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

Prove sqrt(3) is irrational

  1. Jan 12, 2019 #1
    1. The problem statement, all variables and given/known data
    Prove sqrt(3) is irrational

    2. Relevant equations


    3. The attempt at a solution
    (a/b)^2 = 3 assume a/b is in lowest form

    a^2 = 3b^2

    so a^2 is of form 3n

    whenever n is even, a^2 will be even => a will be even whenever n is even

    so a is of form 2l whenever n is even

    => 4l^2 = 3b^2 => 2l/sqrt(3) = b

    rewrite as 2((l*sqrt(3))/3) = b

    contradiction. Since both a and b are divisible by 2 whenever n is even, sqrt(3) must be irrational for even n, This implies that sqrt(3) is irrational
     
  2. jcsd
  3. Jan 12, 2019 #2

    fresh_42

    User Avatar
    2018 Award

    Staff: Mentor

    To what?
    What if ##n## is odd?

    I think you can do it a lot easier. Do you know what a prime number is, and that ##3## is prime?
     
  4. Jan 12, 2019 #3
    Contradicts a/b being a rational in lowest form.

    I don't see why proving its irrational for even n to not be sufficient enough...

    But yes I know that 3 is prime
     
  5. Jan 12, 2019 #4

    fresh_42

    User Avatar
    2018 Award

    Staff: Mentor

    Why should it be sufficient? ##n## odd is a possibility, or not? And I do not see any contradiction, because all you know (by assumption) that ##\sqrt{3}## is rational, so what follows from an expression like ##\dfrac{2\cdot l \cdot \sqrt{3}}{3}=b## other than ##b \in \mathbb{Q}##.

    What about my other question: Do you know what a prime number is?
     
  6. Jan 12, 2019 #5
    Hmm I guess you're right. I'll continue with the possibility of n being odd..

    As for the contradiction. a^2 = 3b^2 => a^2 is of form 3n => a is of form 3n. If a is of form 3n => 3b^n that b will be even as well, and since I assumed (a/b) is in lowest form there is a contradiction since I have two evens numbers.

    Yes a prime number is one whose factors are only 1 and itself. I think you are saying it will be easier to solve this problem if I can show that (a/b)^2 = 3 that if b does not equal one then 3 is not a prime so I would have a contradiction?
     
  7. Jan 12, 2019 #6

    fresh_42

    User Avatar
    2018 Award

    Staff: Mentor

    I don't see this.
    That is wrong. Your definition is the one for an irreducible number, not a prime number. For the integers the two terms are identical, i.e. the prime numbers are exactly the irreducible ones and vice versa. However, the true definition of a prime number is of a great help here:

    A number is ##p## is prime, if it is not ##\pm 1## and if ##p \,|\,a\cdot b \Longrightarrow p\,|\,a \text{ or } p\,|\,b##. With this definition, the proof is more or less straightforward and you don't need cases and expressions at the end, which are composed by elements you want to examine: ##\sqrt{3}\,.##

    Another possibility is to use the prime factor decompositions of ##a## and ##b##.
     
  8. Jan 12, 2019 #7
    Sorry I do not know why I wrote that in the first quote, when that is not what I meant. I meant to write:

    a^2 = 3b^2 => if b^2 is even, then a^2 is also even, => a and b are both even, thus (a/b) cannot be in lowest form.

    I will try to finish this problem using the definition of prime.
     
  9. Jan 12, 2019 #8

    fresh_42

    User Avatar
    2018 Award

    Staff: Mentor

    This is correct. But the case ##b## odd could be a problem.
    Just start with ##3\,|\,a^2## and so ##3\,|\,a## and so on.
     
  10. Jan 13, 2019 #9

    Math_QED

    User Avatar
    Homework Helper

    You can't call a definition wrong when it is equivalent with the one you have in mind, even if yours is more general as it works in the more general settings of rings. The context was very clear here.
     
  11. Jan 13, 2019 #10

    fresh_42

    User Avatar
    2018 Award

    Staff: Mentor

    I can call it wrong because it is. That they are equivalent is a theorem, not given a priori. I don't like it, if students are taught mistakes. Me, too, has learnt it wrong at school, but what is the matter with doing it right? It has nothing to do with "more general", it is a different property.

    You won't define compact by bounded and closed, just because it is the same in ##\mathbb{R}^n##, won't you?
     
  12. Jan 13, 2019 #11

    Math_QED

    User Avatar
    Homework Helper

    It is perfectly fine to define compactness in ##\mathbb{R}^n## as closed and bounded. In fact, there are a couple of books that do this. The author can then prove it is equivalent with the cover definition and then use this as general definition in topological/metric spaces. Whether this is pedagogically sound, is another question. But definitely not wrong. The reason this is done is that covers are hard to grasp at first, so giving an alternative "more intuitive" definition helps students.

