Start with the question of what a truth even is. Philosophers have circled this for a long time without settling it. Some say a truth is whatever matches reality. Others say it’s whatever holds together inside a system of beliefs. Pragmatists say it’s whatever works. Deflationists shrug and say the word adds nothing. They say calling something true is just another way of asserting it. None of these positions has knocked the others out, which is probably the first clue that we’re dealing with something harder than it looks.
If we grant that truth exists, certain constraints fall out of the idea almost on their own. A truth can’t contradict itself. It can’t shift depending on who’s asking. It must be about something. It must be checkable in principle. These are minimum demands. Even philosophers who disagree about what truth is, tend to accept that whatever it is, it must obey something like these rules. Otherwise, the word loses any usefulness.
But constraints aren’t the same as foundations. Knowing what shape a truth must have doesn’t tell you how to get one. So, the real question becomes; what does it take to prove something?
The trouble with proof
Strip proof down to its bones and you find this: a demonstration that, given what we both already accept, the claim follows. Shared starting points, legitimate moves, a chain to the conclusion. That’s the minimum. It works inside mathematics, inside courts, inside everyday argument.
The trouble is that every proof needs premises, and those premises themselves need proving, and so on. You can only escape this in three ways: by going on forever, by going in a circle, or by stopping somewhere and calling it self-evident. The ancient skeptics noticed this and called it a trilemma. None of the three escapes is satisfying. The infinite regress never lands. The self-evident foundation just relocates the problem. And the circle looks like cheating.
Can a circle do real work?
It depends on the circle. A tight one “God exists because the Bible says so, and the Bible is true because God wrote it” is plainly empty. It proves nothing because it presupposes everything. But there’s a different kind of circle, a wide one, where many beliefs prop each other up across a whole system. Coherentists think this kind of mutual support is the best we can do, and that the strength of the web is what makes knowledge possible.
The weakness of that picture is that a perfectly consistent system can also be perfectly wrong. A conspiracy theory holds together. An ideology holds together. Internal consistency doesn’t prove contact with anything outside itself. Circles can lock in falsehoods just as well as truths, and once locked in, they’re hard to escape because everything inside the circle vouches for everything else.
So circularity alone won’t do. It might keep a truth stable once you have one, but it won’t generate one. Something else is needed — something that pushes back from outside.
Triangulation
When I was young I spent many hours with my father mapping. I would stretch out a chain for him, and he would measure the angles of a distant point at each end of the chain. So, mapmakers had solved a version of the truth problem long ago. You can’t measure a distant mountain by walking to it. But you can stand in one known place and take a bearing. Then move to a second known place and take another. The two lines cross, and where they cross is the mountain. Add a third bearing as a check. If all three lines meet at one point, you’ve found it. If they don’t, something is wrong and you know it.
What’s powerful about this is that none of the individual measurements is foundational. Each is fallible. But because they’re independent, their errors don’t compound, they cancel. The convergence is more reliable than any single bearing could be. And the method announces its own failures: when the lines don’t meet, you can’t pretend they do.
This is roughly how science works at its best. Different lines of evidence from fossils, genetics, biogeography, direct observation all meet and point at the same conclusion. The age of the universe is triangulated from several independent calculations that didn’t have to agree but do. None of them is certain on its own. Their agreement is what makes the result hard to dismiss.
But the mapmaker is one person
Here the analogy nearly breaks. The mapmaker chooses where to stand, reads the instruments, judges when the lines converge. Every step passes through one mind. So in what sense is this really triangulation between independent viewpoints? Three measurements taken by the same person, filtered through the same assumptions, might just be one perspective wearing three coats.
This is a serious objection and it doesn’t fully go away. A lone observer with a quiet bias could produce a map that looks internally perfect and is still wrong. The convergence would be an artifact of the mind, not of the world. The philosopher, Davidson, thought the only fix was another mind. He thought that objectivity requires a second perceiver looking at the same world, because only then can you tell the difference between how things seem and how things are.
The instrument as another viewpoint
I worked for most of my life with scientific instruments and I believe they offer a way around the lone-observer problem that doesn’t require a second consciousness. The instruments themselves do part of the work. A compass doesn’t share the mapmaker’s hopes. A theodolite doesn’t care what answer would be convenient. I catch a 13 inch trout, but my ruler tells me is 12 inch. My ruler contradicts me; tells me I am wrong. These tools encode the discoveries of people the user never met. These experts were opticians, geometers, clockmakers, generations of testing and correction. When the mapmaker looks through a lens, they’re already in collaboration with absent minds, and with the physics those minds figured out.
More importantly, the instruments can disagree with their user. A thermometer reading 40 when you expected 20 has contradicted you. It has acted, in a way that another person disagreeing would act. It doesn’t matter that it isn’t conscious. What matters is that the resistance came from outside the mind, not from inside it. The instrument is a delegate of reality; a piece of the world rigged to report on the rest of it.
This changes the picture. Objectivity doesn’t strictly require another consciousness. It requires another source of constraint — something the mind doesn’t control, that can push back on it. Other minds are one such source. Instruments are another. The world’s own behaviour is a third. A child learning that stoves are hot is triangulating against reality without needing a peer. The criterion isn’t how many minds agree. It’s how many independent sources of resistance the belief has survived.
Where this leaves truth
Truth, on this view, isn’t something we derive from certain foundations. There aren’t any. It’s what survives convergence from as many independent angles as we can muster — different people, different methods, different instruments, different eras. Each angle is fallible. None is foundational. But their intersection narrows reality down further than any of them could alone.
This sidesteps the old trilemma without pretending to solve it. We don’t need an infinite regress, or a self-evident starting point, or a vicious circle. We need a wide enough parliament of fallible witnesses that the particular biases of any one of them stop mattering much. The history of human knowledge, looked at this way, is the history of building better otherness. Constructing new instruments, gathering wider communities, finding new ways to let reality talk back.
None of this gives us certainty. A perfectly triangulated belief can still be wrong if every viewpoint shares some hidden bias. What it gives us is something more modest and more honest: belief that has been tested from many directions and hasn’t yet broken. That’s not nothing. It might be everything we ever actually have.
The mapmaker is a single mind. But the map, by the time it’s been used and corrected and redrawn across centuries of sailors who needed it to work, is no longer the product of any single mind. It’s a sediment of many — and of the world’s stubborn refusal to be whatever they wanted it to be. That sediment is, in the end, what we mean by something being true.