I know that two closed fields of caracteristic $0$ and uncountable are isomorphic iff they have the same cardinality. But I don't know why $\mathbb{C}_p$ has the same cardinality as $\mathbb{C}$. Can anyone give me some reference or hint ?

  • $\begingroup$ By $C$, do you mean $\mathbb{C}$? $\endgroup$ – Thomas Andrews May 23 '11 at 7:56

Step 1: Any complete metric space without isolated points has cardinality at least $\mathfrak{c}$ (continuum cardinality). I got a nice explanation of this here. In particular $\mathbb{C}_p$ has cardinality at least $\mathfrak{c}$.

Step 2: As wikipedia knows, a topological space which is Hausdorff, first countable and separable (so in particular a separable metric space) has cardinality at most $\mathfrak{c}$: indeed, every point is the limit of a sequence from a countable set, and $\aleph_0^{\aleph_0} = \mathfrak{c}$. Now $\overline{\mathbb{Q}}$ is dense in $\overline{\mathbb{Q}_p}$ (a consequence of Krasner's Lemma: see e.g. $\S 3.5$ of these notes) and hence also in its completion $\mathbb{C}_p$. So $\mathbb{C}_p$ has cardinality at most $\mathfrak{c}$.

[Added: Here is alternate -- less elegant but more elementary -- argument for Step 2:
(i) For any infinite field $K$, the cardinality of $K$ is equal to the cardinality of its algebraic closure.
(ii) For any metric space $X$, the cardinality of the completion of $X$ is at most $\# X^{\aleph_0}$ (the number of sequences with values in $X$). By standard facts on cardinal exponentation, we have $\# X \leq \mathfrak{c} \iff \# X^{\aleph_0} \leq \mathfrak{c}$.]

Thus by the Schroder-Bernstein Theorem, $\mathbb{C}_p$ has cardinality $\mathfrak{c}$.

  • $\begingroup$ You need a separation assumption (Hausdorff, say) in step 2. Of course, we have that in this case. $\endgroup$ – Chris Eagle May 23 '11 at 8:13
  • $\begingroup$ @Chris: I sure do, thanks. $\endgroup$ – Pete L. Clark May 23 '11 at 8:15
  • 2
    $\begingroup$ Step 1 also has an elementary version : using $p$-adic Hensel expansion, we see that $\mathbb{Z}_p$ is uncountable (which is very similar to what can be done in $\mathbb{R}$). $\endgroup$ – Joel Cohen May 23 '11 at 14:59
  • $\begingroup$ @Joel: that's definitely another nice way to see Step 1, thanks. (I'm not sure whether it's more or less elementary than what I said; it depends on the person, I suppose.) $\endgroup$ – Pete L. Clark May 24 '11 at 5:14

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.