3 ms·
Just finished a book club for Gödel, Escher, Bach, which gives a lay -- but pretty detailed and thorough -- explanation of the proof. Here's how I'd say lay out
by SilasX 6y ago
Just finished a book club for Gödel, Escher, Bach, which gives a lay -- but pretty detailed and thorough -- explanation of the proof. Here's how I'd say lay out the key insights as they applied to the parts I didn't understand before.
You want to use pure math to say "this statement cannot be proven". To do that, you:
1) Express the concept of statements following from other strings by the axioms or other theorems.
2) Express the concept of "this is a valid chain of statements, each of which follows from the previous". (GEB calls it a proof-pair.)
3) Express the concept of "Statement A does not have such a valid chain" (there exists no such chain with A at the end).
That allows you to say "Statement A cannot be proven."
From that point, it's a matter of extending it to the statement (call it G): "There exists a statement S such that S has no proof and S statisfies criteria A/B/C."
...and then constructing the criteria A/B/C such that G is the one statement that can satisfy it. The details of how you construct the criteria so that you are specifying G are the Gödel numbering that the Quanta article describes.
- ProfHewitt 6y agoUnfortunately, the above proof sketch does not work. Mathematics becomes inconsistent if a proposition G is allowed such that G<=>¬⊦G. For a simple proof that allowing such a G makes mathematics inconsistent, please see the following: https://papers.ssrn.com/abstract=3603021 https://papers.ssrn.com/abstract=3603021
- SilasX 6y agoThe proposition G in the sketch above just says G can't be proven, not that it's false, just like the one in GEB.