Irreducible components of hypersurface

356 Views Asked by At

This is exercice 5.4 from Miles Reids Undergraduate Commutative Algebra.

Let $k$ be an algebraically closed field. For $f \in k[x_1,...,x_n]$, write $V(f) \subset k^n$ for the hypersurface defined by $f=0$.

Using Hilberts Nullstellensatz, one can show that:

Claim: if $f$ is irreducible and $f$ does not divide $g$, then $V(f) \not\subset V(g)$ (see this post of mine).

From this result, I want to deduce the following.

If $g= $ cst $\cdot \, \,\Pi \, \,f_i^{n_i}$, then the irreducible components of $V(g)$ are of the form $V(f_i)$.

If I understand correctly, this follows from the counterpositive statement of the claim above.

Indeed, for any $f_i$, we have that $V(f_i) \subset V(g)$ and hence $f_i$ is either irreducible or $f_i$ divides $g$. If $f_i$ is irreducible, we are done. If not, we can reiterate the process until we get irreducibility.

Does this make sense?

Thank you for all your comments and remarks.

1

There are 1 best solutions below

1
On BEST ANSWER

You can't just use the contrapositive of the quoted result, you need the if and only if form. I.e., that $f\mid g$ if and only if $V(f)\subseteq V(g)$.

Your proof is phrased rather weirdly, and I can't quite tell what you're saying. I thought that the $f_i$s were the irreducible factors of $g$, but then you conclude that $f_i$ is either irreducible or $f_i$ divides $g$, which makes no sense as a conclusion if the $f_i$s are irreducible factors, since we know both of those things about $f_i$ from the start.

The correct proof goes like this.

Let $g=c\prod_i f_i^{n_i}$ with the $f_i$s irreducible polynomials with $f_i$ not an associate of $f_j$ for $i\ne j$, and $c\in k$. Then $f_i\mid g$, so $V(f_i)\subseteq V(g)$. Moreover, if $a\in k^n$ with $g(a)=0$, then $\prod_i f_i(a)=0$, so for some $i$, $f_i(a)=0$. Thus $a\in V(f_i)$ for some $i$. Thus $V(g)\subseteq \bigcup_i V(f_i)$. Thus we have that $$ V(g) = \bigcup_i V(f_i).$$

It just remains to check that $V(f_i)$ is irreducible. Suppose not, then there are $g$ and $h$ such that $V(f)\subseteq V(gh)$, but $V(f)\not\subseteq V(g)$ and $V(f)\not\subseteq V(h)$. Using that $f\mid g$ if and only if $V(f)\subseteq V(g)$, this translates to the existence of $g$ and $h$ such that $f_i\mid gh$, but $f_i\nmid g$ and $f_i\nmid h$, which contradicts the irreducibility of $f_i$, since irreducibles are prime in $k[x_1,\ldots,x_n]$.