I'm trying to find the Galois group of the polynomial $p(X)= X^5+X+1$ over $\mathbb Q$.

First, one notes that, if $\omega$ is a primitive cubic root of unity, then it is a root of $p(X)$. So, $X^2+X+1$ is a factor of $p(X)$. We have the following factorisation: \begin{equation} p(X) = (X^2+X+1)(X^3-X^2+1). \end{equation} Now, the derivative of $p(X)$ is $p'(X) = X^4+1$, which is always positive. Hence, $p(X)$ has only one real root, which is a root $\alpha$ of $X^3-X^2+1$. Let me denote by $\eta$ and $\overline{\eta}$ the other two complex roots of $X^3-X^2+1$. Then, a splitting field of $p(X)$ over $\mathbb Q$ is clearly given by \begin{equation} K = \mathbb Q(\alpha, \omega, \eta). \end{equation} Next, I have to compute the degree $[K : \mathbb Q]$ of the extension. It is simple to prove that $[\mathbb Q(\alpha, \eta) : \mathbb Q] = 6$: in fact, $\mathbb Q(\alpha, \eta)$ is a splitting field of $X^3-X^2+1$ over $\mathbb Q$, and this polynomial has only one real root. So, I try to compute as follows: \begin{equation} [K : \mathbb Q] = [K : \mathbb Q(\alpha,\eta)][\mathbb Q(\alpha, \eta) : \mathbb Q]. \end{equation} Here comes my problem: clearly, the degree $[K : \mathbb Q(\alpha,\eta)]$ is $1$ or $2$. The point is to understand whether $\omega \in \mathbb Q(\alpha, \eta)$ or not, and I'm unable to give an answer. I think that actually $[K : \mathbb Q(\alpha,\eta)] = 2$, and that eventually one can prove that the Galois group $\mathrm{Gal}(K | \mathbb Q)$ is isomorphic to $S_3 \times C_2$.

How can I prove this?

  • $\begingroup$ See math.stackexchange.com/a/38898/589 and math.stackexchange.com/questions/45893/…. $\endgroup$ – lhf Jun 6 '14 at 16:30
  • 4
    $\begingroup$ Nicely presented and worked out question. +1 $\endgroup$ – DonAntonio Jun 6 '14 at 16:32
  • 1
    $\begingroup$ Can anyone explain why for the splitting field we only need to adjoin $\eta$? I don't see how we can generate $\overline\eta$ from that. $\endgroup$ – Jack M Jun 6 '14 at 16:38
  • 1
    $\begingroup$ $\overline{\eta}$ is the complex conjugate of $\eta$. Adding $\eta$ to $\mathbb Q(\alpha)$ is the same as adding the root of the discriminant of the quadratic equation whose roots are $\eta$ and $\overline{\eta}$, which is the same as adding $\overline{\eta}$. $\endgroup$ – Francesco Genovese Jun 6 '14 at 16:47
  • $\begingroup$ Related: math.stackexchange.com/questions/1375747 $\endgroup$ – Watson Dec 26 '16 at 13:07

The only quadratic subextension of $\Bbb Q \subset \Bbb Q(\alpha,\eta)$ is given by the subfield fixed by the index $2$ subgroup $A_3$ of $S_3$. It is generated by $(\alpha-\eta)(\eta-\bar{\eta})(\bar{\eta}-\alpha)$, which is the square root of the discriminant $\Delta$ of $X^3-X^2+1$

A computation gives $\Delta = 23$. Since $\Bbb Q(\sqrt{23})$ and $\Bbb Q(\sqrt{-3})$ are different extensions, $\sqrt{-3} \notin \Bbb Q(\sqrt {23})$ and so $\sqrt{-3} \notin \Bbb Q(\alpha,\eta)$


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.