Why I'm Not a Christian

TL;DR: the prior probability is too low

Before I say why I’m not a Christian, I should disclose my religious background.

I grew up vaguely Christian, and converted to Catholicism around the age of 14. At 17, I stopped taking communion, after realising that Catholics are committed to accepting the doctrine of divine simplicity, a devilishly complicated doctrine which I think is almost certainly false. After that, I became an agnostic about theism, an then became a theist again in my first term at Oxford after hearing maybe the best argument for theism ever invented. Now, I’m a religious nomad, looking for a religious home.

Probably because of my past, I find Christianity the most emotionally attractive religion. Despite this, I’m not a Christian, because I don’t think Christianity is very probable.

Here, in brief, is why.

I. The Case In Favour

I want to give credit where credit is due. There is some evidence for Christianity—namely, there’s historical evidence for the resurrection. Whatever happened after Jesus was killed, the sceptic is going to have to think something pretty unlikely happened after the crucifixion. There are some weird historical facts that need explaining, and I don’t have a preferred theory of how to explain them.

In general, when I’m not an expert on something, I try to defer to the majority of experts. They might be wrong, but I’m even more likely to be wrong, given that they’re academic weapons and I’m nothing but an academic potato gun. That in mind, I think most New Testament scholars who have studied the resurrection don’t think the evidence is compelling. There are things an informed sceptic can say. If you want to argue that the evidence is so strong that it should force us to think a miracle happened, that’s fine, but you have your work cut out for you.  

I think most parties to the resurrection debate agree—on both sides—that though the resurrection evidence isn’t compelling, it’s not terrible, either. Whether we should think it happened will boil down to our priors. In particular: the prior probability we assign to God being a trinity, becoming incarnate, and dying for our sins. (The prior probability of something, btw, is the probability we should assign it prior to looking at the evidence. In this case, I’m asking: what is the prior probability of God being a trinity, incarnating, and atoning, conditional on theism? Put another way: if all we knew was that a perfect being existed, to what extent should we expect him to (a) be a trinity, (b) incarnate as a human, and (c) die for the sins of the world?)

Currently, I think the prior probabilities of all of these—trinity, incarnation, vicarious atonement—are really pretty low.

Intellectually, at least, that is why I’m not a Christian.

II. Trinity

According to most Christians, God is one being but three persons, who all share in the divine essence. Thinking about it in Bayesian terms, the trinity doesn’t help explain anything about reality that a unitarian God couldn’t explain just as well. So on the face of things, the trinity seems like a deadweight cost.  

To quote Graham Oppy: “It seems [...] the doctrine of the Trinity adds significantly to the ontological and ideological costs of Christianity without leading to any improvement in the explanation of data.”¹ 

Some Christians have tried to get around this by giving a priori arguments for why a perfect being would have to be tri-personal. In particular, maybe there’s some fact about God’s love that entails he’d be tri-personal.  

The main argument to this effect—the Love Argument—runs roughly as follows. Suppose God the Father exists, as all Abrahamic faiths maintain that he does. If the Father exists, he is perfectly loving, and essentially so. After all, perfect love is an attribute God would have by definition. God just wouldn’t be God if he wasn’t perfectly loving.

But, the arguers argue, if the Father is always perfectly loving, he will always be in a state of loving someone perfectly, at every moment He exists. After all, doesn’t it just seem intuitive that someone who is actively in a state of loving someone perfectly is both greater, in some respect, and more loving than someone who only has the disposition to love someone perfectly but doesn’t exercise it?

But if the Father is always in a state of loving someone perfectly, he will eternally generate—or necessarily co-exist with; take your pick—at least one other perfect (and therefore divine) person with whom He can enjoy a perfect, loving relationship. From this, we get something like the claim that God must be multi-personal.

But how do we get from multi-personal to tri-personal? Why think there are exactly three divine persons in God and not, say, two, or ten, or twelve, or 860,527?

Well, arguably—the arguers again argue—there are three (and only three) types of love, types of love that are fundamentally irreducible to one another. First, there’s self-love—the kind of love any self-respecting person should have for themselves; second, there’s love of another—the kind of love we see mutually expressed in two-person friendships and monogamous marriages; finally, third, there’s co-operative love—the kind of love where (a minimum of) two persons co-operate in order to jointly love (a minimum of) of one other person, as with parents who co-operate in order to jointly love their children.

