0
$\begingroup$

The original problem comes from corollary (10.10.6), chapter 10, Volumn I, EGA.

I state it in the language of modules here for convenience.

Claim. If an $R$-module $F$ is a limit of two inverse(or projective) systems $\{F_n\}_{n \geq 1}$ and $\{G_n\}_{n \geq 1}$, where the morphisms $\phi_{m,n} : F_n \longrightarrow F_m, \psi_{m,n} : G_n \longrightarrow G_m, m \leq n$ satisfy $\phi_{m,n}(F_n) = F_m$ and $\psi_{m,n}(G_n) = G_m$, then the two inverse systems are isomorphic.

I believe that this claim is wrong. Are there any examples?

Edit: EGA I corollary(1, 10.10.6) is correct due to (0,7.2.9).

$\endgroup$
4
  • 2
    $\begingroup$ Is that really exactly what EGA claims? And what have you tried so far in looking for an example? For example, have you thought about systems whose limit is the zero module? $\endgroup$ Commented Apr 27 at 10:21
  • $\begingroup$ No, the EGA statement uses the sheaf language. I have tried to prove the claim. But then I think the conditions are enough to prove it and it should be a false claim. I can not find an example to overthrow the claim. $\endgroup$
    – Functor
    Commented Apr 27 at 13:14
  • $\begingroup$ I tried searching for (10.10.6) of EGA but there is no such section. Chapter 10 only has 10.1, 10.2, and 10.3 as far as I see. $\endgroup$
    – Shrugs
    Commented Apr 27 at 19:41
  • $\begingroup$ @Shrugs. It is Volumn I of EGA. Chapter 10 is about formal schemes. $\endgroup$
    – Functor
    Commented Apr 28 at 0:33

1 Answer 1

0
$\begingroup$

The claim is false already for vector spaces.

For example, one inverse system can have increasing even dimensions and the other can have increasing odd dimensions, and the inverse limits will both have dimension $2^{\aleph_0}$.

There may still be map between inverse systems that induces an isomorphism on the inverse limit though.

EDIT:

For example, define two inverse systems by projection onto the leading factors ($m\leq n$)

$$\phi_{m,n}: k^{2n} \rightarrow k^{2m} $$

$$\psi_{m,n}: k^{2n+1} \rightarrow k^{2m+1}$$

The inverse limits are both isomorphic to $\prod_\omega k$.

$\endgroup$
2
  • $\begingroup$ Can you give more details? The morphisms $\phi_{m,n}$ should be surjective. $\endgroup$
    – Functor
    Commented Apr 28 at 0:31
  • $\begingroup$ @Functor sure I added a bit now $\endgroup$
    – Ben
    Commented Apr 28 at 7:08

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .