Inverse Galois Problem for Direct Product of Groups












3












$begingroup$


Assume that $G = operatorname{Gal}(L/Bbb{Q}), H = operatorname{Gal}(K/Bbb{Q})$.



Can we construct a field extension of $Bbb Q$ from $L$ and $K$ such that $G times H$ is its Galois group?



For $L cap K = Bbb Q$, the field extension we need is just $LK$. What if $L cap K = M, Bbb Q < M$?



Are there corresponding results for fields other than $Bbb Q$?










share|cite|improve this question











$endgroup$












  • $begingroup$
    Doesn't the product of their minimal polynomials work?
    $endgroup$
    – Tobias Kildetoft
    Dec 3 '18 at 11:20










  • $begingroup$
    In a similar thread on mathoverflow someone said that the Galois group of the corresponding splitting field is a subgroup of the product of the Galois groups we started with.
    $endgroup$
    – Lukas Kofler
    Dec 3 '18 at 11:24






  • 1




    $begingroup$
    @TobiasKildetoft: It cannot work if $Lcap K$ properly contains $mathbb{Q}$: the degrees don't match. Say $L=mathbb{Q}(a)$, $K=mathbb{Q}(b)$ (possible by primitive element theorem). The splitting field of the product of the corresponding irreducible polynomials is $mathbb{Q}(a,b)=LK$. But then $[LK:mathbb{Q}]=[LK:L][L:mathbb{Q}] < [K:mathbb{Q}][L:mathbb{Q}] = |Gtimes H|$.
    $endgroup$
    – Arturo Magidin
    Dec 3 '18 at 17:45






  • 1




    $begingroup$
    It's hard to see how you could do it with $L$ and $K$. For example, take $L=K=mathbb{Q}(sqrt{2})$, so $G=H=C_2$. How could you use $L$ to get an extension of degree $4$ with Galois group $C_2times C_2$?
    $endgroup$
    – Arturo Magidin
    Dec 3 '18 at 17:48










  • $begingroup$
    @ArturoMagidin Yeah, I think I had implicitly assumed that the intersection was trivial. I wonder if there is some way to "twist" a field extension to make it have smaller intersection with some other given extension and at the same time keeping the Galois group the same.
    $endgroup$
    – Tobias Kildetoft
    Dec 3 '18 at 18:08
















3












$begingroup$


Assume that $G = operatorname{Gal}(L/Bbb{Q}), H = operatorname{Gal}(K/Bbb{Q})$.



Can we construct a field extension of $Bbb Q$ from $L$ and $K$ such that $G times H$ is its Galois group?



For $L cap K = Bbb Q$, the field extension we need is just $LK$. What if $L cap K = M, Bbb Q < M$?



Are there corresponding results for fields other than $Bbb Q$?










share|cite|improve this question











$endgroup$












  • $begingroup$
    Doesn't the product of their minimal polynomials work?
    $endgroup$
    – Tobias Kildetoft
    Dec 3 '18 at 11:20










  • $begingroup$
    In a similar thread on mathoverflow someone said that the Galois group of the corresponding splitting field is a subgroup of the product of the Galois groups we started with.
    $endgroup$
    – Lukas Kofler
    Dec 3 '18 at 11:24






  • 1




    $begingroup$
    @TobiasKildetoft: It cannot work if $Lcap K$ properly contains $mathbb{Q}$: the degrees don't match. Say $L=mathbb{Q}(a)$, $K=mathbb{Q}(b)$ (possible by primitive element theorem). The splitting field of the product of the corresponding irreducible polynomials is $mathbb{Q}(a,b)=LK$. But then $[LK:mathbb{Q}]=[LK:L][L:mathbb{Q}] < [K:mathbb{Q}][L:mathbb{Q}] = |Gtimes H|$.
    $endgroup$
    – Arturo Magidin
    Dec 3 '18 at 17:45






  • 1




    $begingroup$
    It's hard to see how you could do it with $L$ and $K$. For example, take $L=K=mathbb{Q}(sqrt{2})$, so $G=H=C_2$. How could you use $L$ to get an extension of degree $4$ with Galois group $C_2times C_2$?
    $endgroup$
    – Arturo Magidin
    Dec 3 '18 at 17:48










  • $begingroup$
    @ArturoMagidin Yeah, I think I had implicitly assumed that the intersection was trivial. I wonder if there is some way to "twist" a field extension to make it have smaller intersection with some other given extension and at the same time keeping the Galois group the same.
    $endgroup$
    – Tobias Kildetoft
    Dec 3 '18 at 18:08














3












3








3





$begingroup$


Assume that $G = operatorname{Gal}(L/Bbb{Q}), H = operatorname{Gal}(K/Bbb{Q})$.



Can we construct a field extension of $Bbb Q$ from $L$ and $K$ such that $G times H$ is its Galois group?



For $L cap K = Bbb Q$, the field extension we need is just $LK$. What if $L cap K = M, Bbb Q < M$?



Are there corresponding results for fields other than $Bbb Q$?










share|cite|improve this question











$endgroup$




Assume that $G = operatorname{Gal}(L/Bbb{Q}), H = operatorname{Gal}(K/Bbb{Q})$.



