I was wondering whether a maximal ideals of $K[x_1,\cdots,x_n]$ such that the quotient field equals to $K$ must be the form of $(x_1-a_1,\cdots,x_n-a_n)$? Here $K$ is not necessary to be algebraically closed.

I tried to consider the Zariski's lemma, if $\mathfrak m$ is a maximal ideal of finitely generated $K$-algebra $A$, then $A/\mathfrak m$ is a finite extension of $K$. But I don't know the degree $[A/\mathfrak m:K]=1$ means what?


It depends on what you mean by "equal". A field can be isomorphic to a non-trivial finite extension of itself, e.g. $\mathbb{C}(Y^2) \cong \mathbb{C}(Y)$, so the residue field being isomorphic to $K$ is not a strong enough condition. Concretely, a counterexample is given by $(X^2 - Y^2) \subset \mathbb{C}(Y^2)[X]$, a maximal ideal with quotient field $\mathbb{C}(Y)$.

However, if the quotient $K[X_1, \ldots , X_n] / \mathfrak{m}$ is actually isomorphic to $K$ as a $K$-algebra, then, if $a_i$ denotes the image of $X_i$ in $K$, it is easy to see that $X_i - a_i$ lies in $\mathfrak{m}$ for each $i$, and therefore $\mathfrak{m} = (X_1-a_1, \ldots , X_n - a_n)$.


I can give a positive answer when $k$ is algebraically closed.

From the Hilbert Nulstellensatz, we know that

$$k^n \to \{\mathrm{maximal \ ideals \ in \ k[X_1, \dots, X_n]}\}: (a_1, \dots, a_n) \mapsto I(\{a_1, \dots, a_n\}) = (X_1-a_1, \dots, X_n-a_n)$$ is a bijection and the result readily follows.

See also: basic question on maximal ideals in $K[X_1,...,X_n]$


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.