Let $M,N$ be finitely generated projective modules over a commutative Noetherian ring $R$ such that $M_P \cong N_P \cong R_P, \forall P \in Spec(R)$.

If $\exists n\ge 1$ such that $M \oplus R^n \cong N \oplus R^n$, then is it true that $M \cong N$ ?

  • $\begingroup$ Yes, this is true. One approach uses exterior powers: is this a tool which is accessible for you? $\endgroup$ – Alex Wertheim May 9 at 23:46
  • $\begingroup$ @AlexWertheim: Sure, you can use exterior powers, I would like to see that ... if there are other approaches, I would like to see those too ! $\endgroup$ – user102248 May 9 at 23:48

I claim that for any (finitely generated) projective $R$-module $P$ of constant rank $1$, $P \cong \Lambda^{k+1}(P \oplus R^{k})$ for any $k \in \mathbb{N}$. From this, it will follow that $$M \cong \Lambda^{n+1}(M \oplus R^{n}) \cong \Lambda^{n+1}(N \oplus R^{n}) \cong N$$ We thus focus on proving the claim. Recall that for any two $R$-modules $P, Q$ and any $n \in \mathbb{N}$, there is a natural isomorphism of $R$-modules $$\Lambda^{n}(P \oplus Q) \cong \bigoplus_{k=0}^{n} \Lambda^{k}(P) \otimes_{R} \Lambda^{n-k}(Q)$$
From this formula, we deduce $$\Lambda^{k+1}(P \oplus R^{k}) \cong \bigoplus_{i=0}^{k+1} \Lambda^{i}(P) \otimes_{R} \Lambda^{(k+1)-i}(R^{k})$$ The first summand $\Lambda^{0}(P) \otimes_{R} \Lambda^{k+1}(R^{k})$ is trivial, since $\Lambda^{k+1}(R^{k}) = 0$. The second summand is $\Lambda^{1}(P) \otimes_{R} \Lambda^{k}(R^{k}) \cong P \otimes_{R} R \cong P$. We claim that $\Lambda^{i}(P) = 0$ for all $i > 1$, so that all the higher summands vanish. Indeed, for any maximal ideal $\mathfrak{m} \subseteq R$, we have the following sequence of isomorphisms $R_{\mathfrak{m}}$-modules: $$(\Lambda^{i}(P))_{\mathfrak{m}} \cong \Lambda^{i}(P_{\mathfrak{m}}) \cong \Lambda^{i}(R_{\mathfrak{m}})$$ since $P$ has constant rank $1$. But $\Lambda^{i}(R_{\mathfrak{m}}) = 0$ for all $i > 1$, so every localization of $\Lambda^{i}(P)$ at a maximal ideal $\mathfrak{m} \subset R$ is $0$, whence $\Lambda^{i}(P) = 0$, as desired.

  • $\begingroup$ I should add that this result fits into a larger picture, namely the beginnings of algebraic K-theory. This is a beautiful story, which is wonderfully told in the ''K-book'' by Weibel. (In particular, this result appears in the first chapter, which can be found here: sites.math.rutgers.edu/~weibel/Kbook/Kbook.I.pdf) $\endgroup$ – Alex Wertheim May 10 at 0:05
  • $\begingroup$ thanks ... it is good to know that the result is in Weibel's K-theory book ... my motivation in fact was also $K$-theoretic, I was trying to prove that $Pic(R)$ embeds in the group of units of the ring $K_0(R)$ ... my question is the injectivity of the natural map $\endgroup$ – user102248 May 10 at 0:29

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.