I just proved that a commutative ring $R$ is an integral domain iff $R$ is isomorphic to a subring of a field.

My question is why can't $R$ be a field with these conditions? Aren't we satisfying all of the field's properties?


  • 3
    $\begingroup$ Being a sub-thing of a field is pretty different from being a field. The integers are subring of the reals, but can you invert 2 and stay in the integers? $\endgroup$
    – Randall
    Sep 1 '17 at 13:42
  • $\begingroup$ It is not excluded that under these conditions $R$ is a field. But also not guaranteed. It might be that elements of $R$ have no inverse. Looking at it as subring of field $F$ that means that their inverse is in $F-R$. $\endgroup$
    – drhab
    Sep 1 '17 at 13:45

For a counter-example, let's have a look at $\mathbb{Z} \subseteq \mathbb{Q}$.
Here $\mathbb{Z}$ is an integral domain which is not a field;
also you can check that $\mathbb{Z}$ is a sub-ring of the field of rational numbers $\mathbb{Q}$.

Note that $\mathbb{Z}$ satisfies all of the field's properties; except the property which concerns the
existence of multiplicative inverses for non-zero elements. For example $2^{-1} \notin \mathbb{Z}$ .

  • $\begingroup$ I overlooked that example I guess. I think it is because I kept thinking about finite integral domains for some reason :(. Don't know why I couldn't wring myself out of there... Thanks. $\endgroup$
    – Friedman
    Sep 1 '17 at 14:07
  • 5
    $\begingroup$ @Friedman By the way, in case you don't already know, a finite integral domain $\textit{is}$ always a field. $\endgroup$
    – wgrenard
    Sep 1 '17 at 14:14

No, subring of a field does not satisfy all the field's axioms. Namely, the problem is twofold: the subring doesn't have to contain $1$ and even when it does, there is trouble with inverses.

Let $1\in R\subseteq \mathbb F$, where $\mathbb F$ is a field and let $0\neq r\in R$. Sure, $r$ is invertible in $\mathbb F$, but what guarantees that $r^{-1}\in R$? Well, nothing. See the answer by Famke where they give example of $\mathbb Z\subseteq \mathbb Q$.

  • $\begingroup$ "The subring doesn't have to contain $1$" is not necessarily a problem. If you work with unitary rings such problem doesn't exist. $\endgroup$
    – Xam
    Sep 1 '17 at 20:05
  • $\begingroup$ But if you don't it does. Anyway, it's obvious that the emphasis was put on inverses, @Xam. $\endgroup$
    – Ennar
    Sep 1 '17 at 20:07

A yet another example are polynomials $R[x]$ over a field $R$. While (obviously) all other properties are fine, there are much less inverses than there are required for a field.

Now formal power series offer somewhat a remedy...


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.