Skeptical Theism: Can We Really Know All of God's Reasons?

What if there are some evils in this world we simply can't explain, no matter which of the defenses we've covered so far we throw at them? Is that a knockdown case against God's existence? Or is it just evidence that we don't know enough to judge? Let's consider it.

Escalating Through the Defenses

Let's run through a series of cases, each one designed to knock out whatever defense worked on the case before it.

Evil 1: An arsonist burns down someone's house. Handled by the Free Will Defense (Episode 8): it's bad the arsonist chose this, but good that God gave people freedom to choose between good and evil, even knowing some would misuse it.

Evil 2: A wildfire, started by no one, burns down a house. Free will doesn't apply — a forest doesn't make choices. So we reach for Character Building (Episode 9): the struggle to rebuild makes the person deeper and stronger.

Evil 3: The wildfire burns down the house of "Perfect Patty," who (let's pretend) already has flawless character. Character-building doesn't apply — there's no more character to build. So we reach for the Love Defense (Episode 11): Patty's neighbors now get to serve her, forging a depth of relationship a shared struggle makes possible.

Evil 4: Same fire, but Patty lives in "Isolationville," a hundred miles from anyone, population one. Love doesn't apply — nobody's there to love her. So we reach for the Megaphone Argument (Episode 12): even alone, the disaster gives her the chance to refocus off temporary goods and onto God Himself.

Evil 5: The fire burns down Saint Perfect Patty's house in Isolationville, and she simply burns to death instantly, no time to reflect. Saint Perfect Patty spends her days in prayer and knitting sweater vests for the poor. Now Megaphone doesn't even apply — she can't get any closer to God than she already was, and there's no time for refocusing anyway.

The Arms Race Problem

You might say: fine, let's find a sixth reason. But notice the pattern — this is becoming an arms race, one side inventing wackier scenarios, the other scrambling for a bespoke purpose each time. That's not great philosophy, and it reveals something about how the original argument is structured.

Premise (2) of the Logical PoE — "if God exists, He would prevent evil" — can be read two ways: God couldn't have any reason for allowing evil in general, or God couldn't have a reason for allowing every particular instance of evil. Our four defenses handle the first reading. Do they handle the second? Possibly not — and that's what Evil 5 is designed to press on.

The Problem of a Specific Evil

6. If God exists, then no irredeemable evil exists. ("Redeem" means bringing about a good from the evil that makes it worth it — through any defense we've covered. "Irredeemable" means no such good is available.)
7. There exists some evil, x, for which no one has discovered a redeeming explanation.
8. The Noseeum Inference: If x is an evil and no one has discovered a redeeming explanation for it, then x is irredeemable. ("Noseeum," as in the bug — "if you don't see 'em, they're not there.")
9. So, x is irredeemable.
10. So, some irredeemable evil exists.
11. So, God doesn't exist.

This improves on the original Logical PoE by isolating the pressure point precisely: not "evil in general is inexplicable" but "here's a specific evil nobody has explained, so there must be no explanation."

Response 1: Induction Is Weak Anyway

One response challenges premise 8 by noting it's an argument from induction (the classic pattern: swan 1 is white, swan 2 is white... therefore all swans are white). Our methods explained Evils 1 through 4, so we'd expect them to explain Evil 5 too. But inductive arguments are famously fragile — all it takes is one black swan, and Australia had plenty. Maybe there are genuinely inexplicable evils, the moral equivalent of a particle beyond our instruments' reach. That opens the door to a stronger response.

Response 2: Skeptical Theism

This is where Skeptical Theism enters, associated with philosophers like Stephen Wykstra, William Alston, and Michael Bergmann. Wykstra's 1984 paper "The Humean Obstacle to Evidential Arguments from Evil" introduced the CORNEA principle ("Condition Of ReasoNable Epistemic Access"): you're only entitled to infer "there's no reason for X" from "I can't see a reason for X" if, were there a reason, you'd actually be positioned to see it. Alston and Bergmann pushed this further, arguing our cognitive faculties simply aren't built to survey the full range of goods, evils, and their interconnections across a universe and an eternity.

12. The Noseeum Inference is only justified if x is a transparent evil — one whose redeeming properties we're able to discover, if it has any. ("Transparent" doesn't mean obvious, just discoverable.)
13. Any evil whose redeeming properties we can't discover is an opaque evil.
14. So, the Noseeum Inference isn't justified — we can't tell whether a hard case is transparent (genuinely irredeemable) or merely opaque (redeemable, but beyond our current sight).