    In the same way, I think the definition of the OP is more intuitive than the one you propose. Probably the reason that most books use this definition. A quick search also shows that wikipedia uses the OP's definition.
     
  13. Jan 13, 2019 #12

    fresh_42

    User Avatar
    2018 Award

    Staff: Mentor

    Sorry, but I'm against the method of doing something wrong and correct it afterwards. It implies the hidden assumption that students / kids are not smart enough to do it correctly. Make a test and start a thread by: "1. Given ##X\subseteq \mathbb{R}^n## compact, show that ... 2. Compact = bounded and closed 3. ...." And use a statement which is easy to proof be covering compactness and hard by the other one, as it is here the case.
    I bet it won't take two posts, until this definition will be corrected here!

    Sorry, but irreducibility is not prime! That it is the same for some rings, doesn't justify to introduce it wrong and correct it afterwards. I hate this didactic method: "(1st year) Today we learn addition and subtraction. .... No, 3-5 is not possible, you cannot give away 5 bonbons if you have only 3." and then "(some time later) Today we learn that 3-5 = -2" I really do hate this method!

    Irreducibility is not prime! Stop assuming kids are stupid.
     
  14. Jan 13, 2019 #13

    fresh_42

    User Avatar
    2018 Award

    Staff: Mentor

    And even if I accepted the wrong definition at schools, that is no reason to accept it on PF!
     
  15. Jan 13, 2019 #14

    Math_QED

    User Avatar
    Homework Helper

    It isn't a matter of thinking kids are stupid.

    Why not start with general topological spaces and then treat metric spaces as a special case? Simply because people wouldn't see the point.

    Or maybe, let us start with measure theory before doing probability theory.

    Sometimes, you must make people confident in a certain area, way of thinking before they can appreciate/understand the whole picture. This has nothing to do with being stupid or not.

    That's why we teach kids things IN SMALL STEPS. One thing at a time.
     
  16. Jan 13, 2019 #15

    fresh_42

    User Avatar
    2018 Award

    Staff: Mentor

    I do not agree. You contradict yourself:
    The former statement implies that you know what's building up confidence and what does not. This is a hidden way of saying too complicated which in return hides too stupid.

    But apart from this discussion about pedagogic: Do you really think irreducibility is necessary to understand primality? I may agree with you on measure theory, although I'm less convinced than you are, but I do not agree, that primality is significantly more difficult than irreducibility. The comparison with measure theory is a bit far fetched.
     
  17. Jan 13, 2019 #16

    Math_QED

    User Avatar
    Homework Helper

    I think ##p|ab \implies p|a \lor p|b## is more difficult to comprehend than the alternate definition, yes. But that wasn't the point. The point was that both definitions are good (at least in this context!). And maybe you are right that there are mathematical reasons to prefer the one over the other, but to call it wrong is simply wrong itself. Mathematically, the OP gave a correct definition (again, in this context).

    I disagree that I contradicted myself. "Too complicated" can mean "too soon (YET)" or "not mature enough (YET)". Not "too stupid to understand".

    Another example: starting with calculus before diving in real analysis. It illustrates the case well imo.

    Let us agree to disagree.
     
  18. Jan 13, 2019 #17

    fresh_42

    User Avatar
    2018 Award

    Staff: Mentor

    Agreed. The more as I've chosen "wrong" as adjective to break the automatism with irreducibility rather than to start a debate upon its linguistic implications.
     
  19. Jan 13, 2019 #18
    a^2 = 3b^2 => a = 3((b^2)/a) => 3|a

    so a is of form 3n

    9n^2 = 3((b^2)
    =>
    3n^2 = b^2 => 3((n^2)/b) = b => 3|b

    So (a/b) isnt in lowest form, therefore sqrt3 is irrational
     
  20. Jan 13, 2019 #19
    Fresh would this proof be sufficient enough?
     
  21. Jan 13, 2019 #20

    fresh_42

    User Avatar
    2018 Award

    Staff: Mentor

    I would avoid fractions here, as it requires to demonstrate that ##b\,|\,n^2## which needs a second thought. Why is ##\dfrac{n^2}{b}## an integer?

    Instead you could just repeat the primality argument: ##9n^2=3b^2 \Longrightarrow 3n^2=b^2 \Longrightarrow 3\,|\,b## and if three divides both, you are done.
     
Share this great discussion with others via Reddit, Google+, Twitter, or Facebook

Have something to add?