My understanding of truth

How do we know things?

Here is how I understand the nature of truth. Please comment with things I should read or with errors. I am sure it isn't novel.

1. There is a physical universe. They may or may not be a non-physical universe of math and love and/or a deity, but if there is, it is hard to get to.

2. We are big clusters of atoms which have somehow developed internal narratives of our selves.

3. When we understand things, it is that we have very good predictive models of them. We do not have direct access to the Truth of them.

When we sit on a chair we guess where the chair is, we send signals to our muscles and we receive feedback from our bums that we have landed on the surface. None of these are priveleged Truth packages, signed by God Himself, which we can trust without fail. We might be drunk, or the chair might be part of an optical illusion. Someone might have strapped electrodes to our bums to trick us into thinking we feel the seat.

If we imagine the universe is really a big spreadsheet, which tracks the location and momentum of every particle, we can guess the contents of this spreadsheet, but we can never be certain we know it.

4. When we say we 'know' something, we mean that it is implied by a model that we are confident of.

The position of chairs is implied by our visual and spatial models which give expected results an absurdly high percentage of the time. "What about illusions" *What about them?* I am guessing the positions of 100s of disctinct objects all day for my whole life and maybe less than 1000 times in my adult life that particular model has been wrong. But this doesn't mean that my model of the world predicts where things *actually are*:

• We could be running on a big computer. Then there would be no things to be anywhere

• I could be insane

Perhaps it seems like 99.999999% should round to 100. But I think there is a really important difference here.

If we 'know' something then we have gained some deep insight into *The Way of Things*, into that spreadsheet. We can never be wrong and we could anchor all other knowledge to that point. It seems plausible that we could be sure about the mechanisms we used to figure that thing out as well. Eventually this certainty might apply very broadly - a freezing spell that climbs the anchor link by link and then freezes the whole ship.

This does not seem to be the the case. We are wrong about many things, including some we are very confident of. And it's worth pretending that some things are certain, so as to save ourselves doing the calculations, but 99.999999999 is still different than 100.

5. We do no receive Truth from the universe. We guess what the universe is like and then we test it.

This is from Deutsch. Its tempting to think that the Truth is outside of us and that it pours in, via our eyes, ears or some divine inspiration. I'll leave the 3rd one, but I don't think it's the first two. How would the Truth get in?

Instead any truth that there is comes from inside of us. Every time we hear an explanation, we build it from scratch in our own heads and check that we understand it by spinning up examples or images, by asking questions. The knowledge of a pig was not copied into your brain like a .png file, you saw many pigs and made up your own definition changing it whenever you wrere corrected. Even for things where you might think you were explicitly told, I doubt that's the case. I don't think the equation for a circle is encoded in some file in your brain.

If that weren't the case - if the triangle crawled into your brain and wrote "I am a triangle on the walls" how could that explain why some people misunderstand triangles? Did they not absorb it properly? Were they rude to their first semi-platonic triangle and it pettily confused them? I bet some people can repeat "a 2D shape with 3 straight sides, each line connected at its ends to the two other lines" and still not understand it. What's going on?

5. When we are trying to figure things out, we are building models and getting feedback on them

We have models we trust more and less and we are guessing what will happen if we do this or that. In some cases we test instinctively, such as when I kick a ball, or deliberately, when I reckon I can bluff a hand of poker, or in discourse, when I test a new idea of how housing shortages damage our lives, or via experiment or mathematical proof.

6. Some feedback is better than others.

In Bayesian terms, some things cause me to update more. I would prefer a science experiment to my personal jotting. I prefer a proof where I understand all the steps to a sketch of one.

But these rules are only rough. Sometimes I update surprisingly. I would trust my friend Usama's opinion on an economics paper more than if I read it and thought I understood it all myself. I would trust that physhics still works even if someone showed me an experiment where it broke - probably the experiment is a trick, the first time at least.

7. This model seems better than any my mind can think of, but it isn't True

My understanding of the world is like a cluster of ships on a deep sea. None are achored, but slowly the flotilla grows and people can live there. It really is possible that in some deep way I am mistaken - perhaps I am insane or simulated - but for the sake of this discussion, being sane in a physical universe is the best I've got.

So to recap:

• The universe is real

• We do not have definte contact with that universe

• We do not receive Truth from the universe, at best, we make it up at worse we are deeply deluded

• This is okay. It seems to work almost all of the time

• Experiments, discussion, proofs are funamentally the same type of thing - they are the way we test our models with sense data external to those models

• Experiments can be misleeding, proofs can have subtle errors, discussions can be biased

Again, I’d appreciate reading recommendations.