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

Homework Help: Irrationality Proofs

  1. Jun 23, 2010 #1
    1. The problem statement, all variables and given/known data

    a) Prove that [tex]\sqrt{3}, \sqrt{5}, \sqrt{6}[/tex] are irrational. Hint: To treat [tex]\sqrt{3}[/tex], for example, use the fact that every integer is of the form [tex]3n[/tex] or [tex]3n + 1[/tex] or [tex]3n + 2[/tex]. Why doesn't this proof work for [tex]\sqrt{4}[/tex]?

    b) Prove that [tex]2 ^ 1/3[/tex] and [tex]3 ^ 1/3[/tex] are irrational.

    2. Relevant equations

    Can't think of any, really.

    3. The attempt at a solution

    a) I futzed around with fractions involving [tex]3n[/tex] and [tex]3n + 1[/tex] and so forth but I never really got anywhere. I figure that the proof must have something to do with 3, or [tex]\sqrt{3}^2[/tex], but I can't tell exactly what.

    b) I actually got something here: I reasoned that [tex]\sqrt[3]{2} = (\sqrt[3]{2}^2)\sqrt{2}[/tex]. Since we know that [tex]\sqrt{2}[/tex] is irrational, and that (rational) x (irrational) is irrational unless one of the terms is [tex]0[/tex], and that (rational) x (rational) is rational, we can conclude that it is impossible for [tex]\sqrt[3]{2}[/tex] to be rational because if it was, then a rational number ([tex]\sqrt[3]{2}^2[/tex]) multiplied by an irrational number ([tex]\sqrt{2}[/tex]) would equal a rational number, which here cannot be possible. Is this a proof by contradiction?
    Last edited: Jun 23, 2010
  2. jcsd
  3. Jun 23, 2010 #2


    Staff: Mentor

    What you meant to say but your LaTeX script wasn't quite right was this:
    Prove that [tex]2 ^ {1/3}[/tex] and [tex]3 ^ {1/3}[/tex] are irrational.

    If your exponent inside a [ tex] tag takes two or more characters, surround them with braces - { }.

    You can also use cube roots, using \sqrt[3]{ ... }, like this:
    Prove that [tex]\sqrt[3]{2}[/tex] and [tex]\sqrt[3]{3}[/tex] are irrational.

  4. Jun 23, 2010 #3
    Sorry about that. I went back and fixed the [tex]\sqrt[3]{2}[/tex] in the second half but forgot about the first!
  5. Jun 23, 2010 #4


    Staff: Mentor

    To prove that, say, sqrt(2) is irrational, assume the opposite: that it is rational and so can be written as a/b, with no factors in common in a and b.

    sqrt(2) = a/b ==> 2 = a2/b2

    You should arrive at a contradiction, which means that your assumption that sqrt(2) is rational must not be true.
  6. Jun 23, 2010 #5
    Oh yes, I know that (Spivak actually provides what is essentially the same proof). I'm just looking to see if my reasoning is correct [b) in "The attempt at a solution"]
  7. Jun 23, 2010 #6


    Staff: Mentor

    In b, you have [tex]\sqrt[3]{2} = (\sqrt[3]{2}^2)\sqrt{2}[/tex]
    Why would you think this? The left side is about 1.4422; the right side is about 2.2449.
    I don't get what you're trying to do here.
  8. Jun 23, 2010 #7
    Oh dear. It's an arithmetic error that I overlooked (I multiplied two of the exponents when I should have added them).

    I guess there's a reason Spivak's proof looked so much longer than mine!

    Should probably turn attention back to a), then.
  9. Jun 24, 2010 #8

    Gib Z

    User Avatar
    Homework Helper

    This doesn't use the hint, but you could generalize the proof in Spivak for sqrt 2 to any non-perfect square. Keep an eye out for which step its crucial that its not a perfect square.
  10. Jun 24, 2010 #9
    The step that most sticks out to me is the one in which Spivak asserts that [tex]p^2 = 2q^2[/tex]. If 2 were some perfect square, then the equation would look like [tex]p^2 = (aq)^2[/tex] where [tex]a[/tex] is some integer, which would allow one to solve for [tex]p/q[/tex]. But I'm puzzled as to how I could generalize that proof, since it seems to rest on the fact that the numbers involved are even, which wouldn't work for 3.

    In other words, I'd end up with [tex]p^2 = 3q^2[/tex]. All this seems to tell me is that [tex]p[/tex] and [tex]q[/tex] are both even or both odd. But whereas Spivak's proof for 2 relied on [tex]p[/tex] and [tex]q[/tex] being even, and so having a common factor of 2, that doesn't seem to be possible with [tex]p[/tex] and [tex]q[/tex] being odd.
  11. Jun 24, 2010 #10
    You can also look at the factorization in prime numbers. If you have:

    sqrt[p] = r/s

    where p is a prime number and r and s are integers that have no divisirs in common, then you have:

    p = r^2/s^2 ---------->

    r^2 = p s^2

    If you were to factor both sides and count (by multiplicity) how many prime factors there are on each side, what would you find?
Share this great discussion with others via Reddit, Google+, Twitter, or Facebook