3

I am doing some research of my own and I have a brief question and I don't recall studying this particular thing. Is it generally true that if a set coupled with a binary operation is closed under inverses and has an identity then $ab=cd$ implies $ba=dc$ for all $a,b,c,d$ in the set? If not, would this statement hold true for groups?

sdcvvc
  • 10,528
Jebruho
  • 1,700

4 Answers4

18

As long as your operation has an identity, the property

$$ab = cd \Rightarrow ba=dc$$

is equivalent to commutativity: if the operation is commutative, the implication is obvious, for the other direction, put $c=1$ and $d=ab$ to get $ab = ba$.

sdcvvc
  • 10,528
6

Consider $GL(2,\mathbb R)$. It is a non empty set with binary associative composition, unit and inverses.

Let $A$ be the unit matrix, $B=\begin{bmatrix}0 & 1\\1 & 0\end{bmatrix}$, $C=\ \begin{bmatrix}1 & 1\\0 & 1\end{bmatrix}$ and $D=\ \begin{bmatrix}-1 & 1\\1 & 0\end{bmatrix}$.

Then $AB=B=CD$; but $B=DC$ is not true.

Avitus
  • 14,018
5

Counterexample: Consider the group $S_3$. $$(13)(12)=(123)e$$ However: $$(12)(13)\not=e(123)$$

Amr
  • 20,030
3

Another one may be arisen in $Q_8$, the quaternion group. Just take $a=x,y=b=c,d=x^{-1}$.

Mikasa
  • 67,374