Can we construct a field extension of $Bbb Q$ from $L$ and $K$ such that $G times H$ is its Galois group?



For $L cap K = Bbb Q$, the field extension we need is just $LK$. What if $L cap K = M, Bbb Q < M$?



Are there corresponding results for fields other than $Bbb Q$?







abstract-algebra group-theory field-theory galois-theory






share|cite|improve this question















share|cite|improve this question













share|cite|improve this question




share|cite|improve this question








edited Dec 3 '18 at 11:30







Lukas Kofler

















asked Dec 3 '18 at 11:15









Lukas KoflerLukas Kofler

1,2552519




1,2552519












  • $begingroup$
    Doesn't the product of their minimal polynomials work?
    $endgroup$
    – Tobias Kildetoft
    Dec 3 '18 at 11:20










  • $begingroup$
    In a similar thread on mathoverflow someone said that the Galois group of the corresponding splitting field is a subgroup of the product of the Galois groups we started with.
    $endgroup$
    – Lukas Kofler
    Dec 3 '18 at 11:24






  • 1




    $begingroup$
    @TobiasKildetoft: It cannot work if $Lcap K$ properly contains $mathbb{Q}$: the degrees don't match. Say $L=mathbb{Q}(a)$, $K=mathbb{Q}(b)$ (possible by primitive element theorem). The splitting field of the product of the corresponding irreducible polynomials is $mathbb{Q}(a,b)=LK$. But then $[LK:mathbb{Q}]=[LK:L][L:mathbb{Q}] < [K:mathbb{Q}][L:mathbb{Q}] = |Gtimes H|$.
    $endgroup$
    – Arturo Magidin
    Dec 3 '18 at 17:45






  • 1




    $begingroup$
    It's hard to see how you could do it with $L$ and $K$. For example, take $L=K=mathbb{Q}(sqrt{2})$, so $G=H=C_2$. How could you use $L$ to get an extension of degree $4$ with Galois group $C_2times C_2$?
    $endgroup$
    – Arturo Magidin
    Dec 3 '18 at 17:48










  • $begingroup$
    @ArturoMagidin Yeah, I think I had implicitly assumed that the intersection was trivial. I wonder if there is some way to "twist" a field extension to make it have smaller intersection with some other given extension and at the same time keeping the Galois group the same.
    $endgroup$
    – Tobias Kildetoft
    Dec 3 '18 at 18:08


















  • $begingroup$
    Doesn't the product of their minimal polynomials work?
    $endgroup$
    – Tobias Kildetoft
    Dec 3 '18 at 11:20










  • $begingroup$
    In a similar thread on mathoverflow someone said that the Galois group of the corresponding splitting field is a subgroup of the product of the Galois groups we started with.
    $endgroup$
    – Lukas Kofler
    Dec 3 '18 at 11:24






  • 1




    $begingroup$
    @TobiasKildetoft: It cannot work if $Lcap K$ properly contains $mathbb{Q}$: the degrees don't match. Say $L=mathbb{Q}(a)$, $K=mathbb{Q}(b)$ (possible by primitive element theorem). The splitting field of the product of the corresponding irreducible polynomials is $mathbb{Q}(a,b)=LK$. But then $[LK:mathbb{Q}]=[LK:L][L:mathbb{Q}] < [K:mathbb{Q}][L:mathbb{Q}] = |Gtimes H|$.
    $endgroup$
    – Arturo Magidin
    Dec 3 '18 at 17:45






  • 1




    $begingroup$
    It's hard to see how you could do it with $L$ and $K$. For example, take $L=K=mathbb{Q}(sqrt{2})$, so $G=H=C_2$. How could you use $L$ to get an extension of degree $4$ with Galois group $C_2times C_2$?
    $endgroup$
    – Arturo Magidin
    Dec 3 '18 at 17:48










  • $begingroup$
    @ArturoMagidin Yeah, I think I had implicitly assumed that the intersection was trivial. I wonder if there is some way to "twist" a field extension to make it have smaller intersection with some other given extension and at the same time keeping the Galois group the same.
    $endgroup$
    – Tobias Kildetoft
    Dec 3 '18 at 18:08
















$begingroup$
Doesn't the product of their minimal polynomials work?
$endgroup$
– Tobias Kildetoft
Dec 3 '18 at 11:20




$begingroup$
Doesn't the product of their minimal polynomials work?
$endgroup$
– Tobias Kildetoft
Dec 3 '18 at 11:20












$begingroup$
In a similar thread on mathoverflow someone said that the Galois group of the corresponding splitting field is a subgroup of the product of the Galois groups we started with.
$endgroup$
– Lukas Kofler
Dec 3 '18 at 11:24




$begingroup$
In a similar thread on mathoverflow someone said that the Galois group of the corresponding splitting field is a subgroup of the product of the Galois groups we started with.
$endgroup$
– Lukas Kofler
Dec 3 '18 at 11:24