Two examples of the difference. I once dated a girl from Rhode Island who was an absolute wreck of a relationship. My friend Stuckie told me flatly, "That is the dumbest thing you've ever said. She sucks." We broke up, and shortly after I met my wife — easy, in hindsight, to see the redemptive value. That's transparent.

Here's a less obvious one. When I left the Air Force, I wanted to teach university-level philosophy. Instead I had to grind out eleven years teaching high school philosophy, plenty of it frustrated with God for not giving me the career I wanted. Looking back, it made me a far better philosopher — high schoolers won't let you get away with any BS, so I was forced to present philosophy clearly, engagingly, honestly. Not easily discoverable — it took years — but still discoverable. Still transparent, just not obvious.

Now imagine an evil where we search for years and come up empty. Skeptical Theism doesn't say that proves it's redeemable — it just refuses to concede it's irredeemable. We simply can't tell opaque from truly irredeemable, given how limited our vantage point is.

Why Think We're in This Position?

Finite comprehension of the infinite. If God is infinite in knowledge and power and we are finite, it follows that God would have redemptive reasons beyond our capacity to grasp — not just "might," but would, given the sheer gap between infinite and finite minds. This doesn't mean we'll never understand any of them; it might be that God only permits "indeterminable" evils we'll understand once it's all explained, which connects to 1 Corinthians 13:12: "Now I know in part; then I shall know fully, even as I am fully known."

Incompleteness of moral intuition. Our moral intuitions feel obvious to us, but they've shifted substantially over history and show signs of running in cycles. Given how much moral opinion has changed even in recent generations, it would be remarkable if we'd just landed on the complete, correct picture now. If our intuitions are incomplete, we're poorly positioned to declare some evil couldn't possibly serve a good purpose.

Complexity of the whole story. There's an old image of embroidery: the underside looks like a chaotic mess of knotted thread, while the top shows a beautiful pattern. Looking only at the underside, you'd swear no thread had a purpose — but it does.

I think about this with one of the hardest things I've processed: the death of Katie Kyser, wife of a close friend, from cancer, leaving behind a five-year-old daughter. Was that redeemable? There was enormous suffering — it triggered my first anxiety attack, just imagining leaving my own son behind someday. But there was real redemption too: Katie was, by all accounts, joyful throughout much of her illness, and for me personally, it forced a reckoning with mortality that made me value life more and focus more on my own kids. I won't speak for anyone else in that story. But what economy of goods and evils would you even need to properly weigh a case like that? We may simply not be positioned to see the whole pattern — evaluating goods and evils across a life, a community, and possibly eternity is too complex for finite minds to be confident about.

Objection: A Good Parent Would Explain

If God is good, wouldn't a loving parent at least explain things rather than leave a child totally in the dark? This lands differently depending on who's asking — appealing to Job won't move an atheist, fairly enough. But for a Jewish or Christian reader, Job actually supports Skeptical Theism. Job is our Saint Perfect Patty: a righteous man suffering catastrophic losses who directly demands an explanation. You'd expect God to say "character building." Instead, in Job 38–41, God describes His own power and knowledge in sustaining the universe — essentially telling Job he isn't positioned to evaluate the full picture, and needs to trust.

I think about this with my own kids. I took them out to surf waves that were safe but a little outside their comfort zone — that builds courage. But if I'd explained beforehand exactly why facing that fear was good for them, they wouldn't have understood it; they could only grasp it after facing the fear and thinking, "I'm glad I did that." Before that, they just had to trust me. That's close to what faith is: not merely assenting to a proposition, but trusting that God loves you, knows better than you, and wants what's genuinely best — especially when He doesn't explain Himself in the moment. It's also entirely possible God explains it all eventually, in this life or the next.

Where This Leaves Us

The Noseeum Inference — "if I can't find a reason, there is no reason" — rests on an assumption we have no grounds for: that we're always positioned to detect redemptive purposes if they exist. Given the gap between infinite and finite minds, how much our moral intuitions have shifted, and how complex a single life's moral economy really is, I think Skeptical Theism gives solid grounds to say: an evil looking inexplicable to us isn't the same as it having no explanation.