in Codicalism, here are two senses of ‘consistency’: linguistic consistency (the model of some aspect of physical reality written in a particular language — or possibly multiple languages — is consistent) and substrative consistency (physical reality itself is consistent).
It could be that there is substrative consistency and we just haven’t found the right language for linguistic consistency.
Inconsistent formal systems based on paraconsistent and dialetheic logics do have contradictions but it’s not the case that that all sentences are provable (called “explosion”). There are paraconsistent logic programming languages, for example. Some think physical theory will turn out to be paraconsistent.
That’s linguistic inconsistency (within or between two languages). But is there substrative (actual physical events) inconsistency?
I don’t think so either (and once it is defined “what that would mean”, a language is involved), but I don’t know.
I would eliminate all vestiges of platonism by defining mathematics simply as this:
That what can be computed on some physical computer.
Now what physical computers can exist? (Black-hole computers? Bio-computers? etc.) That’s the question!
I question whether “simplest” is a principle that is the “best” for what a physical theory should be.
What if there were a multiverse theory generator — based on some sort of genetic programming — that output random, “messy” physical theories (all written in ActorScript, Go, Racket, …) and it turned out that our physical universe pretty much was modeled d by one of those.