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

Consider the group $G=AB$ with $A$ and $B$ abelian. Is it true that $[a,b] \in [A,B]$ commutes with $[x,y]$ (with $x\in A$, $y \in B$) implies $[a,b]$ commutes with $[x^{-1},y^{-1}]$?

share|cite|improve this question
By a result of Ito, a group that is a product of abelian groups is metabelian, so all commutators commute with each other. – Derek Holt Mar 15 '12 at 13:39
Thanks, but I need that result in order to prove Ito's theorem (yes, it's at the beginning of the proof...). I believe the solution could be achieved substituting words like $a^{-1}ba$ with elements like $a^\prime b^\ast$ (or $b^\prime a^\ast$, by the fact that $G=AB=BA$), but I was not able to obtain the exact result. – Oo3 Mar 15 '12 at 16:43
I have looked at the proof of Ito's theorem, which is in German but very short, and I don't understand why you think you need to prove the result you are asking about. Which step exactly in the proof of Ito's Theorem do you not understand? – Derek Holt Mar 15 '12 at 23:05
What I do not understand is that we prove $[a,b]^{xy}$ and $[a,b]^{yx}$ are equal, so it follows that $[a,b]^{xyx^{-1}y^{-1}} = [a,b]$. But this means that $[a,b]$ commutes with $[x^{-1},y^{-1}]$, I wish it was $[x,y]$ instead. – Oo3 Mar 16 '12 at 17:58
But $x$ and $y$ are general elements of $A$ and $B$ here, so you can just replace $x,y$ by $x^{-1}$ and $y^{-1}$. – Derek Holt Mar 16 '12 at 18:31
up vote 3 down vote accepted

If it helps, you do need to assume A and B permute.

For example, take $a=x$ and $b=y$. Then obviously $\{a,x\}$ comes from an abelian group, and $\{b,y\}$ comes form an abelian group, and $[a,b]=[x,y]$ commutes with itself. However, taking: $$a=\begin{bmatrix}1&1\\0&1\end{bmatrix}, \quad b=\begin{bmatrix}0&-1\\1&0\end{bmatrix}, \quad [a,b]=\begin{bmatrix}2&-1\\-1&1\end{bmatrix}, \quad [a^{-1},b^{-1}]=\begin{bmatrix}2&1 \\ 1&1\end{bmatrix}$$ and the last two do not commute.

Of course $ba$ cannot be written in the form $a^i b^j$ (look at the zero pattern) and so the subgroups A and B do not permute (AB is not a group).

To prove Ito's theorems along your lines, it seems easiest just to show (as Derek Holt indicated) that all commutators commute. I believe this is done by by writing $b^x = a_1 b_1$ and $a^y=a_2b_2$ and then computing $[a,b]^{xy} = [a,b]^{yx}$ (simply by expanding) so that $[a,b]^{[x^{-1},y^{-1}]} = [a,b]$ and the two commutators commute. Since $x,y$ are arbitrary, it does not matter if they have an inverse or not. Similarly, since $a,b$ are arbitrary, $xy$ and $yx$ act as the same operator, so the commutator $[x,y]$ acts trivially. Ito's (1) and (2) help to show $[AB,AB]=[A,B]$ so that one gets an action on a nice object.

Ito's original proof is extremely similar (equations (3) and (4) on the first page):

  • Itô, Noboru. "Über das Produkt von zwei abelschen Gruppen." Math. Z. 62 (1955), 400–401. MR71426 DOI:10.1007/BF01180647
share|cite|improve this answer
I thought that assuming $AB$ as a group $G$ would made $A$ and $B$ automatically permutable. – Oo3 Mar 16 '12 at 18:01
@Oo3: yes, and that assumption is necessary. – Jack Schmidt Mar 16 '12 at 18:32

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.