This is an example in Karen Smith's An Invitation to AG. I don't follow the explanation that is given and would appreciate if someone can explain the following:

$V=\{(x,y) \in \mathbb{C}^2 : |x| \le 1, |y| \le1\}$ is a closed set that is not an algebraic variety. This follows from the fact that no nontrivial algebraic variety in $\mathbb{C}^2$ can have interior points, since the zero set of one nonzero polynomial has no interior points.


Suppose that $f(x, y)$ is a polynomial. You're asking why its set of zeroes can not have any interior points.

As a preliminary observation, consider a fixed value of $x$. Substitute this value into $f$, and you'll see that all of the powers of $x$ are just coefficients for the powers of $y$, and so for a fixed value of $x$, we can view $f(x, y)$ as a polynomial in $y$. This is then either the $0$ polynomial, or it only has finitely many roots.

How often can it arise that $f(x, y)$ is the zero polynomial in $y$?

Write the polynomial $f(x, y)$ as

$$ f(x, y) = \sum_{k=0}^{d} P_k(x) y^k. $$

We see that for $f(x, y)$ to be the zero polynomial in $y$, that $x$ must be a root of each of the polynomials $P_0, P_1, \dots, P_d$. In particular, there are only finitely many possible values of $x$ for which this occurs.

We see that except possibly for a finite number of values of $x$, that for a particular value of $x$, there are only finitely many values of $y$ such that $(x, y)$ is a root of $f$.

Now suppose that the zero set of $f$ has an interior point. Then there is an open ball around this point contained in the zero set. But then there is some $x$ such that there are infinitely many values of $y$ such that $(x, y)$ is in the open ball, and hence such that $(x, y)$ is a zero of $f$. This contradicts our earlier observation. (Choose an $x$ that is not common zero of $P_0, P_1, \dots, P_d$, and look at all of the points in the open ball with that $x$ value for their $x$-coordinate.)

  • $\begingroup$ Thank you for your help. Can you please explain why "there is some $x$ such that there are infinitely many values of $y$ such that $(x,y)$ is in the open ball"? $\endgroup$ – Mark Mar 8 '17 at 11:16
  • $\begingroup$ Just imagine what an open ball looks like. For every $x$ value there are infinitely many $y$ values such that $(x, y)$ is a point in the ball. The important thing here though is because there are infinitely many different $x$-values amongst the points in the ball, we can also choose the value of $x$ so that it is not a common root of $P_0, P_1, \dots, P_d$. $\endgroup$ – Dylan Mar 8 '17 at 19:07

Suppose that $f\in\mathbb C[x,y]$ is a polynomial which vanishes on your set $V$. As $V$ contains an open set which contains $(0,0)$, you can easily check that $f$ and all its derivatives vanish at $(0,0)$, and then, since for example $f$ is an analytical function and coincides with the sum of its Taylor series at $(0,0),$ the function $f$ vanishes on all of $\mathbb C^2$.

This shows that the only polynomial in $\mathbb C[x,y]$ which vanishes on $V$ is the zero polynomial. As $V$ is not the zero set of the zero polymonial, $V$ is not an algebraic set.

  • $\begingroup$ Thank you. Please correct me if I do not understand completely: you're picking an arbitrary polynomial and then assuming $V$ is a variety containing the interior point $(0, 0)$, you arrive at the contradiction that $f$ is zero. I do not know how this rules out the possibility that there cannot be another interior point? $\endgroup$ – Mark Mar 7 '17 at 5:36
  • $\begingroup$ No, $V$ denotes the set in your question, which contains $(0,0)$ in its interior: there is no need to assume that: it simply does. Then I am showing that if a polynomial vanishes on $V$ then it vanishes on all of $\mathbb C^2$. That implies that if an algebraic set $X$ contains $V$, then $X$ is all of $C^2$. $\endgroup$ – Mariano Suárez-Álvarez Mar 7 '17 at 6:52
  • $\begingroup$ If you wouldn't mind, can you please explain also how any algebraic variety in $\mathbb{C}^2$ can't contain interior points? $\endgroup$ – Mark Mar 8 '17 at 6:07
  • $\begingroup$ The proof proves exactly this...If you have any interior point, you can just use a translation to assume the origin is an interior point...Or you can even use the same proof with any interior point since a polynomial coincides with its taylor series at every point. $\endgroup$ – MooS Mar 8 '17 at 6:13

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.