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

If $G$ is abelian, $T(G)$ is the torsion subgroup, then $G/T(G)$ is torsion free.

If $T(G) = \{1\}$, then $G$ is called a torsion-free group. Below is what I did to prove this statement.

$G/T(G) = \{T(G)g\mid g\in G\}$. If $g \in G$, then $T(G)g = T(G)$, otherwise $T(G)g = G - T(G)$. $So T(G/T(G))= \{T(G)\}$. $1$ is in $T(G)$ for sure, but it's also possible that there exists other elements in $T(G)$.

Could anybody point out where I am wrong?

share|cite|improve this question
"If $g \in G$" is probably not what you meant. – Dylan Moreland Mar 15 '12 at 3:17
up vote 2 down vote accepted

First: it is false that if $g\in G$ then $T(G)g = T(G)$, but I suspect you meant "$g\in T(G)$", which would make that statement true.

Second: it is false that if $g\notin T(G)$ then $T(G)g=G-T(G)$. It's true that the coset $T(G)g$ is disjoint from $T(G)$, and hence is contained in $G-T(G)$, but it need not be equal to $G-T(G)$. In general it won't be (it will only be the case if $T(G)$ is of index $2$).

Here's what you need to do: take an element of $G/T(G)$, which looks like $T(G)g$ with $g\in G$, and assume that it is a torsion element: that is, there exists $n\gt 0$ such that $(T(G)g)^n = T(G)$. This will occur if and only if $T(G)g^n = (T(G)g)^n = T(G)$, if and only if $g^n\in T(G)$.

Can you now prove that in fact $g\in T(G)$? If so, then $T(G)g = T(G)$, and you will have proven that:

If $T(G)g\in T(G/T(G))$, then $T(G)g = T(G)$.

which will show that $T(G/T(G)) = \{T(G)\}$ (which happens to be the identity of $G/T(G)$).

share|cite|improve this answer
Is the group of rational numbers a torsion free group? – Shannon Mar 15 '12 at 4:55
@Shannon: You tell me: the rational numbers are a group under addition, so an element $a\in\mathbb{Q}$ is torsion if and only if there exists $n\gt 0$ such that $\underbrace{a+\cdots+a}_{n\text{ summands}} = 0$. Is the only torsion element $0$, or are there other torsion elements? – Arturo Magidin Mar 15 '12 at 5:00
sorry, I mean the rational numbers under multiplication. – Shannon Mar 15 '12 at 5:04
@Shannon: The rational numbers are not a group under multiplication (zero has no inverse). For nonzero rationals, you have two possibilities: the positive rationals under multiplication, or the nonzero rationals under multiplication. Again the question is: suppose that $q$ is a nonzero [positive] rational, and $n\gt 0$ a positive integer. If $q^n=1$, does that mean that $q=1$, or not? – Arturo Magidin Mar 15 '12 at 15:27
I think the rational numbers under addition is a torsion free group. And for nonzero rational numbers under multiplication is a torsion free group too. Pls correct me if I am wrong. Thank you so much. – Shannon Mar 15 '12 at 16:54

It's not too hard to see that $G/T(G)$ is torsion free,

The idea is that we're sending all the torsion elements to $0$ by modding out by $T(G)$, so of course there are none left after we do that!

To put it a little more formally

Suppose $a\in G/T(G)$ is torsion. Then $a^{k} = 0$ for some $k$, i.e. $a^k \in T(G)$. But this means that $a^k$ is torsion in $G$; but of course this menas that $a$ itself was torsion, so $a\in T(G)$ i.e $a = 0$ in $G/T(G)$. Thus we see that $G/T(G)$ is torsion free.

As for your proof the statement "If $g\in G$ $T(G)g = T(G)$, otherwise $T(G)g = G - T(G)$" is certainly false.

share|cite|improve this answer

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.