibility' is brought up: the existence of a square circle was epistemically possible to these people even while metaphysically impossible. Likewise, God's existence or non-existence can be impossible without our epistemic access to that impossibility.

 

This suggests that imagination is a success term: it's not enough to take oneself to be imagining something; you have to succeed in imagining it in some sense.

 

And maybe this implies that there are layers of imaginability, because if God turns out to be impossible and thus unimaginable, then what were believers doing while believing in and thinking about God? Either they were imagining something that is possible and incorrectly calling that God, or they were imagining something impossible, but in some superficial sense.

 

I suspect that all objects of imagination are impossible, but this comes out of my suspicion that necessitarianism is true. Only objects of experience, which are actual, are possible. Even if you imagine something and it comes to pass, what you imagine is only a rough approximation of what really comes to pass; it never could be that your vision exactly comes to pass. Same with imagining things that you know happened but weren't there to see yourself, or even with things you remember. If you remember perfectly, then it's not an object of imagination, but of experience. If you remember imperfectly, then the imperfections are products of your imagination, and thus that past event didn't happen exactly like that.

 

One thing that pulls my intuition in this direction is the fact that our imagination never has the perfect fidelity that reality has, and so for reality to truly match perfectly with our imagination, it would have to be just as fuzzy and missing in the details, which is impossible. But if this is right, then while imagination might succeed in the sense of approximating reality, it never truly succeeds. But I guess imagination as a success term was never about accurately representing reality, but rather successfully combining properties in the mind. Incoherent things can never be successfully imagined exactly because the presence of one property means the lack of the other, so the two can never be combined in the mind. So imagination can be plenty successful in this way while still producing impossible objects and events. Incoherence needn't be the only source of impossibility; nomological impossibility is another. But even if all impossibility ends up being incoherence, incompatible properties could be successfully combined in the mind due to a failure to grasp the incompatibility. Right, it's not like God's impossibility is easy to see, lest the epistemic environment be completely different to what we see. The incompatibility is only visible, the failure of imagination only occurs, when the properties are explicitly contradictory. 

 

But doesn't belief depend on imagination? Are all our beliefs technically false in this case? No, because if we allow degrees of truth, then many beliefs of ours will be practically true. Setting degrees of truth to the side, propositional beliefs that contain nothing but first-person data would still be certainly true, as they are beliefs that map perfectly onto experience. If you say that beliefs that contain nothing but first-person properties must be non-propositional, and if you allow  non-propositional beliefs, then these beliefs would still be certainly true. 

 

Side note 2:

 

Because p's being false's being impossible entails that it cannot be imagined, my inability to imagine p being false is evidence that p cannot be false. This gets tricky if I have epistemic peers who say they can imagine p being false. If these peers are nowhere to be found, then quickly the best explanation for why we cannot imagine p being false becomes because p cannot be false.

 

*** 

 

I'm just giving a simple conditional: If there really is someone who can truly imagine p being false, then, and only then, could my imagination in theory become like theirs and I have a change of mind; but even then I will remain unable to imagine p being false until my imagination has been so changed. But if it turns out that no one can truly imagine p being false, then the antecedent isn't true, and we are all stuck unable to imagine p being false. I'm being no more dogmatic than everyone else in that case.

 

This then hinges on what it means to imagine a proposition being true or false. For example, could someone imagine...

 

p = at least one thing exists.

 

...being false?

 

Let's say someone says, "Yes, I can imagine it turning out that nothing at all exists, not even one thing. I can imagine that I am guilty of a verbal mistake and confused about the verb 'exists', and it turns out that nothing does it. So I can imagine p being false."

 

Either this person is lying, mistaken, or telling the truth. I cannot imagine p being false here, so I'm stuck with my certainty that p is true. This means I also cannot imagine it being the case that this person is telling the truth.

 

But if this person is right, and they really can imagine p being false, then I grant the possibility of my imagination becoming like theirs and me having a change of mind. But if they are lying or mistaken, then there might be no such possibility, and if it's the case that anyone who similarly claims to be able to imagine p being false is also lying or mistaken, then there is no such possibility at all.

 

This way I keep both my certainty and open-mindedness. If anything, this appears less dogmatic than skepticism because it's open to certainty.