3 ms·
You are making a philosophical choice that was heavily debated in the early 20th century, namely, you contend that a mathematical foundation is truly foundation
by pron 10y ago
You are making a philosophical choice that was heavily debated in the early 20th century, namely, you contend that a mathematical foundation is truly foundational (perhaps even unique), and that math is built on top of one. Another alternative (favored by Turing[1]) is that any mathematical foundation is just like any other calculus, only one that deals with the lower-levels of mathematics.
Also, by picking computation as the "true" foundation, you are being a bit arbitrary. On the one hand, you can get more grounded. Computability was justified on physical arguments, and Turing and others recognized that computability is only a rough approximation of feasibility, which was given a more precise treatment much later. So if you want to base the foundations of math on the physical -- which is what you're doing if you're basing them on Turing computability -- then you can go a lot further down towards "grounded". On the other hand, there is really no reason to limit math at the computable. Brouwer thought there was, but Turing didn't. If non-constructive math yields results that are useful, compatible with constructive math and is easier to work with in some cases, what justification is there to reject it other than by taking a view that is both fundamentalist (in the sense I described) and somewhat arbitrary? After all, constructive math is a very different math, and if classical math rests on shaky foundations, how is it so useful in practice, and how come it agrees with constructive math on everything that is physically observable?
[1]: https://mdetlefsen.nd.edu/assets/201037/jf.turing.pdf https://mdetlefsen.nd.edu/assets/201037/jf.turing.pdf
- Kutta 10y agoI'm fine with classical reasoning and uncomputability, and I acknowledge the intuition for classical truth (it got popular for a reason). But pretty much all the useful classical math can be performed all the same in type theory without significant change, therefore the presence of "useful math" is not an argument for or against type theory and classical foundations. Rather, the main thing is that type theory is just better for formalizing things, and while type theory can conveniently embed classical reasoning and talk about various levels of constructivity, classical foundations is not convenient for constructive reasoning. Math with justification in ZFC is useful in practice, but ZFC itself is mostly not. I would like to see eventually many or event most things formalized, and for that we need fundamentals for more than justification. On the more philosophical side, "philosophers should care about computational complexity", certainly, but it's not like one cannot have gradation in philosophical appeal. Also, it definitely helps to have computation at hand if we want to talk about feasible computation.
- pron 10y ago> therefore the presence of "useful math" is not an argument for or against type theory and classical foundations I wasn't making an argument against type theory (although constructive analysis is quite different from classical analysis) but an argument against a weak argument against classical math. Useful math is very much an argument in favor of a foundation. In fact, it was the winning argument in favor of ZFC, when it was thought that intuitionism would mean radically changing analysis. > Rather, the main thing is that type theory is just better for formalizing things In mechanical proof checkers? Then say that type theory is a more convenient small core for a mechanical proof checker. But you're arguing for constructive math, and constructive math is significantly less convenient for proving many things (and constructivists openly acknowledge that). > ZFC itself is mostly not. Useful for what? Mathematical foundations are not used when doing math unless you're doing formal math. ZFC seems a lot more useful that type theory when, say, teaching mathematical foundations (which, to reiterate, don't really mean the actual foundations of math). > I would like to see eventually many or event most things formalized, and for that we need fundamentals for more than justification. OK, formalizing mathematics is a 100 year old dream and a noble one, but one that may not be shared today by most mathematicians today. Your argument would be clearer and less religious if you said: I want to formalize math mechanically, and currently type theory seems more convenient for that particular task than ZFC. > but it's not like one cannot have gradation in philosophical appeal Sure, but I don't understand the aesthetic gradation. Does a mathematical foundation become more appealing to you the more closely it constrains math to the physical? Why? Would a mathematical foundation that doesn't allow expressing speed quantities greater than the speed of light be appealing to you? I would understand if you said you're an intuitionist, but the intuitionists really did have a problem explaining the effectiveness of classical math (which is simply invalid to them). Classical analysis coincides with constructive analysis on all physically observable propositions and is still easier to work with (at least currently). > Also, it definitely helps to have computation at hand if we want to talk about feasible computation. There is absolutely nothing wrong, missing, or inconvenient with how classical math models computation. Type theory or any other constructive math has no advantage there. It certainly doesn't make talking about feasible computation easier.
- deleted 10y ago