Mathematics Stack Exchange is a question and answer site for people studying math at any level and professionals in related fields. 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

Let $\mathbb{A}$ be a polynomial ring in $n$ variables over an algebraically closed field $\mathbb F$. Given a maximal ideal $\mathfrak{m}$ of $\mathbb A$, consider the quotient $\mathbb{A}/\mathfrak{m}$ as a left $\mathbb{A}$-module ($\mathbb{A}/\mathfrak{m}$ is one-dimensional over $\mathbb F$).

Suppose $M$ is a right finitely generated $\mathbb{A}$-module which is infinite dimensional (as $\mathbb F$-vector space) and such that $\dim_{\mathbb F}(M\otimes_{\mathbb{A}} \mathbb{A}/\mathfrak{m})<\infty$ for any maximal ideal $\mathfrak m$.

Question: Can we say that $\dim_{\mathbb F}(M\otimes_{\mathbb{A}} \mathbb{A}/\mathfrak{m})$ is independent of $\mathfrak m$?

The motivation of the question is to show that $M$ is projective (therefore free, by Quillen's Theorem).

share|cite|improve this question
up vote 5 down vote accepted

No. If $M$ is f.g. over $A$ then $M\otimes_A (A/\mathfrak m)$ will be finite dimensional over $F$ for every maximal ideal $\mathfrak m$ of $M$ (since $A/\mathfrak m = F$, and change of scalars preserves the property of being finite dimensional).

So your question amounts to asking whether every f.g. module over $A$ that is infinite dim'l over $F$ is free, and the answer is no.

E.g. Suppose that $n = 1$, and take $M = A \oplus F,$ with $x$ and $y$ acting via $0$ on the second summand.

E.g. Suppose that $n = 2$, so $A = F[x,y].$ Let $M = F[x]$, regarded as an $A$-module by having $y$ act via $0$.

E.g. Suppose that $ n =2$, and let $M = (x,y) \subset F[x,y]$.

share|cite|improve this answer
Thank you for the examples. My next question is very general and unprecise (sorry about that), but it is: What is the slightest hypothesis to add in $M$ in a way that $M$ becomes a projective module? – Matt Elly Aug 14 '12 at 13:55
@MattElly: Dear Matt, If $n = 1$ then (since $A$ is a PID in that case) you have that f.g. and torsion-free implies projective. Otherwise, the only generally applicable hypothesis that I know is the one that you mention in your question, namely that the dimensions $M\otimes_A (A/\mathfrak m)$ should be constant. (There are other, more homological criteria, which can be applicable in certain particular circumstances: e.g. if M can be written as an $n$th syzygy, i.e. if there is an exact sequence $0 \to M \to P_{n-1} \to \cdots \to P_1 \to P_0,$ with all the $P_i$ projective, then $M$ is ... – Matt E Aug 14 '12 at 14:37
... itself projective. This generalizes the torsion freeness criterion in the $n = 1$ case, but you usually have to be in very particular situations for this to be helpful.) Regards, – Matt E Aug 14 '12 at 14:39
Thank you again! Do you have any reference for this last part between parentheses? – Matt Elly Aug 15 '12 at 3:45
@MattElly: Dear Matt, Any text on homological algebra should discuss it; the key point is that $F[x_1,\ldots,x_n]$ has projective dimension $n$. I would try Weibel's book, or even just wikipedia. Sorry not to be more specific; I just don't have a good knowledge of accessible homological algebra texts/sources. Also, if you're interested in this criterion, you could always ask a question about it which includes a reference request! Regards, – Matt E Aug 15 '12 at 4:29

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.