Show that the ideal $(XY+XZ+YZ,XYZ) = (X,Y)(Y,Z)(X,Z)$ and the irreducibility of the vanishing sets of the factors.

666 Views Asked by At

Consider $K[X,Y,Z]$ over a field $K$ of characteristic $0$. I want to show that

  1. EDIT (thanks to user26857): This equality does not hold, for example $X^2Y \notin (XY+XZ+YZ,XYZ)$. I want to show that $(XY+XZ+YZ,XYZ) = (X,Y)(Y,Z)(X,Z)$ in order to conclude that the vanishing sets of the left hand side ideal is equal to the union of the three ideals on the right hand side,
  2. the vanishing sets of $(X,Y),(Y,Z)$ and $(X,Z)$ are irreducible.

Concerning Part 1 EDIT (thanks to user26857): This equality does not hold, for example $X^2Y \notin (XY+XZ+YZ,XYZ)$. I have shown explicitely that $$V_{(XY+XZ+YZ,XYZ)}(K) = V_{(X,Z)} \cup V_{(X,Y)} \cup V_{(Y,Z)} $$ by solving the systems of equations (for example the vanishing set of $(X,Y)$ is the set of points $(0,0,z)$).

However, I wanted to try and show it using the property that the vanishing set of a product of ideals equals the union of the vanishing sets of the respective ideals.

An element in $(XY+XZ+YZ,XYZ)$ is of the form $f(XY+XZ+YZ)+gXYZ$ where $f,g$ are polynomials in $K[X,Y,Z]$. Now $gXYZ$ is in $(X,Y)(Y,Z)(X,Z)$, since, for instance, $gX\in (X,Y)$, $Y\in (Y,Z)$ and $Z \in (X,Z)$. I have problems for the term f(XY+XZ+YZ), because there are only two factors in each summand and seeing $XY$ as $XY1$ does not help because $1$ is not in any of the ideals. I also tried rewriting the expression as

$$(fY+fZ+gYZ)X + fYZ\text{ or }(fY+fZ)X + (fY+gXY)Z$$ but the main problem is that I do not see how any of the forms corresponds to an element in the product ideal $(X,Y)(Y,Z)(X,Z)$. I hope that my definition of the product of ideals $a,b,c$ is correct:

$$abc= \{\sum_{j=1}^m (\sum_{i=1}^n a_ib_i) c_j : a_i\in a, b_i\in b, c_i\in c, n,m\in \mathbb{N}\},$$ which I derived from the definition $$ab=\{\sum_{i=1}^n a_ib_i:a_i\in a, b_i\in b, n\in \mathbb{N}\}.$$

Concerning Part 2 Take for example $V_{(X,Z)} = \{(0,y,0)\}$.

  • My first idea was to use the fact that $$V_{(X,Z)}(K) \subseteq V_{XYZ}(K).$$ The polynomial $XYZ$ is irreducible, which implies that the coordinate ring of $V_{XYZ}(K)$ is an integral domain. So $V_{(X,Z)}(K)$ is a subset of an irreducible set. This does not allow me to conclude, because as far as I know one can only say that open subsets of irreducible sets are irreducible. Now $V_{(X,Z)}(K)$ is closed and I would guess that it is not open (I have no argument for this though).
  • As a second attempt, I tried to prove that the vanishing ideal of $V_{(X,Z)}(K)$ is prime, which is equivalent to the set being irreducible. First of all, what is the vanishing ideal of $V_{(X,Z)}(K)$? Is it $(X,Z)$? The inclusion $(X,Z) \subseteq I_{V_{(X,Z)}(K)}$ is obvious and for the other inclusion I argue that any polynomial in $K[X,Y,Z]$ that vanishes for any $(0,y,0)$ must have a zero constant term and in every term there must be at least an $X$ or $Z$, so that we can factor out $X$ and $Z$ to form a polynomial in $(X,Z)$. As for the primeness, I do not have a rigorous argument, however, I think that the ideal should be prime, since the product of two polynomial that are not in the ideal can never lie in the ideal, for there will be a constant term or term without an $X$ or $Y$ in it that will mess things up.