Currently I am going through Representation theory of semi-direct products by Reyes. At the begining of the article, the author defines the semidirect product as follows.

$G = H \cdot B$ is a semidirect product where $H$ is a normal subgroup of $G$ and $B$ is isomorphic with a group of automorphisms of $H$.

This doesn't match with the definition given here which is:

Semidirect products.

We say that $G$ is (isomorphic to) a semidirect product of $M$ by $N$ if and only if there exist subgroups $H$ and $K$ of $G$ such that:

  • $H\cong M$ and $K\cong N$;
  • $H\triangleleft G$;
  • $H\cap K=\{e\}$;
  • $G=HK$.

Because, in the second definition $H$ is not required to be isomophic with a group of automorphism of $K$. So, can I say that the first definition is not a general definition of semidirect product?


See "outer" and "inner" semi-direct products in https://en.wikipedia.org/wiki/Semidirect_product

  • $\begingroup$ so it looks, Reyes was talking about the outer semidirect product, right? $\endgroup$ – Omar Shehab Oct 23 '16 at 22:40
  • 1
    $\begingroup$ Yes, and the two constructions are closely related. $\endgroup$ – David Towers Oct 24 '16 at 7:19
  • $\begingroup$ could you please explain how they are related? $\endgroup$ – Omar Shehab Oct 24 '16 at 18:15
  • 1
    $\begingroup$ A group $G$ is an "inner" semidirect product of $H$ and $K$ if and only if it is isomorphic to an "outer" semidirect product. "Outer" means giving an actual construction of the semidirect product; "inner" means recognizing it in the wild. $\endgroup$ – Ravi Fernando Oct 25 '16 at 3:39
  • 1
    $\begingroup$ Also a slight comment on Reyes's line: when talking about (outer) semidirect products, we need a homomorphism from $K = B$ to $Aut(H)$, but this homomorphism does not need to be injective; i.e. it does not need to identify $K$ as a subgroup of $Aut(H)$. For example, a direct product $H \times K$ is the case where $K \to Aut(H)$ is the trivial homomorphism. This is definitely a special case of semidirect products, but it seems like a stretch to call $K$ "isomorphic with a group of automorphisms of $H$" when they all act as the identity. $\endgroup$ – Ravi Fernando Oct 25 '16 at 3:43

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.