Characterization of Truth in Mathematics
Date: 11/07/2003 at 00:43:48 From: Kiki Subject: Characterization of Truth in Mathematics I am interested in researching the correspondence between mathematics and reality. One of the relevant questions I need to answer is "What does it mean for a mathematical statement to be true - from the viewpoint of a working mathematician?" In other words, if a non- mathematician were to ask a mathematician what it means for a mathematical fact or idea to be true, how would the mathematician reply? My feeling is that the mathematician would say that a mathematical statement is true just in case it adheres to the rules of logic, i.e. just in case it can be logically deduced from other established truths of mathematics. But then, where do the rules of logic come from and what makes the rules of logic true? So, my question is really two-fold. I am looking for a simple, "textbook" answer to this question as opposed to a long, involved philosophical discussion about the nature of mathematical truth, logic, etc. To put it another way, in an encyclopedia of mathematics, what would the brief entry under "mathematical truth" say? (Are there any such encyclopedias or books that could answer my question?) Thanks for your help. - Kiki
Date: 11/07/2003 at 10:18:05 From: Doctor Vogler Subject: Re: Characterization of Truth in Mathematics Kiki, Your feeling is right, but "logically deduced from other established truths of mathematics" might cause you to wonder where you start. Mathematics begins (logically) from a set of assumptions that we call axioms. For example, we assume that the set of positive integers (1, 2, 3, ...) is infinite, is in a progression (the first number is 1 and then every number has a "next" number), and is the smallest such set. This gives us a place to start, and then we use these axioms to prove other things, such as that every subset of positive integers has a smallest number. Mathematics continues with definitions. Actually, you could consider the above to be the definition of natural numbers (which it is), as definitions often come equipped with axioms. We define <name> to be a set of <something> which satisfies these <axioms>. For example, after defining natural numbers, we define addition and multiplication, and then we define prime numbers. Then we prove things about what we've defined, such as that there are infinitely many prime numbers, and that every number can be factored uniquely into prime factors. That's how we mathematicians view "mathematical truth." How does it relate to reality? Well, reality doesn't actually come into play in the logical axioms and definitions, so it only relates because reality follows many of the laws of mathematics. It relates because there really is a set of "natural numbers" that obeys the three axioms I mentioned previously. - Doctor Vogler, The Math Forum http://mathforum.org/dr.math/
Date: 11/07/2003 at 11:08:31 From: Doctor Ian Subject: Re: Characterization of Truth in Mathematics Hi Kiki, Nice to hear from you again! Our first conversation, a few years back, made it into the Dr. Math archives: Understanding Mathematics http://mathforum.org/library/drmath/view/52289.html It mentions a few different books and articles that would be helpful in researching your question. One that isn't mentioned, which might also be useful, is _The Character of Physical Law_, by Richard Feynman, which is essentially a book-length exploration of the question "Why does mathematics seem to do such a great job of explaining what happens in the world?" I've checked all the resources I have, and none of them provides the kind of concise, dictionary-length entry that you're looking for. I suspect that's because the authors and editors of those resources realize that an attempt to provide one would generate more questions than it would answer. Having said that, I'll go ahead and state that the questions [W]here do the rules of logic come from and what makes the rules of logic true? have a single answer: 'Agreement'. Mathematicians have agreed on a particular set of rules for deducing theorems from axioms, in much the same way that chess players have agreed on a particular set of rules for playing chess. The subtle but important insight explored by Lewis Carroll in his dialog, "What the Tortoise Said to Achilles", http://www.ditext.com/carroll/tortoise.html is that there is nothing that _compels_ a particular mathematician to accept the rules of logic that are being used by anyone else. Any mathematician is free to make up his own rules. All that means is that he's playing a different game than everyone else. The essential difference between mathematical truth and scientific truth lies in the direction of the reasoning. In science, there is (we assume) some set of rules that defines 'truth'; and we try to guess what they are by observing outcomes. In math, we _know_ what the rules are, because we get to make them up; and we then try to see what outcomes follow from those rules. Another way to express this is that truths in mathematics are 'formal', that is to say, self-contained. No mathematical truth exists independently of a set of axioms, and a set of agreed-upon rules for deducing theorems from axioms. They all come together as a system; and strictly speaking, they have no necessary connection to anything outside the system. Applying mathematics to the real world is ultimately just a special case of analogical reasoning. We notice a resemblance between relationships that exist between numbers, and relationships that exist between things in the world, and we take advantage of that resemblance to do our reasoning using numbers, which are cheap, plentiful, and very well-behaved. Then we translate our conclusions back to the world, and hope that they still apply. But the operative word is 'hope'. All analogies break down when we try to make use of attributes that exist in one of the analogs, but not the other. If we rely on attributes of numbers (e.g., that intervals of numbers can be subdivided indefinitely) that are not shared by the things we're using numbers to model, then our mathematical truths don't correspond to physical truths. It doesn't make the mathematical truths any less true! It just means that our analogy doesn't stretch as far as we'd like it to. I hope this helps! - Doctor Ian, The Math Forum http://mathforum.org/dr.math/
Date: 11/08/2003 at 22:05:21 From: Kiki Subject: Thank you (Characterization of Truth in Mathematics) Thank you Doctors Ian and Vogler, for your detailed responses! Good to hear from you again too, Doctor Ian. What I want to do with this research project is to "zoom in" on the connection point between mathematics and the real world. I suspect that there are two types of correspondences: the type exemplified by the correspondence between the facts of Euclidean geometry and physical objects and the type exemplified by the correspondence between differential equations and the processes they describe. For example, the mathematics of Newtonian physics would be of this latter type. What do you think? Does my idea make sense? - Kiki
