9
$\begingroup$

I found a set of practice questions, one of which asks whether or not $f(x) \in \mathbb{Q}[x]$ irreducible implies $f(x^2)$ irreducible.

Is this true? I'm having trouble thinking of a counterexample. Is there an irreducibility criterion that we could use?

$\endgroup$
25
$\begingroup$

No, $f(x)=x$ is irreducible and $f(x^2)=x^2$ is not.

| cite | improve this answer | |
$\endgroup$
  • 1
    $\begingroup$ Sometimes the easiest examples are the most difficult to come up with :) $\endgroup$ – Wojowu Mar 15 '17 at 20:02
  • 1
    $\begingroup$ Similarly, $f(x)=x-a^2$ is irreducible and $f(x^2)=x^2-a^2=(x-a)(x+a)$ is not. $\endgroup$ – Mike F Sep 10 '17 at 21:43
15
$\begingroup$

As Tsemo already answered, this is false.

But now you might wonder if it's still true that $f(x^2)$ is irreducible if we are given any irreducible $f(x) \in \mathbb Q[x]$ with $\deg(f) > 1$. Now, if $\alpha \in \mathbb Q$ is a root of $f(x^2)$, then $\alpha^2 \in \mathbb Q$ is a root of $f(x)$ and hence $(x - \alpha^2)$ is a factor of $f(x)$. This means if $f(x)$ is not linear then the degree of the factors of $f(x^2)$ must be at least $2$. So I looked for examples with $\deg(f) = 2$ with $f(x^2)$ having two quadratic factors. Remembering a factoring trick, I found the following example.

Take $f(x) = x^2 + 4 \in \mathbb Q[x]$ which is irreducible since it has no roots in $\mathbb Q$. Then we have, $$ f(x^2) = x^4 + 4 = x^4 + 4x^2 + 4 - 4x^2 = (x^2 + 2)^2 - (2x)^2 = (x^2 - 2x + 2)(x^2 + 2x + 2) $$ and so $f(x^2)$ is not irreducible.

| cite | improve this answer | |
$\endgroup$

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.