Mathematics Stack Exchange is a question and answer site for people studying math at any level and professionals in related fields. Join them; it only takes a minute:

Sign up
Here's how it works:
  1. Anybody can ask a question
  2. Anybody can answer
  3. The best answers are voted up and rise to the top

I just learned about the transfer homomorphism, and I am having trouble internalizing it. I am learning from 'A Course in the Theory of Groups', and I was hoping that perhaps someone had a more intuitive explanation of the transfer to supplement the book's construction.

share|cite|improve this question
up vote 11 down vote accepted

The transfer tries to capture conjugacy in a homomorphism. This is a little hard to do so the definition is a little complicated. I give a simpler to verify definition from Kurzweil–Stelmmacher, then the standard product-of-conjugates formula and its specialization to the center and to a Hall subgroup.

Smoother definition

Kurzweil–Stellmacher gives a slightly nicer definition of the transfer: Let $H \leq G$ be a subgroup of finite index in the group $G$. If $R,S$ are transversals of $H$ in $G$, define $$(R|S) = \prod_{(r,s) \in R \times S \atop Hr = Hs} \overline{ rs^{-1} } \in H/[H,H]$$ Then one easily verifies that $$(R|S)^{-1} = (S|R), \quad (R|S)(S|T) = (R|T), \quad (Rg|Sg) =(R|S), \quad \text{and} \quad (Rg|R) =(Sg|S)$$ (At the moment, I couldn't do the last in my head, but $(Rg|R) = (Rg|Sg) (Sg|S) (S|R) = (Sg|S)$ since the outer factors cancel by the previous properties and the fact that $(R|S)$ lives in an abelian group.)

Hence the map $tr:G \to H/[H,H]: g \mapsto (Rg|R)$ is well-defined (by the last property) and a homomorphism since $$(Rgh|R) = (Rgh|Rh)(Rh|R) =(Rg|R)(Rh|R)$$

Some important properties

For g in G, consider the orbits of $\langle g \rangle$ on the cosets $G/H$. Let them have representatives $g_i \in R$ and size $n_i$. Then $$tr(g) = \prod g_i g^{n_i} g_i^{-1}$$ is a product of conjugates of $g$. In particular, if $g \in Z(G)$, then all the conjugates are the same, so $$tr(g) = g^{[G:H]}, \quad \text{if } g \in Z(G)$$ and a similar formula holds if we consider $tr(g)$ mod the subgroup $$H^* = \langle h^{-1} h^g : h,h^g \in H, g \in G \rangle$$ since modulo this subgroup, all $G$-conjugates are the same: $$tr(g) \equiv g^{[G:H]} \mod H^*, \quad \text{if } g \in H$$

In particular, if $G$ does not sufficiently swirl $H$, that is, if $H^* < H$, and if $\gcd([H:H^*],[G:H])=1$ (for instance if $H$ is Hall or Sylow), then the transfer map restricts to an automorphism of $H/H^*$ and in particular $G$ has a proper normal subgroup $K=\ker(tr)$ such that $H \cap K = H^*$. When $H$ is a Sylow $p$-subgroup, this $H^*$ is called the focal subgroup and is equal to $H \cap [G,G]$.

Notice that $H^* \geq [H,H]$ measures how much extra conjugacy (fusion) $G$ causes in $H$. The transfer allows us to pull $H^*$ back into $G$, and so we get a normal subgroup of $G$ that somehow measures fusion in $H$. Again this is most useful and precise when $H$ is a Hall or Sylow subgroup of $G$.

share|cite|improve this answer
Other good answers would include the definition from homology of groups, and the definition from character theory. This answer is closest to the material in Robinson, if I recall correctly. – Jack Schmidt Oct 2 '12 at 21:45

There is a way (in my opinion, quite natural) which leads not only to transfer map, but also to the induced representations.

share|cite|improve this answer
Please explain your downvotes. – Alex W Jun 5 '15 at 16:48

Your Answer


By posting your answer, you agree to the privacy policy and terms of service.

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