Date: 11/10/2003 at 08:31:49 From: Doctor Ian Subject: Re: Thank you (Characterization of Truth in Mathematics) Hi Kiki, That sounds interesting, but be careful about the way your use of language (and particularly of metaphor) shapes how you think about the problem. For example, do you think there's one 'point of connection' between math and the real world? What would be the attributes of such a point? We already know that Newtonian physics is wrong, so you may want to choose a different correspondence as an example. Within physics, you have at least two choices: relativistic physics, and quantum physics, each of which appears to be at odds with the other, in that the former requires space to be continuous, while the latter is inconsistent with that requirement. I also wonder if you're leaving some correspondences out of the picture. For example, what about the correspondence between the facts of arithmetic and the apparently discrete nature of objects at the atomic level? (For example, atoms are made from discrete numbers of nucleons, which are made from discrete numbers of quarks; electrons arrange themselves into discrete orbitals; and so on.) Also, probability seems to play a large role in processes at the nuclear level. I'm not sure that fits neatly into either geometry or differential equations. As to whether your idea makes sense, I'd have to know more about what your idea _is_. For example, by 'correspondence', do you mean something different from 'analogy'? If I arrange some physical objects in a circle, they don't really form a circle in the Euclidean sense. For one thing, circles are made of points, and points don't have extent. But objects in the world certainly have extent! So what I have is a rough (inexact) analogy between a purely geometric 'object' and an arrangement of physical objects in the world. By reasoning about the geometric object, I can draw conclusions about the physical one, although - because the analogy isn't exact - the conclusions may not be valid. Similarly, if I strike the surface of a drum, I can draw an analogy between the drum and a mathematical surface subject to certain boundary conditions. By reasoning about what would happen to the surface under particular deformations, I can draw conclusions about what will happen to the drum if I strike it with a stick. But - becuse the analogy isn't exact - the conclusions may not be valid. Apart from integer arithmetic, which seems to describe things like conservation of charge, it's hard to think of a mathematical concept that doesn't break down at some point when used to describe natural phenomena. If you haven't yet seen Stephen Wolfram's book, _A New Kind of Science_, you might want to take a look at it. The thesis of the book is that as we continue to look more closely at nature, we'll find that the analogy between (1) continuous mathematical spaces and equations and (2) what's 'really' going on, will become increasingly strained. If I understand him correctly, he's suggesting that the 'proper' representation of nature will ultimately be not mathematical, but algorithmic. It sounds like you've found a very interesting topic! I suspect that the hardest part of pursuing it will be to frame your question in a way that will make the answer intelligible to others, as well as to yourself: Is Geometry a Language? http://mathforum.org/library/drmath/view/55427.html But that's an important step. Don't skip it. I hope this helps! - Doctor Ian, The Math Forum http://mathforum.org/dr.math/
Date: 11/14/2003 at 00:16:18 From: Kiki Subject: Characterization of Truth in Mathematics Hi. To clarify my ideas: I want to research the relationship between mathematics and its applications to reality (the spatio-temporal world surrounding us). More specifically, I want to examine the relationship between: a set of mathematical concepts and their use in answering a question about a real-world phenomenon (object, event or process). Particularly, for any application of a set of mathematical concepts, I want to examine the correspondence that exists/is made between the true mathematical propositions comprising those concepts and the real- world phenomenon that is to be understood. I believe that there are two types of correspondences, which I've termed "opaque" and "transparent." (1) "Transparent" refers to correspondences in which the phenomenon is actually an instantiation of the mathematical concept(s) (or, equivalently, in which the mathematical concept is an abstraction of the phenomenon): e.g. Triangular objects in the physical world can be viewed as instantiations of the idealized triangles of Euclidean geometry, and hence, their geometrical properties can be studied simply by studying the geometrical properties of idealized triangles. Another excellent example is the application of knot theory to the study of DNA strands in biology: DNA strands can be seen as instantiations of idealized mathematical knots; their knotting/tangling properties can be studied simply through studying the corresponding idealized knots. (2) "Opaque" refers to correspondences in which the mathematics merely describes or encodes certain properties of the phenomenon, but is not a literal abstraction of it. For example, the growth of a population of rabbits can be described with a differential equation, but a population of rabbits is in no way an instantiation of a differential equation. I believe that these two types completely characterize all correspondences between mathematical ideas and their applications. Is this clearer? Is there anything I haven't accounted for? Thanks for your help!
Date: 02/06/2004 at 15:29:34 From: Doctor Ian Subject: Re: Characterization of Truth in Mathematics Hi Kiki, Sorry for taking so long to get back to you! To begin with, I would disagree with the statement that A population of rabbits is in no way an instantiation of a differential equation. Certainly an individual rabbit isn't an instantiation of a differential equation. But when you talk about a 'population' of rabbits, you're not talking about a physical object. You're talking about an abstraction. In particular, you're talking about the cardinality of an abstract set, which changes over time - i.e., a function of a continuous independent variable. A differential equation is precisely a description of a class of such functions. So I think that as you refine the precision of your definitions, you may find that the distinction between 'transparent' and 'opaque' may evaporate. Also, offhand, I don't see where logic would fit into your scheme. - Doctor Ian, The Math Forum http://mathforum.org/dr.math/
Search the Dr. Math Library:
Ask Dr. MathTM
© 1994- The Math Forum at NCTM. All rights reserved.