1




1




$begingroup$
@TobiasKildetoft: It cannot work if $Lcap K$ properly contains $mathbb{Q}$: the degrees don't match. Say $L=mathbb{Q}(a)$, $K=mathbb{Q}(b)$ (possible by primitive element theorem). The splitting field of the product of the corresponding irreducible polynomials is $mathbb{Q}(a,b)=LK$. But then $[LK:mathbb{Q}]=[LK:L][L:mathbb{Q}] < [K:mathbb{Q}][L:mathbb{Q}] = |Gtimes H|$.
$endgroup$
– Arturo Magidin
Dec 3 '18 at 17:45




$begingroup$
@TobiasKildetoft: It cannot work if $Lcap K$ properly contains $mathbb{Q}$: the degrees don't match. Say $L=mathbb{Q}(a)$, $K=mathbb{Q}(b)$ (possible by primitive element theorem). The splitting field of the product of the corresponding irreducible polynomials is $mathbb{Q}(a,b)=LK$. But then $[LK:mathbb{Q}]=[LK:L][L:mathbb{Q}] < [K:mathbb{Q}][L:mathbb{Q}] = |Gtimes H|$.
$endgroup$
– Arturo Magidin
Dec 3 '18 at 17:45




1




1




$begingroup$
It's hard to see how you could do it with $L$ and $K$. For example, take $L=K=mathbb{Q}(sqrt{2})$, so $G=H=C_2$. How could you use $L$ to get an extension of degree $4$ with Galois group $C_2times C_2$?
$endgroup$
– Arturo Magidin
Dec 3 '18 at 17:48




$begingroup$
It's hard to see how you could do it with $L$ and $K$. For example, take $L=K=mathbb{Q}(sqrt{2})$, so $G=H=C_2$. How could you use $L$ to get an extension of degree $4$ with Galois group $C_2times C_2$?
$endgroup$
– Arturo Magidin
Dec 3 '18 at 17:48












$begingroup$
@ArturoMagidin Yeah, I think I had implicitly assumed that the intersection was trivial. I wonder if there is some way to "twist" a field extension to make it have smaller intersection with some other given extension and at the same time keeping the Galois group the same.
$endgroup$
– Tobias Kildetoft
Dec 3 '18 at 18:08




$begingroup$
@ArturoMagidin Yeah, I think I had implicitly assumed that the intersection was trivial. I wonder if there is some way to "twist" a field extension to make it have smaller intersection with some other given extension and at the same time keeping the Galois group the same.
$endgroup$
– Tobias Kildetoft
Dec 3 '18 at 18:08










2 Answers
2






active

oldest

votes


















2












$begingroup$

It's not possible to do this using no information about $mathbb{Q}$ other than that it's a field, because there are fields $F$ for which the collection of Galois groups of their finite Galois extensions is not closed under taking products. The simplest example is any finite field; here the Galois groups of finite Galois extensions are precisely the finite cyclic groups $C_n$, and e.g. $C_2 times C_2$ is not cyclic. But any quasi-finite field will work, some of which have characteristic $0$, so even that assumption won't save you.



More abstractly, the Galois groups of finite Galois extensions of $F$ are precisely the finite quotients of the absolute Galois group $text{Gal}(overline{F}/F)$. The class of profinite groups which arise in this way is quite diverse, and there's no reason to expect the collection of finite quotients of some profinite group (which for finite fields is the profinite integers $widehat{mathbb{Z}}$) to be closed under products.






share|cite|improve this answer











$endgroup$





















    2












    $begingroup$

    Perhaps a $p$-adic approach could shed some light on some cases where the OP problem has a negative answer. Fix an odd (for simplicity) prime $p$. Given a number field $F$, let $S$ be a finite set of primes of $F$ containing the set $S_p$ of primes above $p$, and consider the maximal pro-$p$-extension $F_S$ of $F$ which is unramified outside $S$, with $G_S(F):=Gal(F_S/F)$. It follows from global CFT that the group $G_S(F)^{ab}$ (obvious notation) is a $mathbf Z_p$-module of finite type, and Leopoldt's conjecture (proved for abelian fields) gives its $mathbf Z_p$-structure: (1) $G_S(F)^{ab}cong mathbf Z_p^{1+c} times T_S(F)$, where $T_S(F)$ denotes the (finite) $mathbf Z_p$-torsion submodule and $c$ is the number of complex primes of $F$. Let us keep the OP notations $K, L, G=Gal(K/mathbf Q), H=Gal(L/mathbf Q)$.



    1) If $F=mathbf Q$, then $T_{S_p}(mathbf Q)=0$ and the decomposition (1) reads $G_S(mathbf Q)^{ab}cong mathbf Z_p times T_S(mathbf Q)$, where $mathbf Z_p$ is the Galois group of the cyclotomic $mathbf Z_p$-extension of $mathbf Q$ (which is $mathbf Q_{S_p}$), and $T_S(mathbf Q)$ is the product over the primes in $T=S-S_p$ of the local $p$-primary groups of roots of unity. This gives explicitly the structure of the finite quotients of $G_S(mathbf Q)^{ab}$ and provides a negative answer to the OP question when it is restricted to abelian $p$-groups. Indeed, we can always enlarge $S$ in order to include the primes of $mathbf Q$ which ramify in $K$ and $L$, and then fix an extension $E/mathbf Q$ with Galois group $cong T_S(mathbf Q)$. Then $K$ (and $L$ as well) is the compositum of a sub-extension of $E$ and a sub-extension of $mathbf Q_{S_p}$; and so will be $K.L$ . But $Gtimes H$ cannot be a quotient of $G_S(mathbf Q)^{ab}$ as described above, even if $Kcap L=mathbf Q$. This is analogous to the example over a finite field given by @Quiaochu Yuan.



