True or False:

Let $A \supseteq R$ be an integral ring extension of $R$, with $A$ being a domain. If every non-zero prime ideal of $R$ is a maximal ideal, then every non-zero prime ideal of $A$ is also maximal.

This is a follow up to another problem that's exactly identical except that in the first problem $A$ is not necessarily a domain. I found a counterexample for that problem, using $\mathbb{Z} = R$ and $\mathbb{Z}[x]/(x^2) = A$.

I'm thinking that being a domain changes something important here, but I can't figure out what. By the Going Up/Maximality Theorems in my book, I know that every prime ideal in $A$ lying over a prime ideal in $R$ is maximal, so I really only need to show that A has no prime ideals that don't lie over a prime ideal in R or that prime ideals of A that aren't lying over a prime ideal in R are maximal, but I'm getting stuck here.

There's a hint in my book that says that if {$0$} is a prime ideal in $R$, then the only prime ideal of A lying over {$0$} is {$0$}, but I'm not sure how that's helpful since we're dealing specifically with non-zero prime ideals in the proof.

Insight is appreciated!


Any non-zero prime ideal $\mathfrak{q}\subseteq A$ lies over some non-zero prime ideal $\mathfrak{p}=\mathfrak{q}\cap R \subseteq R$. Then you are done by those going up/maximality theorems that you mention.

Added: since the answer is rather short and you also asked for insight, I will add some more. In an integral ring extension $R\subseteq A$ we have $$ \dim_{\text{Krull}}(A)=\dim_{\text{Krull}}(R) $$ The exercise you are proving is a particular case of this (for $\dim_{\text{Krull}}(R)=1$, assuming it is not a field).

You probably knew a particular case of this already: an integral extension of a field is a field, meaning that the smallest ring containing your field and some algebraic element is again a field. But if you add a transcendental element $T$ you don't get a field anymore, you get a principal ideal domain.

This has a nice geometric interpretation in terms of actual dimension of topological spaces: fields are points, principal ideal domains are curves, ... (the ring being the functions on the geometric space). This doesn't apply on arbitrary rings and is only a particular case, of course, but I think it is a nice way to think of it.

  • $\begingroup$ That makes perfect sense. I appreciate it! $\endgroup$ – Joe Apr 9 '18 at 21:31

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.