Lang's Algebra says that if an abelian group $A$ is free and finitely generated by $(x_i), i=1,\dots, n$ , then it is isomorphic to $\mathbb{Z}x_1 \bigoplus \cdots \bigoplus \mathbb{Z}x_n$, which is isomorphic to $\mathbb{Z}\bigoplus\cdots\bigoplus\mathbb{Z}$ ($n$-fold). But if $A = A_{\text{tor}}$, i.e. a torsion group, then isn't $A \approx \mathbb{Z}/m_i\mathbb{Z}\bigoplus\cdots\bigoplus\mathbb{Z}/m_n\mathbb{Z}$, where $m_i = $ exponent of $x_i$, finite? So can we have a finite group isomorphic to an infinite one?

  • $\begingroup$ In the first sentence, $A$ is a finitely generated free abelian group. In the second sentence, $A$ is a finitely generated torsion abelian group. $\endgroup$ – Zev Chonoles Sep 28 '13 at 20:52
  • $\begingroup$ which is a free abelian group which is torsion. $\endgroup$ – StudySmarterNotHarder Sep 28 '13 at 20:53
  • $\begingroup$ The only abelian group that is both free and torsion is the trivial group. $\endgroup$ – Zev Chonoles Sep 28 '13 at 20:53
  • $\begingroup$ So nontrivial torsion groups don't have a basis? $\endgroup$ – StudySmarterNotHarder Sep 28 '13 at 20:54
  • $\begingroup$ @EnjoysMath Yes. The independence condition implies that a torsion element can't be part of a basis. $\endgroup$ – user61527 Sep 28 '13 at 20:56

No, a finite group and an infinite group can't be isomorphic. $\mathbb Z$ (or $\mathbb Z\cdot x$) is not isomorphic to $\mathbb Z/m\mathbb Z$.

If one group has an element $y$ of infinite order, and a second group has no such element, then it's easy to show that the groups are not isomorphic. (Where would the isomorphism map $y$?)


Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

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