    2) The non abelian case is perhaps more "visible" over a general number field $F$. Let $S, S_p, T$ be as before (relatively to $F$). Modulo Leopoldt's conjecture, the following can be proved. Let us say that $F$ is $p$-rational if $T_{S_p}(F)=0$, which is equivalent to saying that $G_{S_p}(F)$ is a free pro-$p$-group. Example: $mathbf Q$ is $p$-rational. Let $tilde F$ be the compositum of all the $mathbf Z_p$-extensions of $F$, so that $tilde F$ is the fixed field of $T_S(F)$; the set $S$ is called primitive if the Frobenius automorphisms $phi_w$ of $Gal(tilde F/F)$ for all $win T$ generate a free $mathbf Z_p$-submodule $Gal(tilde F/F)$ of rank equal to $mid T mid$. If $F$ is $p$-rational and $S$ is primitive, then (2) $G_S(F) cong G_{S_p}(F) * (*_{win T} G_w)$, where * denotes the free product in the category of pro-$p$-groups and $G_w$ is the inertia subgroup of $G_S(F)$ relative to $w$ ((2) can be viewed as the non abelian lift of (1)). This theorem shows that: (i) Any $p$-group whose minimal number of generators is $le 1+c-mid T mid$ can be realized as the Galois group over $F$ of an extension unramified outside $p$; (ii) the OP question has a negative answer, by the same type of argument as in 1).



    NB: The conclusion cannot be extended to extensions which are not $p$-extensions (for a given $p$).One could as previously introduce the maximal extension $E_S$ of $F$ which is unramified outside $S$, but this stops short here. It is not even known whether $Gal(E_S/F)$ is finitely generated.






    share|cite|improve this answer











    $endgroup$













      Your Answer





      StackExchange.ifUsing("editor", function () {
      return StackExchange.using("mathjaxEditing", function () {
      StackExchange.MarkdownEditor.creationCallbacks.add(function (editor, postfix) {
      StackExchange.mathjaxEditing.prepareWmdForMathJax(editor, postfix, [["$", "$"], ["\\(","\\)"]]);
      });
      });
      }, "mathjax-editing");

      StackExchange.ready(function() {
      var channelOptions = {
      tags: "".split(" "),
      id: "69"
      };
      initTagRenderer("".split(" "), "".split(" "), channelOptions);

      StackExchange.using("externalEditor", function() {
      // Have to fire editor after snippets, if snippets enabled
      if (StackExchange.settings.snippets.snippetsEnabled) {
      StackExchange.using("snippets", function() {
      createEditor();
      });
      }
      else {
      createEditor();
      }
      });

      function createEditor() {
      StackExchange.prepareEditor({
      heartbeatType: 'answer',
      autoActivateHeartbeat: false,
      convertImagesToLinks: true,
      noModals: true,
      showLowRepImageUploadWarning: true,
      reputationToPostImages: 10,
      bindNavPrevention: true,
      postfix: "",
      imageUploader: {
      brandingHtml: "Powered by u003ca class="icon-imgur-white" href="https://imgur.com/"u003eu003c/au003e",
      contentPolicyHtml: "User contributions licensed under u003ca href="https://creativecommons.org/licenses/by-sa/3.0/"u003ecc by-sa 3.0 with attribution requiredu003c/au003e u003ca href="https://stackoverflow.com/legal/content-policy"u003e(content policy)u003c/au003e",
      allowUrls: true
      },
      noCode: true, onDemand: true,
      discardSelector: ".discard-answer"
      ,immediatelyShowMarkdownHelp:true
      });


      }
      });














      draft saved

      draft discarded


















      StackExchange.ready(
      function () {
      StackExchange.openid.initPostLogin('.new-post-login', 'https%3a%2f%2fmath.stackexchange.com%2fquestions%2f3023918%2finverse-galois-problem-for-direct-product-of-groups%23new-answer', 'question_page');
      }
      );

      Post as a guest















      Required, but never shown

























      2 Answers
      2






      active

      oldest

      votes








      2 Answers
      2






      active

      oldest

      votes









      active

      oldest

      votes






      active

      oldest

      votes









      2












      $begingroup$

      It's not possible to do this using no information about $mathbb{Q}$ other than that it's a field, because there are fields $F$ for which the collection of Galois groups of their finite Galois extensions is not closed under taking products. The simplest example is any finite field; here the Galois groups of finite Galois extensions are precisely the finite cyclic groups $C_n$, and e.g. $C_2 times C_2$ is not cyclic. But any quasi-finite field will work, some of which have characteristic $0$, so even that assumption won't save you.



