Dismiss Notice
Join Physics Forums Today!
The friendliest, high quality science and math community on the planet! Everyone who loves science is here!

Is mathematics empirical?

  1. Dec 2, 2004 #1
    There is no absolute truth in mathematics and logics. There are many consistent mathematical systems, consistent in the sense that they are not breaking the basic rules. But there is no absolute way of knowing that the basic rules are true. And there are limitless basic rules that can be chosen. Different basic rules gives different systems, like the intuitionist logic,predicate logic or fuzzy logic.

    So why then is mathematics/logic interesting and why are some systems studied instead of others? I would argue it is because mathematics is empirical. There are some logical and mathematical systems that are more true than others. Those systems that more closely follow the real world are more true than others.

    So truth in mathematics is ultimately derived from physics. Those mathematics that gives physicists more accurate models are more true.
  2. jcsd
  3. Dec 2, 2004 #2

    matt grime

    User Avatar
    Science Advisor
    Homework Helper

    I wouldn't use the word true, and I would never use a comparitive form for it in this sense. It [the set of rules/things we use] is not "more true", but it is more reasonable, and better yet more *useful* than anything else we have. That is why we use the things we do in the models we make: it works. If there were a better one, we'd use that.
  4. Dec 2, 2004 #3
    I tend to agree, but then I see little difference between "truth" and "useful".

    Since there is no known ultimate cause and therefore "truth", I think the best we can get is more "useful". In the sense that it allows us to better manipulate the world.

    Regarding the current prefered system, predicate logic. Suppose physicists start making better predictions using some theories based on intuitionist logic. Would not this require that most of current mathematical proofs be examined again, and accepted or discarded depending on how they fare in this new logical system that seems better suited to the real world?
    Last edited: Dec 2, 2004
  5. Dec 4, 2004 #4
    How do you derive from physics, the truth that [tex] \pi [/tex] is transcendental?

    Can you provide an example of what you have in mind? I seem to recall that intuitionistic theories are weaker than classical theories, any theorem of the former is a theorem of the later, but I may be wrong. An example of what you have in mind might make things clearer.
  6. Dec 5, 2004 #5
    Pi is transcendental according to the basic rules chosen. Physics (and in extension the real world) do not affect the theorems that can be proven when the basic rules have been chosen. The real world affects which basic rules to chose.

    We can look at an example higher up in hierarchy. Euclidean geometry was long considered the only possible system and that it also reflected the real world, restricting mathematical research to within this area. Then is was discovered that non-euclidean geometry could be constructed in logically consistent way. And also that physicists can make better predictions assuming non-Euclidean geometry. Since then, much research has been done in non-Euclidean geometry, probably much more than in Euclidean geometry.

    Regarding intuitionist logic, let us assume that it is found that the most basic fundamental particles seems to follow the basic rules of intuitionist logic better than predicate logic. That when constructing theories and when using
    Reductio_ad_absurdum one gets poor predictions. But if instead if using intuitionist logic one can construct simple theories that give good predictions. I think this would lead to a paradigm shift in mathematics, away from predicate to intuitionist logic.
  7. Dec 5, 2004 #6

    matt grime

    User Avatar
    Science Advisor
    Homework Helper

    OK, next challenge - can we have one physical theory constructed using reductio ad absurdum?

    Surely, since physics is attempting to explain what is there, as opposed to what one may show may exist, irrespective of its apparent use, or lack of, physics is reasonably intuitionist.

    And, as someone has pointed out, anything that can be proved true in intuitionistic logic is true in predicate logic.

    Fuzzy logic already is used in some of the applied science, as are other logic systems.
    Last edited: Dec 5, 2004
  8. Dec 6, 2004 #7


    User Avatar
    Science Advisor

    Why do you choose physics specifically? If calculus is used in economics would you say that it works because of physics?? As far as your "basic rules" are concerned, the whole point of mathematics is that all statements in mathematics are of the form "if A then B". One doesn't have to know IF A is true or not, only follow the consequences IF it were true.
  9. Dec 9, 2004 #8
    I say physics since this since some people here question whether economics or psychology is science and I do not want to discuss that in this thread. However, with physics I mean empirical science. My point is that mathematics is also an empirical science, but one step further removed from the real world than for example physics. Another step away would be logic. And logic is not automatically true, many different kinds of logics can be constructed. For example:

    Yes. But if you follow the the sequence backward you will find axioms that have been picked among many possible other axioms. There is no ultimate principle from which everything can be derived. And again, my point is that is the real world that determines which axioms to pick. Those that help makes better predictions are chosen. Chosen in the sense that mathematicians concentrate most of their efforts on those mathematics.

