In a semigroup $S$ we define $a\le_\mathcal{J} b$ iff $a=xby$ for some $x,y\in S^1$. Defining $a\equiv_\mathcal{J} b$ by $a\le_\mathcal{J} b$ and $b\le_\mathcal{J} a$ gives a partial order on $S/\equiv_\mathcal{J}$. Is this order always total?
2026-03-25 20:34:27.1774470867
On
Does Green's $\mathcal{J}$-relation define a total order on the equivalence classes?
139 Views Asked by Bumbble Comm https://math.techqa.club/user/bumbble-comm/detail At
2
There are 2 best solutions below
8
On
It might be good to look at this order as the inclusion of principal two-sided ideals instead (as it's isomorphic). Perhaps the fact that you didn't look at it like this is why you didn't immediately answer the question yourself. By your nickname I assume you like rings. In rings, principal ideals in the ring-theoretic sense and principal ideals in the semigroup-theoretic sense are the exact same thing. In the commutative case, two-sided and one-sided principal ideals are the same thing too. Can you see examples of commutative rings you know in which the principal ideals don't form a chain?
No it is not always a total order.
For example, in the semigroup $S$ with elements $a$, $b$, and $c$ and multiplication in the table below:
\begin{equation*} \begin{array}{c|ccc} &a&b&c\\\hline a&a&a&a\\ b&a&b&a\\ c&a&a&c \end{array} \end{equation*}
The $\mathscr{J}$-order on $S$ has $b$ and $c$ incomparable, $a\leq_{\mathscr{J}} b$ and $a\leq_{\mathscr{J}} c$.
More generally, every partially ordered set with a minimum element is order-isomorphic to the partial order of the $\mathscr{J}$-classes of some semigroup. A reference for this is Theorem 4(i) in the paper