This question originates from Pinter's Abstract Algebra, Chapter 27, Exercise G1.

Let $F$ be a field, and let $c$ be transcendental over $F$. Prove {$a(c):a(x) \in F[x]$} is an integral domain isomorphic to $F[x]$.


In general, if $A$ is an integral domain, then $A[x]$ is an integral domain.

Let $S_c =$ {$a(c):a(x) \in F[x]$}. $S_c$ is an integral domain iff $S_c$ is a commutative ring with unity that has no divisor of zero.

  1. $S_c$ is obviously commutative and closed under addition and multiplication by the properties inherited from $F[x]$, which is an integral domain. Note $F[x]$ is an integral domain for $F$ is a field.
  2. $a(x)=1\in F[x]\implies a(c)=1\in S_c$.
  3. Suppose $a_1(c)a_2(c) = 0$ for nonzero polynomials $a_1(x), a_2(x)\in F[x]$. Let $b(x)=a_1(x)a_2(x)$. Then $b(c) = 0$, contradicting that $c$ is transcendental. Hence $a_1(c)a_2(c) = 0$ implies either $a_1(x)$ or $a_2(x)$ is zero; that is, $S_c$ has no divisor of zero.

This completes the proof that $S_c$ is an integral domain.

Let $\sigma_c: F[x]\rightarrow S_c$ such that $\sigma_c(a(x)) = a(c)$. Note $\sigma_c$ is a ring homomorphism from $F[x]$ onto $S_c$. Furthermore, $a(c)= b(c)\implies a(x)= b(x)$, for otherwise suppose $a(x)\ne b(x)$, then $a(x)-b(x)\ne 0\implies a(c)-b(c) \ne 0$, as $c$ is transcendental, contradicting $a(c)=b(c)$. Hence $\sigma_c$ is bijective and therefore an isomorphism; that is, $S_c\cong F[x]$.

  • 1
    $\begingroup$ $F$ might be a field, but $F[x]$ never is. $\endgroup$
    – user239203
    Apr 4 '20 at 20:32
  • 5
    $\begingroup$ Shorter: The map $F[x]\to \{\,a(x):a(x)\in F[x]\,\}$ induced by $x\mapsto c$ is obviously an epimorphism of rings. As $c$ is transcendental, its kernel is trivial,hence it is an isomorphism. $\endgroup$ Apr 4 '20 at 20:51

Your claim that that $\sigma_c$ is injective is rather unconvincing. Also, as suggested in the comments, you can skip the entire proof that $S_c$ is an integral domain; once you prove that $S_c\cong F[x]$ it follows that $S_c$ is an integral domain because $F[x]$ is. Overall your proof focuses on the wrong points; you gloss over key facts without any argument, and elaborate on completely irrelevant facts.

To prove that $S_c\cong F[x]$ indeed the most efficient approach is to note that the ring homomorphism $$\sigma_c:\ F[x]\ \longrightarrow\ S_c:\ a\ \longmapsto\ a(c),$$ is surjective by definition of $S_c$, and injective because $a(c)=0$ implies $a=0$ because $c$ is transcendental. Then $S_c$ is an integral domain because $F[x]$ is.

  • $\begingroup$ Agree. I've fixed the injectivity argument in the OP anyway. Thank you for the answer. $\endgroup$
    – hchar
    Apr 5 '20 at 18:20

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.