CeleryKills

From agnosticism to rejecting the terms of the God debate

Ok, this was written on a rare PNW sunny day, but I like the rain allusions, so sue me.

Rain has a way of flattening big claims out here. Not disproving them, just pressing on them long enough to see where they leak. I think my position on God started that way when I was ten or eleven, sitting with questions that felt too large but also oddly ungrounded. Even then, something about the debate itself felt off. Not wrong exactly. Just… improperly framed. Years later, after pushing it against everything from Russell to more contemporary analytic work, the core thought hasn’t really moved. It sharpened, sure. It lost patience.

The simplest way I can put it is this:

“I don’t reject God because it’s false. I reject the question because it never becomes a coherent claim in the first place. If a concept cannot survive its own evidential standards without collapsing into circularity, underdetermination, or unfalsifiability, then it never qualifies as something I can meaningfully evaluate.”

That sounds stronger than agnosticism because it is. It steps outside the belief-versus-knowledge spectrum and asks whether the proposition itself ever earns entry. That’s the pivot. Not “is it true,” not even “can we know if it’s true,” but “has it achieved the status of a truth-apt claim at all.” If the answer is no, then suspension of judgment starts to look less like caution and more like a category mistake.

A steelman version of this position would go something like this. The concept of God, across its many formulations, fails to stabilize into anything that is both clearly defined and evidentially accountable. Some definitions drift into abstraction so thin they cannot be distinguished from metaphor. Others embed their own justification, becoming epistemically circular. Still others retreat into unfalsifiability, insulating themselves from any possible counterevidence. In each case, the concept avoids disconfirmation at the cost of explanatory power. What remains is not a hypothesis in competition with alternatives but a moving target that cannot lose. And if it cannot lose, it also cannot win. It exits the field. At that point, withholding belief is not quite right, because there is nothing sufficiently formed to withhold belief from.

This lands closest to ignosticism, or igtheism, and more formally to theological noncognitivism. Both hold that statements like “God exists” fail to express a proposition with clear cognitive content unless “God” is defined in a precise and testable way. A. J. Ayer pushed a related line when he argued that metaphysical statements lacking empirical verification are literally meaningless, not false but nonsensical (Ayer, Language, Truth and Logic, 1936). That’s a hard edge version. My view is a bit less doctrinaire but runs along the same channel. If a claim cannot, even in principle, differentiate itself from alternatives or expose itself to disconfirmation, what work is it doing?

Ignosticism, sometimes called igtheism, takes its starting point not from disbelief but from definitional failure. The term is most often linked to Paul Kurtz, who argued that theological disputes routinely proceed without a clear, agreed-upon definition of “God,” rendering the debate prematurely framed (Kurtz, The New Skepticism, 1992). The core idea is procedural rather than doctrinal. Before asking whether God exists, one must specify what “God” refers to in a way that is coherent, internally consistent, and capable of distinguishing itself from competing concepts. Without that, the statement “God exists” lacks a stable referent and does not rise to the level of a truth-apt claim. It functions more like an expression of attitude or a placeholder for intuition than a proposition subject to evaluation. Ignosticism therefore suspends not belief but entry into the question, treating most theistic disputes as semantically underdeveloped rather than empirically undecided.

Theological noncognitivism pushes this line further and grounds it more explicitly in twentieth-century analytic philosophy. Influenced by logical positivism, especially A. J. Ayer’s verification principle, it holds that a statement is meaningful only if it is empirically verifiable or analytically true (Ayer, Language, Truth and Logic, 1936). On this view, many theological claims fail not because they are false but because they lack the conditions required for cognitive meaning. Later philosophers such as Antony Flew sharpened the concern by arguing that religious claims often undergo a “death by a thousand qualifications,” retreating from falsifiability whenever challenged, until nothing substantive remains to be tested (Flew, “Theology and Falsification,” 1950). Where the ignostic calls for clearer definitions, the noncognitivist is often skeptical that such definitions can be produced without collapsing into metaphor, circular reasoning, or unfalsifiability. The conclusion is not atheism in the usual sense but a refusal to treat “God exists” as a proposition that successfully enters the space of meaningful assertion.

Russell’s teapot is the obvious benchmark, though I resisted it for years because it feels a little too tidy. Still, the point holds. If someone asserts a china teapot orbiting between Earth and Mars, invisible and undetectable, the burden does not shift to me to disprove it (Russell, Is There a God?, 1952). What’s more interesting is that Russell doesn’t spend much time trying to refute the teapot. He questions whether the claim deserves uptake at all. That’s the move I find more compelling. Not disbelief. Non-admission.

Popper sharpens this with falsifiability. A theory that cannot, in principle, be falsified is not scientific, which is not quite the same as saying it’s false, but it does place it outside a certain domain of rational evaluation (Popper, Conjectures and Refutations, 1963). Now, theology often replies by shifting domains. Different kind of claim, different standards. Fair enough. But then I want to know what those standards are, and whether the claim can meet them without leaning on itself for support.

Elliott Sober’s work on parsimony adds another pressure point. When multiple explanations fit the data, we prefer the simpler one, the one that does not multiply entities beyond necessity (Sober, Ockham’s Razors, 2015). If “God” does no explanatory work that cannot be done by naturalistic accounts, then it is, at best, redundant. At worst, it becomes a placeholder for gaps that steadily shrink. So I find myself asking, not rhetorically, what would count as disconfirming this concept? What observation would look different if God did not exist? If the answer keeps sliding away, then I’m back to the same suspicion I had as a kid. This isn’t a stable question.

