[zkmon]

What can be proven depends on what is allowed be a part of mathematics and logic. Zero, negative numbers, imaginary numbers and a lot other stuff had go through the acceptance first before they can be used in proofs. A lot of foundational concepts in logic, reality, causality, boolean exclusivity, spatial locality - had to be rewritten due to advances in quantum physics etc.

[KK7NIL]

We went through this over a 100 years ago, math now sits on very solid axioms (look up ZFC), they're not questioning that.

[thaumasiotes]

> math now sits on very solid axioms (look up ZFC), they're not questioning that.

People question C all the time. That might be the most prominent ideological difference in mathematical philosophy.

Does it matter? Of course not, but people question it anyway.

[cubefox]

Logicians and philosophers of mathematics have also questioned ZF set theory and "set theory" more generally.

For example the axiom of infinity (by finitists), the power set axiom and first-order theories in general (the downward Löwenheim-Skolem theorem implies that the infinity and power set axioms can't guarantee the existence on an uncountable power set), the fact that ZF doesn't allow a set of everything, and in particular no proper set complements, the fact that the axiom of regularity seems to be useless, etc.

Of course most ordinary mathematicians don't care about all that, because they don't care about ZF(C) or set theory or the foundation of mathematics in general. They rather care about problems in their specific field, like algebraic topology or whatnot.

[zkmon]

200 years passed by between Newton and Einstein. 100 years is tiny in the evolution of thought and is no basis for shutting the questions down.

[KK7NIL]

I wasn't trying to make an appeal to authority due 100 years having passed, that was tangential to my point that almost all modern math now sits on formalized axioms, which it did not do before the foundational crisis in math was resolved (about 100 years ago).

Comparing the axioms of math to relativity in physics is just nonsensical. Math is independent of observation, if a proof is formally correct now, it will always be correct under that chosen axiomatic system. Sure, we can play with different axioms (as others commented, it's common to drop the axiom of choice), but that doesn't invalidate the previous work at all.

[zkmon]

Math is not independent of observation. Math sits on logic which itself sprouts from human experience with the world around them. Math and logic are not alien pure forms isolated from this world. There is not even single concept of logic that is not fully tied to the human experience and perception (of the world).

The concepts such as true, false, equal, greater than - all refer to human experience with counting things or perception of existence etc.

[simiones]

While there used to be resistance to coming up with new formal systems that played with loosening certain restrictions in long-used systems, I think this has not been true for a long time. If you want to come up with a new set of axioms of arithmetic today in which pi = 3, and you can actually come up with a set of meaningful axioms and prove some interesting property of this formal system, I don't think it would be that hard to get mathematicians to accept it and occasionally use it.

[scentoni]

In taxicab geometry, π=4 https://en.wikipedia.org/wiki/Taxicab_geometry