1. Limited time only! Sign up for a free 30min personal tutor trial with Chegg Tutors
    Dismiss Notice
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: Extension fields

  1. Dec 15, 2009 #1
    Two problems from my abstract algebra class...

    Let K be the algebraic closure of a fi eld F and suppose E is a field such that  F [tex]F \subseteq E \subseteq K[/tex]. Then is K the algebraic closure of E?

    Let [tex]n[/tex] be a natural number with [tex]n\geq2[/tex], and suppose that [tex]\omega[/tex] is a complex nth root on unity. Is there a formula for [tex]\left[\mathbb{Q}(\omega) : \mathbb{Q}\right][/tex] ?


    To 1), I must be missing something really silly, because it seems to me like it is obviously the case that K is also the algebraic closure of E, and that the proof should be easy. But I simply can't think of anything.

    To 2) I would say no, but I am not exactly sure that I understand the question. For example, if n=8, then [tex]e^{i2\pi/8}[/tex] is a complex 8th root of unity such that [tex]\left[\mathbb{Q}(e^{i2\pi/8}) : \mathbb{Q}\right]=4[/tex]. However, [tex]i[/tex] is also an 8th root of unity, but [tex]\left[\mathbb{Q}(i) : \mathbb{Q}\right]=2[/tex]. Thus, for a given n, there is not necessarily a formula. Does this sound right?
  2. jcsd
  3. Dec 16, 2009 #2
    1) Well for K to be the algebraic closure of E you must show:
    a) K/E is algebraic.
    b) If [itex]g(x) \in E[x][/itex], then g(x) splits completely in K. (HINT: Remember that K is an algebraic closure of itself so [itex]h(x) \in K[x][/itex] imply that h(x) splits completely in K).

    2) Well [itex]\omega[/itex] is a specific nth root of unity so it's acceptable for your formula to behave differently when given [itex]e^{i2\pi/8}[/itex] and when given [itex]e^{i\pi/8}[/itex]. You should probably look for a formula of the form:
    [tex]\left[\mathbb{Q}\left(e^{ik\pi/n}\right) \, : \, \mathbb{Q} \right] = f(n,k)[/tex]
    so the formula can depend on both n and k, not just n.
  4. Dec 16, 2009 #3
    Ok, for 2), I've got the formula

    \left[\mathbb{Q}\left(e^{ik\pi/n}\right) \, : \, \mathbb{Q} \right] = \phi\left(\frac{n}{\gcd(n,k)}\right)

    where phi is Euler's totient funciton.

    Is that right?

    Now I just have to prove it....
Share this great discussion with others via Reddit, Google+, Twitter, or Facebook