There is a famous theorem saying that

Let $\Bbb F$ be a field and $f(x)$ an irreducible polynomial in $\Bbb F[X]$. Then there exists a field extension $\Bbb L$ of $\Bbb F$ such that $f(x)$ has a root in $\Bbb L$.

I know there is a well-known argument saying that $$f(x)\in\Bbb F[x]\;\;\text{irreducible}\implies \alpha:=x+\langle f(x)\rangle\in\Bbb K:=\Bbb F[x]/\langle f(x)\rangle\;\text{is a root of}\;f(x)$$

Yes, $\Bbb K:=\Bbb F[x]/\langle f(x)\rangle$ contains a copy of $\Bbb F$. But it is painful for me to regard $$\alpha:=x+\langle f(x)\rangle \in K:=\Bbb F[x]/\langle f(x)\rangle$$ as a root of $f(x)\in\Bbb F[x]$. Can we evaluate a polynomial with anything we want, even something that has nothing to do with the coefficient ring of that polynomial?

What I really want is to construct a field $\Bbb L \supset \Bbb F$ (in the sense of set theory) and find an element $u \in \Bbb F$ which is a root of $f \in \Bbb F[x]$, can anybody show me how to construct such an $\Bbb L$ and find such a $u$ explicitly?

  • $\begingroup$ What you have is the explicit construction of $\mathbb L$ and $u$. And it's not true that $\mathbb K$ has nothing to do with $\mathbb F$, it's a field extension of $\mathbb F$. Evaluating $f$ at a point of $\mathbb K$ will yield an element of $\mathbb K$ (which may or may not lie in $\mathbb F$), just like evaluating a polynomial with real coefficients at some complex number $\alpha \in \mathbb C$ will yield a complex number (which may or may not be real). $\endgroup$ – Jim Apr 6 '15 at 4:39
  • 1
    $\begingroup$ don't you mean, find an element $u\in \mathbb{L}$? $\endgroup$ – Rolf Hoyer Apr 6 '15 at 4:54

Two comments

  • By saying that $f(x)$ has a root in $\mathbb{L}$, it is usually meant that the image of $f$ under the inclusion $\mathbb{F}[x] \hookrightarrow \mathbb{L}[x]$ has a root in $\mathbb{L}$. While you can't plug in elements of totally arbitrary sets into polynomials, the instructions comprising a polynomial make sense for any $\mathbb{F}$-algebra.

  • I think my set-theoretic aesthetics do not match yours, but you might be happier to think of $\mathbb{L}$ as a set to be a vector space over $\mathbb{F}$ with basis $\{1, \alpha, \alpha^2, \ldots, \alpha^{\deg(f) - 1}\}$, so that each element is uniquely given by a polynomial in $\alpha$ of degree less than $\deg(f)$.


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.