If the Father is perfectly loving, and always loves perfectly, he’s going to have to express each kind of love on this list, since if he failed to express any one of them, at any time, he wouldn’t be the lover-than-which-no-greater-lover-can-be-conceived.

If we accept this, the rest of the argument falls into place. For the Father to express self-love, all he needs is himself. But for the Father to express love-of-another, and express it perfectly at every moment, there needs to be at least one other divine person for God to love on, and be loved by. And for the Father and this other divine person (let’s call him, I don’t know, the Son) to express co-operative love, they need at least one other divine person to co-operate in the loving of.

They might even name him the Holy Ghost.

Now, you might wonder: even if the Love Argument shows that a minimum of three divine persons are needed for God to be perfectly loving, why would the headcount stop there? Well, while the Love Argument doesn’t rule out there being more than three persons in God, it doesn’t give us reason to posit any more than three, and so—by Ockham’s Razor—we should stick with three and go no further.

I think this is the best a priori argument for the Trinity, but—alas—I’m not at all convinced by it. I have a number of worries, but I’ll stick to two. (Note: two is “a number”, so it’s logically possible that I’m bluffing, misleading you into thinking I have more objections than I really do. Whether I am is a psychoanalytic exercise left for the reader.)

First, as Dale Tuggy has pointed out,

Perfect lovingness is a character trait. Like other character traits, it may be had but not expressed; it doesn’t imply being in an actual interpersonal relationship. It’s like the trait friendly. Imagine a friendly man who has been shipwrecked on a desert isle. Now, sadly, the man is friendless; he has no other with whom to share his life. Still, he’s a friendly man; if we put another suitable castaway on the island, he’ll form a friendship with that person—or at least, he’ll have a strong tendency to do that. That’s what it is to be friendly: it’s to be disposed, in appropriate circumstances, to enter into and remain within (at least superficial) friendships. Just so with the quality perfectly loving. It is having the disposition to act and reacting perfectly loving ways, if and when there is another to love. In isolation, one may still be perfectly loving.²

Second, as I’ve argued elsewhere, I think this argument proves too much. If we grant that God must necessarily express each form of love, there’s a plausible case that love of large groups—e.g., love of country, love of football teams—is a form of love that’s irreducible to self-love, love of another, and co-operative love, and couldn’t be expressed with just two other divine persons. Thus, the argument would imply that there is a large—and probably infinite—collection of persons in the divine substance. I won’t flesh this out here, but read this if it tickles your fancy.

There are other anti-unitarian arguments from divine perfection, but I think Dale Tuggy has convincingly dealt with nearly all of them.

III. Incarnation

According the the doctrine of the incarnation, as defined by the councils of Nicaea, Ephesus, Chalcedon, and so forth, the second person of the Trinity took on a human nature in the person of Jesus. Jesus’s human nature is not his divine nature and vice versa, Jesus is fully human and fully divine, but he’s only one person, not two.

I don’t know exactly what I’d say about the incarnation if I were a Christian, but I assume that with all the intellectual manpower that’s been put into finding some way in which the Christological predications can be made logically consistent, someone at some point has probable found some way to make the doctrine cohere in a theologically acceptable way.

But logical consistency is a low bar. It’s logically possible that God’s favourite colour is pink. The question is: What’s the prior probability that God would do this, assuming it’s possible?

One suggestion is that God would probably become incarnate because that’s the only way for him to suffer alongside us. This is meant to be obligatory for God since—if he’s putting us through a life of suffering—it would be unfair of him not to show solidarity with us in that suffering, even if the suffering is justified in light of greater goods.

My reply to this is that we have good reason anyway to think God is passible. That is: that he’s emotionally affected by things that happen in the world. This isn’t to say that good is at the mercy of the world or that events like the holocaust take God by surprise. God is omniscient, and he chose to create a world with genocides in it. To say that God is passible is just to say that events like the holocaust grieve the heart of God, just like they grieve the hearts of anyone with properly functioning emotional faculties.

Why think God is passible? Well, just as beliefs can be rationally appropriate or inappropriate, so too can emotions. If a genocide is happening, that gives us a reason to feel aggrieved—if we felt pure, undisturbed happiness instead, that would be an irrational happiness, since the emotion doesn’t suit the world it’s responding to.