      More abstractly, the Galois groups of finite Galois extensions of $F$ are precisely the finite quotients of the absolute Galois group $text{Gal}(overline{F}/F)$. The class of profinite groups which arise in this way is quite diverse, and there's no reason to expect the collection of finite quotients of some profinite group (which for finite fields is the profinite integers $widehat{mathbb{Z}}$) to be closed under products.






      share|cite|improve this answer











      $endgroup$


















        2












        $begingroup$

        It's not possible to do this using no information about $mathbb{Q}$ other than that it's a field, because there are fields $F$ for which the collection of Galois groups of their finite Galois extensions is not closed under taking products. The simplest example is any finite field; here the Galois groups of finite Galois extensions are precisely the finite cyclic groups $C_n$, and e.g. $C_2 times C_2$ is not cyclic. But any quasi-finite field will work, some of which have characteristic $0$, so even that assumption won't save you.



        More abstractly, the Galois groups of finite Galois extensions of $F$ are precisely the finite quotients of the absolute Galois group $text{Gal}(overline{F}/F)$. The class of profinite groups which arise in this way is quite diverse, and there's no reason to expect the collection of finite quotients of some profinite group (which for finite fields is the profinite integers $widehat{mathbb{Z}}$) to be closed under products.






        share|cite|improve this answer











        $endgroup$
















          2












          2








          2





          $begingroup$

          It's not possible to do this using no information about $mathbb{Q}$ other than that it's a field, because there are fields $F$ for which the collection of Galois groups of their finite Galois extensions is not closed under taking products. The simplest example is any finite field; here the Galois groups of finite Galois extensions are precisely the finite cyclic groups $C_n$, and e.g. $C_2 times C_2$ is not cyclic. But any quasi-finite field will work, some of which have characteristic $0$, so even that assumption won't save you.



          More abstractly, the Galois groups of finite Galois extensions of $F$ are precisely the finite quotients of the absolute Galois group $text{Gal}(overline{F}/F)$. The class of profinite groups which arise in this way is quite diverse, and there's no reason to expect the collection of finite quotients of some profinite group (which for finite fields is the profinite integers $widehat{mathbb{Z}}$) to be closed under products.






          share|cite|improve this answer











          $endgroup$



          It's not possible to do this using no information about $mathbb{Q}$ other than that it's a field, because there are fields $F$ for which the collection of Galois groups of their finite Galois extensions is not closed under taking products. The simplest example is any finite field; here the Galois groups of finite Galois extensions are precisely the finite cyclic groups $C_n$, and e.g. $C_2 times C_2$ is not cyclic. But any quasi-finite field will work, some of which have characteristic $0$, so even that assumption won't save you.



          More abstractly, the Galois groups of finite Galois extensions of $F$ are precisely the finite quotients of the absolute Galois group $text{Gal}(overline{F}/F)$. The class of profinite groups which arise in this way is quite diverse, and there's no reason to expect the collection of finite quotients of some profinite group (which for finite fields is the profinite integers $widehat{mathbb{Z}}$) to be closed under products.







          share|cite|improve this answer














          share|cite|improve this answer



          share|cite|improve this answer








          edited Dec 3 '18 at 22:18

























          answered Dec 3 '18 at 22:10









          Qiaochu YuanQiaochu Yuan

          278k32584920




          278k32584920























              2












              $begingroup$

              Perhaps a $p$-adic approach could shed some light on some cases where the OP problem has a negative answer. Fix an odd (for simplicity) prime $p$. Given a number field $F$, let $S$ be a finite set of primes of $F$ containing the set $S_p$ of primes above $p$, and consider the maximal pro-$p$-extension $F_S$ of $F$ which is unramified outside $S$, with $G_S(F):=Gal(F_S/F)$. It follows from global CFT that the group $G_S(F)^{ab}$ (obvious notation) is a $mathbf Z_p$-module of finite type, and Leopoldt's conjecture (proved for abelian fields) gives its $mathbf Z_p$-structure: (1) $G_S(F)^{ab}cong mathbf Z_p^{1+c} times T_S(F)$, where $T_S(F)$ denotes the (finite) $mathbf Z_p$-torsion submodule and $c$ is the number of complex primes of $F$. Let us keep the OP notations $K, L, G=Gal(K/mathbf Q), H=Gal(L/mathbf Q)$.



              1) If $F=mathbf Q$, then $T_{S_p}(mathbf Q)=0$ and the decomposition (1) reads $G_S(mathbf Q)^{ab}cong mathbf Z_p times T_S(mathbf Q)$, where $mathbf Z_p$ is the Galois group of the cyclotomic $mathbf Z_p$-extension of $mathbf Q$ (which is $mathbf Q_{S_p}$), and $T_S(mathbf Q)$ is the product over the primes in $T=S-S_p$ of the local $p$-primary groups of roots of unity. This gives explicitly the structure of the finite quotients of $G_S(mathbf Q)^{ab}$ and provides a negative answer to the OP question when it is restricted to abelian $p$-groups. Indeed, we can always enlarge $S$ in order to include the primes of $mathbf Q$ which ramify in $K$ and $L$, and then fix an extension $E/mathbf Q$ with Galois group $cong T_S(mathbf Q)$. Then $K$ (and $L$ as well) is the compositum of a sub-extension of $E$ and a sub-extension of $mathbf Q_{S_p}$; and so will be $K.L$ . But $Gtimes H$ cannot be a quotient of $G_S(mathbf Q)^{ab}$ as described above, even if $Kcap L=mathbf Q$. This is analogous to the example over a finite field given by @Quiaochu Yuan.



