Given a = $e^\frac{2i\pi}{18}$, field extension $|\Bbb Q(a):\Bbb Q| = 6$, I was wondering what the basis for $\Bbb Q(a)$ being, or how the elements are represented. I found the minimal polynomial to be $x^6-x^3+1$, which is degree 6 in $\Bbb Q$, so I'm guessing the basis for $\Bbb Q(a)$ would $\{1,a,a^2,a^3,a^4,a^5\}$. Does that make any sense? Normally the bases I have dealt with were degree $2$ so the elements were just like $a+bi$ or something with basis $\{1,i\}$

  • $\begingroup$ Since $f(x) = x^6-x^3+1$ is irreducible and $f(a) = 0$ you know $\mathbb{Q}[a] = \{ \sum_{n=0}^5 c_n a^n, c_n \in \mathbb{Q}\}$. For proving that $\mathbb{Q}(a) = \mathbb{Q}[a]$ you need to show that $\mathbb{Q}[x]/f(x)$ is a field. With $\phi(p(x)) = p(a)$ you have an isomorphism $\phi : \mathbb{Q}[x]/f(x) \to \mathbb{Q}[a]$, ie. $\mathbb{Q}[a]$ is a field. $\endgroup$ – reuns Apr 7 '17 at 5:07
  • $\begingroup$ Oh I see. Thank you very much. This might be much, but how would I attempt to find the galois group for Q(a):Q? If you don't want to answer thats fine. $\endgroup$ – Tesuji Apr 7 '17 at 5:58
  • $\begingroup$ The Galois group of cyclotomic extensions is easy : $\sigma(\zeta_9)=\zeta_9^k$ where $gcd(k,9) = 1$. There are $\varphi(9) = 6$ such automorphisms. $\endgroup$ – reuns Apr 7 '17 at 6:18
  • $\begingroup$ Oh man we didn't learn about the zeta stuff yet. Is there a link you can provide me to read up on this? Despite your comment, I can't list the different $\sigma$ elements, since our professor wants a product table and then show what group this is isomorphic too.. If you can't, regardless, thank you very much for your help. $\endgroup$ – Tesuji Apr 7 '17 at 6:34
  • $\begingroup$ $\zeta_9 = a = e^{2 i \pi / 9}$ and $\sigma$ is a ring morphism $\mathbb{Q}[a] \to \mathbb{Q}[a]$, so $\sigma(a) = a^k$ means $\sigma(\sum_{n=0}^5 c_n a^n) =\sum_{n=0}^5 c_n \sigma(a^n)= \sum_{n=0}^5 c_n a^{nk}$ $\endgroup$ – reuns Apr 7 '17 at 6:39

Your finding of the minimal polynomial and the basis are fine, it proves that you have well understood the setup. A trick to retrieve the Galois group is to find another root for $x^6-x^3+1$. Some trial and error gives you that $-a^2$ is such a root. So there is an element $\sigma$ of the Galois group that maps $a \mapsto -a^2$, so we can then ask what is the image $\sigma(-a^2)$? It's $--a^4$ and we verify that it is indeed another root! we can go on and find the three other roots: $\sigma(-a^4) = -a^8 = a^2-a^5$, $\sigma(a^2-a^5) = \sigma(a^2)-\sigma(a^5) = -a^4 + a^{10} = -a^4 -a$ and $\sigma(-a^4 -a) = \sigma(-a^4) - \sigma(a) = a^2 + a^5 + a^2$. If we change basis to these roots $$ \{a, -a^2,-a^4,a^2-a^5,a^4-a, a^5\} $$ we have an easy expression for $\sigma$ as a matrix.
When we revert this matrix to the original basis $\{1,a,a^2,a^3,a^4,a^5\}$ this yields the matrix $$ \sigma = \begin{pmatrix}1&0&0&1&0 &0\\ 0&0&0&0&0 &1\\ 0&-1&0&0&-1&0\\ 0&0&0&-1 & 0&0\\ 0&0&1&0&0&0\\ 0&0&0&0&1&0 \end{pmatrix} $$ This matrix generates the Galois group, a cyclic group of order $6$.


Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.