    One could maybe argue that predicate logic have already failed in physics. That wave-particle duality would be an example.

    I do not know enough physics or logic to know if QM is consistent with predicate logic. But if not, another logic would seems to be better choice since QM seems to make the best current predictions in its field of explanation.

    Another interesting idea. What if the it turns out that that real world is discrete, that there is no infinity or continuity in nature. Would this mean that the current concept of limit and differentiation would have to be reworked?
    Last edited: Dec 9, 2004
  10. Dec 10, 2004 #9

    matt grime

    User Avatar
    Science Advisor
    Homework Helper

    What has wave particle duality to do with predicate logic? Apart from your opinion that anything that *ought* to be dichotomic is the same as predicate calculus. The answer by the way is nothing.

    As was pointed out in another thread, discrete models of space currently don't work, and for the umpteenth time limits and derivatives are mathematics, they are not physical objects! The nature of the real world would not invalidate it, in fact it is quite obviously independent of the nature of the physical world in some sense.

    QM in case you didnt' notice has not stopped Newton's Laws of Motion from still being taught and used.
  11. Dec 10, 2004 #10


    User Avatar
    Science Advisor

    Then I think the problem is that you have only studied mathematics that was designed for applications (and you seem to be using the word "physics" in a way I find peculiar). I know many forms of mathematics in which the axioms (what you call the "basic rules") have nothing to do with "the real world", for example the various finite geometries. Of course, those are not then used in applications.

    I think you are looking through the "wrong end of the telescope". Mathematics, in its truest sense, does not derive axioms from "the real world". Of course, when you want to apply mathematics to "the real world", you choose the particular types of mathematics whose axioms do, in fact, appear to correspond to the real world.
    Sometimes, of course, it happens that newly discovered properties of the "real world" turn out to correspond to axioms of forms of mathematics that had been developed previously without reference to the real world. I'm thinking in particular of the application to General Relativity of Riemannian spaces which were develope before the "basic rules" of General Relativity were known.
  12. Dec 10, 2004 #11
    Actually there seem to have been a quite interesting discussion about this:

    I am not saying that the universe is discrete or continuous, in my limited understanding that is one point where QM and GR disagree.

    It is true that the current limits and derivatives would not be logically falsified if the universe proved discrete. They are consistent according to the basic rules chosen. My point was that if physicists could make better predictions with limits and derivatives defined somewhat differently, they and mathematicians would abandon the current limits and derivatives. Hypothetically, a discrete universe could simulate a continuous universe at the macro level while being discrete at the micro level. The interesting predictions would be in the borderland, where the discrete and continuous mix. If the new limits and derivatives would make better predictions here than the old which ignore the discrete, they would be better suited for the real world.

    Again, I am not saying this will be. Just an example of how mathematics could change which axioms to study, similar to the shift from Euclidean to non-Euclidean before.
  13. Dec 10, 2004 #12
    Yes, there are areas of mathematics that seem to have no connection with the real world today But most mathematics are used and studied because it is useful in the real world. If it was not useful, why should society spend resources on it? Then society could as well spend the resources on teaching dead languages. So if mathematicians want to keep their status and pay, they have to choose to study systems built from axioms suited to the real world.

    Regarding the second paragraph: Sometimes theory is before experiments. Sometimes experiments are before theory. Still, in the end both experiments and theory must fit together.
  14. Dec 10, 2004 #13

    matt grime

    User Avatar
    Science Advisor
    Homework Helper

    Well, that's an interesting opinion, but factually flawed from my experience of being employed by mathematics departments (I would estimate in my current department less than 1/2 do maths that is applied to the real world, and it's predominantly an applied maths department too; this is what engineering departments are for). There is nothing to suggest all maths must eventually translate to the physical world (or it'll have it's funding cut). Intellectual rigour need not have a practical direct use.

    Sometimes theory never meets experiment, thank goodness.

    Interestingly I appear to be an impossible object in your opinion - a mathematician who's paid, by the state, to do mathematics that isn't practically applicable. At least it's nice when people tell me what I "have" to be.