              2) The non abelian case is perhaps more "visible" over a general number field $F$. Let $S, S_p, T$ be as before (relatively to $F$). Modulo Leopoldt's conjecture, the following can be proved. Let us say that $F$ is $p$-rational if $T_{S_p}(F)=0$, which is equivalent to saying that $G_{S_p}(F)$ is a free pro-$p$-group. Example: $mathbf Q$ is $p$-rational. Let $tilde F$ be the compositum of all the $mathbf Z_p$-extensions of $F$, so that $tilde F$ is the fixed field of $T_S(F)$; the set $S$ is called primitive if the Frobenius automorphisms $phi_w$ of $Gal(tilde F/F)$ for all $win T$ generate a free $mathbf Z_p$-submodule $Gal(tilde F/F)$ of rank equal to $mid T mid$. If $F$ is $p$-rational and $S$ is primitive, then (2) $G_S(F) cong G_{S_p}(F) * (*_{win T} G_w)$, where * denotes the free product in the category of pro-$p$-groups and $G_w$ is the inertia subgroup of $G_S(F)$ relative to $w$ ((2) can be viewed as the non abelian lift of (1)). This theorem shows that: (i) Any $p$-group whose minimal number of generators is $le 1+c-mid T mid$ can be realized as the Galois group over $F$ of an extension unramified outside $p$; (ii) the OP question has a negative answer, by the same type of argument as in 1).



              NB: The conclusion cannot be extended to extensions which are not $p$-extensions (for a given $p$).One could as previously introduce the maximal extension $E_S$ of $F$ which is unramified outside $S$, but this stops short here. It is not even known whether $Gal(E_S/F)$ is finitely generated.






              share|cite|improve this answer











              $endgroup$


















                2












                $begingroup$

                Perhaps a $p$-adic approach could shed some light on some cases where the OP problem has a negative answer. Fix an odd (for simplicity) prime $p$. Given a number field $F$, let $S$ be a finite set of primes of $F$ containing the set $S_p$ of primes above $p$, and consider the maximal pro-$p$-extension $F_S$ of $F$ which is unramified outside $S$, with $G_S(F):=Gal(F_S/F)$. It follows from global CFT that the group $G_S(F)^{ab}$ (obvious notation) is a $mathbf Z_p$-module of finite type, and Leopoldt's conjecture (proved for abelian fields) gives its $mathbf Z_p$-structure: (1) $G_S(F)^{ab}cong mathbf Z_p^{1+c} times T_S(F)$, where $T_S(F)$ denotes the (finite) $mathbf Z_p$-torsion submodule and $c$ is the number of complex primes of $F$. Let us keep the OP notations $K, L, G=Gal(K/mathbf Q), H=Gal(L/mathbf Q)$.



                1) If $F=mathbf Q$, then $T_{S_p}(mathbf Q)=0$ and the decomposition (1) reads $G_S(mathbf Q)^{ab}cong mathbf Z_p times T_S(mathbf Q)$, where $mathbf Z_p$ is the Galois group of the cyclotomic $mathbf Z_p$-extension of $mathbf Q$ (which is $mathbf Q_{S_p}$), and $T_S(mathbf Q)$ is the product over the primes in $T=S-S_p$ of the local $p$-primary groups of roots of unity. This gives explicitly the structure of the finite quotients of $G_S(mathbf Q)^{ab}$ and provides a negative answer to the OP question when it is restricted to abelian $p$-groups. Indeed, we can always enlarge $S$ in order to include the primes of $mathbf Q$ which ramify in $K$ and $L$, and then fix an extension $E/mathbf Q$ with Galois group $cong T_S(mathbf Q)$. Then $K$ (and $L$ as well) is the compositum of a sub-extension of $E$ and a sub-extension of $mathbf Q_{S_p}$; and so will be $K.L$ . But $Gtimes H$ cannot be a quotient of $G_S(mathbf Q)^{ab}$ as described above, even if $Kcap L=mathbf Q$. This is analogous to the example over a finite field given by @Quiaochu Yuan.



                2) The non abelian case is perhaps more "visible" over a general number field $F$. Let $S, S_p, T$ be as before (relatively to $F$). Modulo Leopoldt's conjecture, the following can be proved. Let us say that $F$ is $p$-rational if $T_{S_p}(F)=0$, which is equivalent to saying that $G_{S_p}(F)$ is a free pro-$p$-group. Example: $mathbf Q$ is $p$-rational. Let $tilde F$ be the compositum of all the $mathbf Z_p$-extensions of $F$, so that $tilde F$ is the fixed field of $T_S(F)$; the set $S$ is called primitive if the Frobenius automorphisms $phi_w$ of $Gal(tilde F/F)$ for all $win T$ generate a free $mathbf Z_p$-submodule $Gal(tilde F/F)$ of rank equal to $mid T mid$. If $F$ is $p$-rational and $S$ is primitive, then (2) $G_S(F) cong G_{S_p}(F) * (*_{win T} G_w)$, where * denotes the free product in the category of pro-$p$-groups and $G_w$ is the inertia subgroup of $G_S(F)$ relative to $w$ ((2) can be viewed as the non abelian lift of (1)). This theorem shows that: (i) Any $p$-group whose minimal number of generators is $le 1+c-mid T mid$ can be realized as the Galois group over $F$ of an extension unramified outside $p$; (ii) the OP question has a negative answer, by the same type of argument as in 1).



