im trying to show that the galois group of $x^5+sx^3+t$ over $\mathbb{Q(s,t)}$ is $S_5$. By just looking at the discriminant, it has to be $S_5$ or $F_{20}$. I know i could distinguish between those 2 using the weber sextic resolvent, but this is a degree 6 polynomial with 2 parameters and i have no idea how to show that there is no root in $\mathbb{Q(s,t)}$. Is there another way to see this more easily?

Something else i would like to know is: I think it s true under some conditions that if u specialize s,t in a polynomial as above the galois group can only become smaller. Does anyone know if this is a basic fact or needs more work?

Thx in advance

  • 1
    $\begingroup$ It is not always $S_5$. For $s=10, t=\pm 5$ the group is $F_{20}$. $\endgroup$ – i. m. soloveichik Jul 27 '15 at 15:57
  • $\begingroup$ If $s=0$ then the Galois group is not $S_5$. $\endgroup$ – i. m. soloveichik Jul 27 '15 at 16:05
  • 2
    $\begingroup$ If $s=15, t=81$ then the group is $D_5$. $\endgroup$ – i. m. soloveichik Jul 27 '15 at 16:07

I can only answer your second question.

What you're looking at geometrically is the finite cover of degree 5: $$f : \text{Spec }\mathbb{Q}[s,t][x]/(x^5+sx^3+t)\rightarrow \text{Spec }\mathbb{Q}[s,t]$$

Away from the ramification locus (this is basically the values of $(s,t)$ which make the polynomial $x^5+sx^3+t$ have multiple roots) in $\text{Spec }\mathbb{Q}[s,t]$, the map is unramified, though not galois. The galois group you're referring to is also called the "monodromy group" (this is geometric language). Let $U\subset\text{Spec }\mathbb{Q}[s,t]$, and let $X := f^{-1}(U)$. By the theory of the fundamental group, the cover $X\rightarrow U$ corresponds to a homomorphism

$$\pi_1(U)\longrightarrow S_5$$

The image of this map is called the monodromy group (for you, the galois group).

Specializing to a particular value of $s,t$ corresponds to pulling back the cover $X\rightarrow U$ via a map $p : \text{Spec }\mathbb{Q}\rightarrow U$. The pulled-back cover corresponds to the homomorphism $$\pi_1(\text{Spec }\mathbb{Q})\stackrel{p_*}{\longrightarrow}\pi_1(U)\longrightarrow S_5$$

The galois group (monodromy group) of the specialization is then just the image of the above homomorphism, which can obviously cannot be larger than the image of the original homomorphism $\pi_1(U)\longrightarrow S_5$.

Therefore, if you can find a single value of $s,t$ for which the galois group is all of $S_5$, then you can conclude that the galois group of the unspecified polynomial is also $S_5$.

  • $\begingroup$ Also, can't you do a base change over $\Bbb C$ to reduce it to a geometry problem ? $\endgroup$ – mercio Jul 27 '15 at 19:00
  • $\begingroup$ @mercio Sure...though I'm not entirely sure how that would help.. $\endgroup$ – oxeimon Jul 27 '15 at 19:05

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.