Where this position works is in its insistence on conceptual discipline. It refuses to grant automatic legitimacy to inherited questions. It demands that claims earn their intelligibility. That seems basic, almost procedural, but it cuts deep. It also aligns with a broader empiricist and skeptic tradition that prioritizes clarity and testability without pretending to exhaust all forms of meaning. There’s a kind of intellectual honesty in saying, “I don’t even have enough here to begin.”

Where it may not work is in the possibility that it overextends criteria drawn from science and analytic philosophy into domains that operate differently. William James, for instance, would argue that some beliefs are “live options” that cannot wait for sufficient evidence and are justified by their practical consequences (James, The Will to Believe, 1896). From that angle, demanding strict testability may miss something about how humans actually engage with existential questions. There’s also the risk that the bar for coherence is set so high that many meaningful but messy concepts get excluded along with the weak ones.

I had a recent exchange that brought this into sharper focus than most abstract arguments ever do. Not because it was especially sophisticated, but because it was ordinary in exactly the right way. The kind of conversation you see play out a hundred times, and almost every time it misses its own conclusion by a few inches.

It started simply enough:

“There are only 2 categories: theist or atheist. All other flavours are subsets.”

That framing always sounds clean, almost reassuringly binary. I used to accept it without much resistance. But once the conversation moved into agnosticism, the cracks started showing.

I responded:

“I don’t think there is a valid ‘God’ question to evaluate. There is no positive proof, just metaphysical positions, none of which survive their own evidential standards without circularity, underdetermination, or unfalsifiability.”

That’s where people usually pause. Not because they disagree outright, but because something about it reframes the terrain. The reply came back quickly:

“Of course you can suspend judgment. Just because there is no evidence doesn’t prove nonexistence. That’s the black swan fallacy.”

And this is the hinge point. The conversation drifts toward proof, toward certainty, toward that familiar demand that nonexistence must be demonstrated before disbelief becomes justified. But that’s not actually the claim being made. So I pushed it back a step:

“I’m not asserting that absence of evidence proves absence. I’m saying that when a claim produces no positive evidence where we would reasonably expect it, and cannot specify what would count against it, it exits the domain of meaningful evaluation.”

That distinction matters more than it first appears. Because once you accept it, even provisionally, the tone of the conversation shifts. You’re no longer asking whether a claim is true or false. You’re asking whether it has managed to become evaluable at all.

The discussion kept looping, but narrowing:

“Existence claims are falsifiable. Nonexistence claims are unfalsifiable. The negation is always the default.”

It sounds tidy. Almost mathematical. But it only holds when the existence claim is actually structured to be falsifiable. So we end up circling Russell’s familiar terrain without naming it directly at first. Eventually I do:

“If someone claims a teapot exists in orbit but cannot be detected by any instrument, that’s an existence claim insulated from falsification. It cannot be disproven, but that doesn’t give it evidential weight.”

And then, almost quietly, we converge:

“Until they prove existence, the null hypothesis says they don’t.”

That line stuck with me. Not because it resolves anything, but because it reveals how close the positions already are. We end up in nearly the same operational stance. No acceptance. No practical commitment. No expectation that the claim will be substantiated. And yet the language still clings to this idea that the question remains open, that judgment is suspended, that something is still pending resolution.

But what exactly is pending?

If you pause the conversation right there and look at it from the outside, a strange thing emerges. One side insists that claims without evidence should not be accepted. The other agrees. One side points out that unfalsifiable claims don’t meet normal standards of evaluation. The other implicitly agrees by appealing to the null hypothesis. And yet both still orbit this idea that agnosticism is the safe landing point.

So I find myself asking, not rhetorically but seriously. If we agree that a claim has produced no evidence, offers no testable structure, and resists falsification by design, what exactly are we suspending judgment about? What would resolution even look like? And if no one in the conversation can answer that without shifting definitions midstream, are we disagreeing, or just maintaining the form of a disagreement out of habit?

That’s the moment where my position diverges, not by leaping to certainty, but by stepping off the track entirely. The realization isn’t that God does not exist. It’s that the conversation never stabilized into something that could bear the weight of that conclusion in the first place. And once you see that, it’s hard to unsee.

Still, I keep circling back to a simpler dilemma. If two people use the word “God” and cannot, after sustained effort, specify conditions under which their claims would diverge, are they disagreeing or just gesturing in parallel? And if the latter, what exactly am I being asked to evaluate?

This isn’t disbelief in the usual sense, and it isn’t suspended judgment. It’s more like stepping to the side of the whole exchange and asking why the ball keeps changing shape midgame. I’ve tried, over time, to give the question every advantage. Read the generous interpretations, the sophisticated metaphysics, the careful agnostic positions. I understand the appeal of holding the door open. But at some point, keeping the door open starts to look less like openness and more like reluctance to admit the room was never built.

Maybe that’s too sharp. Maybe it’s just the weather talking.

References 

• Ayer, A. J. Language, Truth and Logic. 1936. 
• Flew, Antony. “Theology and Falsification.” 1950. 
• James, William. The Will to Believe. 1896. 
• Kurtz, Paul. The New Skepticism. 1992. 
• Popper, Karl. Conjectures and Refutations. 1963. 
• Russell, Bertrand. Is There a God? 1952. 
• Sober, Elliott. Ockham’s Razors: A User’s Manual. 2015.