                NB: The conclusion cannot be extended to extensions which are not $p$-extensions (for a given $p$).One could as previously introduce the maximal extension $E_S$ of $F$ which is unramified outside $S$, but this stops short here. It is not even known whether $Gal(E_S/F)$ is finitely generated.






                share|cite|improve this answer











                $endgroup$
















                  2












                  2








                  2





                  $begingroup$

                  Perhaps a $p$-adic approach could shed some light on some cases where the OP problem has a negative answer. Fix an odd (for simplicity) prime $p$. Given a number field $F$, let $S$ be a finite set of primes of $F$ containing the set $S_p$ of primes above $p$, and consider the maximal pro-$p$-extension $F_S$ of $F$ which is unramified outside $S$, with $G_S(F):=Gal(F_S/F)$. It follows from global CFT that the group $G_S(F)^{ab}$ (obvious notation) is a $mathbf Z_p$-module of finite type, and Leopoldt's conjecture (proved for abelian fields) gives its $mathbf Z_p$-structure: (1) $G_S(F)^{ab}cong mathbf Z_p^{1+c} times T_S(F)$, where $T_S(F)$ denotes the (finite) $mathbf Z_p$-torsion submodule and $c$ is the number of complex primes of $F$. Let us keep the OP notations $K, L, G=Gal(K/mathbf Q), H=Gal(L/mathbf Q)$.



                  1) If $F=mathbf Q$, then $T_{S_p}(mathbf Q)=0$ and the decomposition (1) reads $G_S(mathbf Q)^{ab}cong mathbf Z_p times T_S(mathbf Q)$, where $mathbf Z_p$ is the Galois group of the cyclotomic $mathbf Z_p$-extension of $mathbf Q$ (which is $mathbf Q_{S_p}$), and $T_S(mathbf Q)$ is the product over the primes in $T=S-S_p$ of the local $p$-primary groups of roots of unity. This gives explicitly the structure of the finite quotients of $G_S(mathbf Q)^{ab}$ and provides a negative answer to the OP question when it is restricted to abelian $p$-groups. Indeed, we can always enlarge $S$ in order to include the primes of $mathbf Q$ which ramify in $K$ and $L$, and then fix an extension $E/mathbf Q$ with Galois group $cong T_S(mathbf Q)$. Then $K$ (and $L$ as well) is the compositum of a sub-extension of $E$ and a sub-extension of $mathbf Q_{S_p}$; and so will be $K.L$ . But $Gtimes H$ cannot be a quotient of $G_S(mathbf Q)^{ab}$ as described above, even if $Kcap L=mathbf Q$. This is analogous to the example over a finite field given by @Quiaochu Yuan.



                  2) The non abelian case is perhaps more "visible" over a general number field $F$. Let $S, S_p, T$ be as before (relatively to $F$). Modulo Leopoldt's conjecture, the following can be proved. Let us say that $F$ is $p$-rational if $T_{S_p}(F)=0$, which is equivalent to saying that $G_{S_p}(F)$ is a free pro-$p$-group. Example: $mathbf Q$ is $p$-rational. Let $tilde F$ be the compositum of all the $mathbf Z_p$-extensions of $F$, so that $tilde F$ is the fixed field of $T_S(F)$; the set $S$ is called primitive if the Frobenius automorphisms $phi_w$ of $Gal(tilde F/F)$ for all $win T$ generate a free $mathbf Z_p$-submodule $Gal(tilde F/F)$ of rank equal to $mid T mid$. If $F$ is $p$-rational and $S$ is primitive, then (2) $G_S(F) cong G_{S_p}(F) * (*_{win T} G_w)$, where * denotes the free product in the category of pro-$p$-groups and $G_w$ is the inertia subgroup of $G_S(F)$ relative to $w$ ((2) can be viewed as the non abelian lift of (1)). This theorem shows that: (i) Any $p$-group whose minimal number of generators is $le 1+c-mid T mid$ can be realized as the Galois group over $F$ of an extension unramified outside $p$; (ii) the OP question has a negative answer, by the same type of argument as in 1).