But since God is omni-rational, we should think he’s rational in this regard, too. If his beliefs, desires, belief-desire pairings, and corresponding actions are rational, and we don’t think God is a psychopath or a mindless zombie, why would we think differently about his emotions? Why would God enjoy perfect, undisturbed bliss when there are things it’s obviously irrational to feel that way about?

I submit to you that if you believe in a perfect being, you should think that being suffers. But if God suffers anyway, then there’s no need for an incarnation to make it possible for God to suffer alongside us—he already does!

You might worry that even if God suffers, without an incarnation, he couldn’t suffer in the exact same ways as us, and so his ‘sharing in our’ suffering would be inadequate. In reply, two things. First, even if God doesn’t suffer in exactly the same ways as us (i.e., He doesn’t feel pain with a bodily phenomenology), I don’t see why this would make his suffering inadequate: plausibly, if God is passible, his suffering would be much much worse than anything any human or collection of humans has every suffered, since his emotions are perfectly attuned to the horrors of evil in the world, and He can’t engage in self-deception.

Second, if you think God could only suffer in solidarity with us if he suffered in the same ways we suffer, a single incarnation of the kind Jesus underwent is the last thing you’d expect, since Jesus didn’t suffer the whole range of evils that other humans have suffered. If you really think God would have to suffer in the exact same ways as humans, you should go the whole hog and buy Linda Zagzebski’s thesis that God is omni-subjective (that is: that God perfectly grasps the conscious states of all conscious beings from their own point of view, including states of pain and suffering.)

But apart from this argument from the supposed moral requirement on God to suffer in solidarity with us, I can’t see a way to make theism predict an incarnation.

IV. Atonement

According to the doctrine of the atonement, Jesus death somehow made it possible that humans could be reconciled to God, without punishment, even in light of their sin.

On some theories of atonement (e.g., Abelard’s Moral Influence Theory), there’s nothing especially mysterious about this process. For Abelard, the way in which Jesus’ death helped reconcile us to God is by stirring our hearts with its symbolic, emotional power. This is fine, philosophically, but I don’t see a way to make theism predict it.

However, most atonement theorists think that while moral influence is part of the Biblical picture of atonement, it’s incomplete by itself. As far as the other theories, though, all I have are questions. I just don’t think I’ve seen a remotely independently plausible story of why an atonement would be necessary to bridge the divine between God and humans, and how its moral mechanism would operate.

Penal substitutionary models—one’s on which Jesus is made suffer/is retributively punished in our place, despite being morally guiltless—strike me as especially wacky. If philosophers of religion have a weak spot, it’s ethics, and the recent philosophical defences of Penal Substitution are maybe the worst exercises in moral philosophy I’ve ever read, short of Ayn Rand.

If anyone knows a good theory of atonement, though, let me know in the comments.

V. Pascal’s Wager

According to Pascal’s Wager, you have most prudential reason to commit to—or (non-viciously) cultivate belief in—whichever afterlife-promising religion you take to be most probable, given that the eternal stakes are so high.

I accept a modified form of Pascal’s Wager. While I’ve published on Pascal’s Wager from a critical angle before, I think my objection isn’t persuasive enough to overcome the rock-solid decision-theoretic case in it’s favour.

So maybe, despite everything, I should commit to Christianity anyway. After all, I’ve already said there’s some evidence for Christianity—definitely more evidence than there is for, say, Islam—and so maybe Christianity is the most probable religion.

Honestly, if I convert to Christianity in the near future, it’ll probably be for Pascalian reasons, at least to start with. (Maybe after I take the Wager, I’ll come to think the evidence for Christianity is stronger than I initially thought.)

But two things are holding me back.

First, I’m pretty cautious about wagering: I’ll only Wager if my credence in Christianity being the most probable religion stabilised, and wasn’t something I was shifting on every other week. Religious conversions are the most important decisions we can ever make; it’s disrespectful, to God, to make them on a whim.

Second, I think Hinduism is more probable than Christianity, so—as of right now—I’m more inclined to take the plunge in that direction. (Maybe my debate tonight with John Buck will change my mind…).

1. Oppy, G., Oliphint, K. S., McGrew, T., Moser, P., & Gundry, S. N. (2016). Four views on Christianity and Philosophy. Zondervan Academic.

2. Tuggy, D. (2021). Antiunitarian Arguments from Divine Perfection. Journal of Analytic Theology, 9, pp. 263-264.