    Good to see you're so against all the higher aspects of humanity at least - utility should be one of the last things we think about in lots of cases. As the old saying goes, it'll be a great day when the schools are fully funded by the state and the military has to hold a bring and buy say to pay for its aircraft carriers.
  15. Dec 10, 2004 #14


    User Avatar
    Staff Emeritus
    Gold Member
    Dearly Missed

    matt, I want to bring up another sense of empirical. Take a mathematician who sets out to create something new - that is what math research is. The guy has in his mind a conspectus of prior mathematics in his area and he considers the relationships he knows about. Sometimes ther is a little "mix-and-match" novelty he can do, sometimes it's "monkey see monkey do" analogy with other math, and sometimes by building out a train of thought with these humble techniques he can come to a point where insight supervenes and he makes a discovery. (I want to apologize here to all the excellent women mathematicians for my choice of pronouns).

    Now isn't a creation path like this "empirical" in its relationship to the world of existing mathematics, which to a mathematician is as real and present as the sensual world?
  16. Dec 10, 2004 #15

    matt grime

    User Avatar
    Science Advisor
    Homework Helper

    I missed the post replying to me quoting back my post.

    Lots of quantum things in general (such as quantum groups, cohomology etc) use quantum in a certain mathematical sense.

    I still stand by my assertion that wave particle duality and predicate logic are not necessarily linked, but that is a philosophical position about the nature of logic (and mathematics) not being physical entitites. I would take it to say that the ideas of QM have led to a new way of thinking about things by introducing measurements of failure to commute (deformations), rather than it being the other way round, but again that is a personal opinion about philosophy.

    However it is all hypothetical isn't it?

    Mathematicians are obtuse people, they will study what they want to. Some will make connections to the real world some won't.

    The development of mathematics and physics has to varying degrees been linked throughout its course. It diverged strongly in the 40's, but they are starting to come together now, it appears.

    I don't think calculus wll die (something I'm a little sad about...)

    It may interest you to know that there is a non-analytic concept of derivative; we algebraists have been using it for years. It uses something called the space of Dual Numbers.

    Self Adjoint: I haven't actually used the word empirical directly, or addressed what may or may not be meant by it.
  17. Dec 10, 2004 #16
    What does it mean for an arbitrary physical theory to be "consistent" with an arbitrary mathematical theory?

  18. Dec 10, 2004 #17

    You may not have noticed but we live in a world of limited resources. These should be allocated where they are most useful. One must always think of the alternative use, for example saving lives in health-care with more money and intelligent people.

    That many must work so that a few can spend their time in useless games seems absurd. And if theoretical mathematics is of no use in the real world, then it is no better than spending time with a complex computer game. Interesting maybe for the individual, but not something others should be forced to support. Regarding the state, is is noted for its inefficiency in all areas due to lack of competition.

    But of course theoretical mathematics with no apparent utility today is much better than a computer game. Exactly because it may be of use in the future. So that is one of the better ways the state spends its money, compared to many others which often even have a negative effect. But note that since it is possible to construct and study countless useless mathematics which could consume all resources in the world, some calculation regarding future use should be made when giving money to today useless mathematics.

    Regarding the use of mathematics, it is of course used today in most of society. And most of those using mathematics are certainly not using axioms with no connection to the real world.
  19. Dec 10, 2004 #18
    Two arbitrarily chosen theories are mostly useless without some connection to the real world, for example data from experiments.

    If we have data, then we can construct and test a physical theory using a mathematical theory. Note that the mathematical theory will restrict which physical theories are possible to make or affect how good the predictions are or how complex the physical theory must be to fit the data.

    It would be quite difficult to construct a good physical theory that fits the data if one were forced to work only within the mathematical theory of Euclidean geometry.
  20. Dec 10, 2004 #19
    What makes you think THE theory of predicate calculus is related to QM in some obvious way? As far as I can tell, substituting "theory of recursive functions" or "theory of formal languages" in place of "theory of predicate calculus" is an arbitrary choice.

    We all know that mathematics can be used as a tool to build physical theories. What I'm wondering is; what exactly is the process you used to come to the determination that predicate calculus is false based on the empirical evidence that suggests that light has a dual nature?
  21. Dec 11, 2004 #20

    matt grime

    User Avatar
    Science Advisor
    Homework Helper

    Can we get back to the original post? It seems that this isn't a discussion about mathematics or anything in particular of that nature. It is purely a debate about what people mean by "true". And in particular comparitive ideas.
Share this great discussion with others via Reddit, Google+, Twitter, or Facebook