Thomas Zinn

Frameworks · July 2026

What We Can and Cannot Know

After walking the deepest questions one at a time, it is worth stepping back and sorting them not by topic but by how much can be known. One thing is certain. Most is a calibrated bet. And a few questions are not open in the ordinary sense at all: they are closed to evidence, and the honest move is to say so.

The deepest questions have been walked here one at a time: whether it began, what the ground is, whether there is one or many, what holds the laws, and what the traditions were each describing. It is worth stepping back and sorting them, not by topic, but by how much can actually be known. When you do, they fall into a few sharply different kinds, and the honesty of the whole project depends on not blurring them.

The one certain thing

Start where certainty ends, which is almost immediately. Exactly one thing is beyond doubt: that experiencing is occurring. It needs no evidence because it is not a claim about the world; it is the world present to itself. Everything else is downstream of that single point, and everything else is less than certain.

Open: most of the map

Most of the deep questions are open in the ordinary sense. Whether the ground is aware or blind, whether experience is fundamental, what sets the specific laws, whether there is one ground or many: these are real questions that better argument or better evidence could still move. They are held here as leans, each with a defeater stated plainly, the thing that would lower it. Open is not the same as unknowable. It means we do not know yet, and could.

Belief, not knowledge

A second kind is different. A creator mind and a founding miracle sit past all possible proof, because a claim about intent, or about an origin no one witnessed, cannot be closed by evidence. Getting behind those takes belief, not knowledge. That is not a flaw in the ideas; it is their location, past the last thing that can be shown. The honest move is to mark the boundary, and to refuse both easy errors: dressing a belief as a finding, or dismissing it as if its being unprovable made it false.

Undecidable in principle

The third kind is the strangest, and the smallest. A few questions are not waiting on better evidence; they are closed to evidence forever. Whether the world had an absolute first or is only a phase in something beginningless, whether there are causally sealed-off grounds we could never detect: for these, every possible observation sits on the wrong side of the thing being asked about. No measurement could ever reach across. A question like this can be perfectly meaningful and still be permanently undecidable, and where that is so, the honest output is a clearly marked blank, not a guess wearing the clothes of a discovery.

The shape of it

That is the whole epistemic shape of this section, and of the site behind it. Certain at exactly one point. Graded everywhere it can be, with the weights and the defeaters shown. And blank, deliberately, where no grade will ever be earned. The difference between "we do not know yet" and "we can never know" is one most writing about these questions quietly erases. Keeping it visible is most of what honest thinking about them amounts to.


Part of The Open Questions, and resting on Starting From "I Am". The tagged version, with the full map of every question's status, is in the Library: What We Can and Cannot Know.

Questions

What can we actually know for certain?
Exactly one thing: that experiencing is occurring. It needs no proof because it is not a claim about the world; it is the world present to itself. Everything else, the external world, the ground, the origin, is held at less than certainty, as a calibrated bet with a stated defeater.
What is the difference between open and undecidable in principle?
An open question is one that better evidence or argument could still settle; most of the deep questions are open, held as leans. A question undecidable in principle is one where every possible observation sits on the wrong side of what is being asked, so no evidence could ever reach it, such as whether the world had an absolute first or is a phase in something beginningless. It can be perfectly meaningful and still be permanently closed to proof.
Is belief ever the honest response?
For some questions, yes. A creator mind and a founding miracle sit past all possible proof, so getting behind them takes belief rather than evidence. That is not a defect in those ideas; it is their location, past the last thing that can be shown. The honest move is to mark the boundary rather than dress a belief as a finding or a finding as a belief.

Keep reading