                  NB: The conclusion cannot be extended to extensions which are not $p$-extensions (for a given $p$).One could as previously introduce the maximal extension $E_S$ of $F$ which is unramified outside $S$, but this stops short here. It is not even known whether $Gal(E_S/F)$ is finitely generated.






                  share|cite|improve this answer











                  $endgroup$



                  Perhaps a $p$-adic approach could shed some light on some cases where the OP problem has a negative answer. Fix an odd (for simplicity) prime $p$. Given a number field $F$, let $S$ be a finite set of primes of $F$ containing the set $S_p$ of primes above $p$, and consider the maximal pro-$p$-extension $F_S$ of $F$ which is unramified outside $S$, with $G_S(F):=Gal(F_S/F)$. It follows from global CFT that the group $G_S(F)^{ab}$ (obvious notation) is a $mathbf Z_p$-module of finite type, and Leopoldt's conjecture (proved for abelian fields) gives its $mathbf Z_p$-structure: (1) $G_S(F)^{ab}cong mathbf Z_p^{1+c} times T_S(F)$, where $T_S(F)$ denotes the (finite) $mathbf Z_p$-torsion submodule and $c$ is the number of complex primes of $F$. Let us keep the OP notations $K, L, G=Gal(K/mathbf Q), H=Gal(L/mathbf Q)$.



                  1) If $F=mathbf Q$, then $T_{S_p}(mathbf Q)=0$ and the decomposition (1) reads $G_S(mathbf Q)^{ab}cong mathbf Z_p times T_S(mathbf Q)$, where $mathbf Z_p$ is the Galois group of the cyclotomic $mathbf Z_p$-extension of $mathbf Q$ (which is $mathbf Q_{S_p}$), and $T_S(mathbf Q)$ is the product over the primes in $T=S-S_p$ of the local $p$-primary groups of roots of unity. This gives explicitly the structure of the finite quotients of $G_S(mathbf Q)^{ab}$ and provides a negative answer to the OP question when it is restricted to abelian $p$-groups. Indeed, we can always enlarge $S$ in order to include the primes of $mathbf Q$ which ramify in $K$ and $L$, and then fix an extension $E/mathbf Q$ with Galois group $cong T_S(mathbf Q)$. Then $K$ (and $L$ as well) is the compositum of a sub-extension of $E$ and a sub-extension of $mathbf Q_{S_p}$; and so will be $K.L$ . But $Gtimes H$ cannot be a quotient of $G_S(mathbf Q)^{ab}$ as described above, even if $Kcap L=mathbf Q$. This is analogous to the example over a finite field given by @Quiaochu Yuan.



                  2) The non abelian case is perhaps more "visible" over a general number field $F$. Let $S, S_p, T$ be as before (relatively to $F$). Modulo Leopoldt's conjecture, the following can be proved. Let us say that $F$ is $p$-rational if $T_{S_p}(F)=0$, which is equivalent to saying that $G_{S_p}(F)$ is a free pro-$p$-group. Example: $mathbf Q$ is $p$-rational. Let $tilde F$ be the compositum of all the $mathbf Z_p$-extensions of $F$, so that $tilde F$ is the fixed field of $T_S(F)$; the set $S$ is called primitive if the Frobenius automorphisms $phi_w$ of $Gal(tilde F/F)$ for all $win T$ generate a free $mathbf Z_p$-submodule $Gal(tilde F/F)$ of rank equal to $mid T mid$. If $F$ is $p$-rational and $S$ is primitive, then (2) $G_S(F) cong G_{S_p}(F) * (*_{win T} G_w)$, where * denotes the free product in the category of pro-$p$-groups and $G_w$ is the inertia subgroup of $G_S(F)$ relative to $w$ ((2) can be viewed as the non abelian lift of (1)). This theorem shows that: (i) Any $p$-group whose minimal number of generators is $le 1+c-mid T mid$ can be realized as the Galois group over $F$ of an extension unramified outside $p$; (ii) the OP question has a negative answer, by the same type of argument as in 1).



                  NB: The conclusion cannot be extended to extensions which are not $p$-extensions (for a given $p$).One could as previously introduce the maximal extension $E_S$ of $F$ which is unramified outside $S$, but this stops short here. It is not even known whether $Gal(E_S/F)$ is finitely generated.







                  share|cite|improve this answer














                  share|cite|improve this answer



                  share|cite|improve this answer








                  edited Dec 5 '18 at 5:38

























                  answered Dec 4 '18 at 13:56









                  nguyen quang donguyen quang do

                  8,4791723




                  8,4791723






























                      draft saved

                      draft discarded




















































                      Thanks for contributing an answer to Mathematics Stack Exchange!


                      • Please be sure to answer the question. Provide details and share your research!

                      But avoid



                      • Asking for help, clarification, or responding to other answers.

                      • Making statements based on opinion; back them up with references or personal experience.


                      Use MathJax to format equations. MathJax reference.


                      To learn more, see our tips on writing great answers.




                      draft saved


                      draft discarded














                      StackExchange.ready(
                      function () {
                      StackExchange.openid.initPostLogin('.new-post-login', 'https%3a%2f%2fmath.stackexchange.com%2fquestions%2f3023918%2finverse-galois-problem-for-direct-product-of-groups%23new-answer', 'question_page');
                      }
                      );

                      Post as a guest















                      Required, but never shown





















































                      Required, but never shown














                      Required, but never shown












                      Required, but never shown







                      Required, but never shown

































                      Required, but never shown














                      Required, but never shown












                      Required, but never shown







                      Required, but never shown







                      Popular posts from this blog

                      Probability when a professor distributes a quiz and homework assignment to a class of n students.

                      Aardman Animations

                      Are they similar matrix