MathOverflow is a question and answer site for professional mathematicians. 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

My previous question about the theorem (apparently due to Dedekind -- thanks, Arturo Magidin!) that the ring $\overline{\mathbb{Z}}$ of all algebraic integers is a Bezout domain got me thinking about rings of integers in infinite algebraic extensions of $\mathbb{Q}$.

As I mentioned here, I was thinking about Kaplansky's formulation/generalization of Dedekind's theorem: if $R$ is a Dedekind domain with fraction field $K$ such that for all finite degree field extensions $L/K$ the ideal class group of the integral closure $R_L$ of $R$ in $L$ is a torsion abelian group, then the integral closure $S$ of $R$ in $\overline{K}$ is a Bezout domain.

Since we know (by nefarious arithmetic arguments...) that the Picard group of the ring of integers of a number field is finite, Kaplansky's theorem applies with $R = \mathbb{Z}$ and gives Dedekind's theorem.

Suppose we now look at Kaplansky's proof and see if it gives more: do we need to go all the way up to $\overline{\mathbb{Q}}$ in order to show that the ring of integers is a Bezout domain?

Indeed not. The key property of the ring $S$ used in the proof is that it is root closed, i.e., closed under extraction of $n$th roots, for all $n \in \mathbb{Z}^+$. There are lots of smaller -- but still infinite -- algebraic extensions of $\mathbb{Q}$ which have this property: the smallest is $\mathbb{Q}^{\operatorname{solv}}$, the maximal solvable extension of $\mathbb{Q}$.

But what about the ring of integers of $\mathbb{Q}^{\operatorname{ab}}$, the maximal abelian extension of $\mathbb{Q}$? Is this a Bezout domain? If not, what is its Picard group?

In general, if $R$ is a Dedekind domain with fraction field $K$, $L/K$ is any algebraic field extension and $S$ is the integral closure of $R$ in $L$, then if $[L:K]$ is infinite $S$ need not be a Dedekind domain but is always a Prufer domain: every finitely generated ideal is invertible. In particular $S$ is a Bezout domain iff $\operatorname{Pic}(S) = 0$. So one may ask:

What is known about the Picard groups of rings of integers of infinite algebraic extensions of $\mathbb{Q}$? What, if anything, is the connection to more classical algebraic number theory?

share|cite|improve this question
The Picard group of the ring of integers of ${\mathbb Q}^{ab}$ is a a countable direct sum of copies of ${\mathbb Q}/{\mathbb Z}$. See – dke Dec 28 '10 at 19:18
Nice question Pete. – Adam Hughes Dec 28 '10 at 19:22
@dke: Thanks a lot. I think this is worth leaving as an answer. – Pete L. Clark Dec 28 '10 at 19:43
@Pete, I've sent you the paper. Another relevant result is by Kurihara: if you take the maximal abelian subfield of ${\mathbb Q}^{ab}$, the Picard group is trivial. Unfortunately I don't have that paper at hand. – dke Dec 28 '10 at 19:53
Oops, 'maximal abelian' should have been 'maximal real'. Damn uneditable comments... – dke Dec 28 '10 at 19:59

Let $K$ be a cyclotomic extension, and $a, b \in O_K$. If the ideal $A = (a,b)$ becomes principal in some cyclotomic extension $L/K$, then clearly you can solve the Bezout problem, and the converse holds, too: if the Bezout problem is solvable in some cyclotomic extension, then the ideal class of $A$ must capitulate there. Thus the question you are asking is whether the class group of a cyclotomic field capitulates in a cyclotomic extension. For certain pieces of the class group this is believed to be true (Greenberg), but in general I don't know the answer. Georges Gras and Bosca & Jaulent have studied the phenomenon of "abelian capitulation". From a first glance at their article it seems that the answer to your question is yes for the maximal real cyclotomic subfield of all cyclotomic extensions.

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.