A basic form of Hilbert's irreducibility theorem can be formulated as follows:

Let $f(t,x)\in\mathbb{Q}[t,x]\setminus\mathbb{Q}[t]$ be an irreducible polynomial. There exist infinitely many linear polynomials $p(t)\in\mathbb{Q}[t]$, $p(t)=t-t_0$ such that $(p(t),f(t,x))$ is a maximal ideal of $\mathbb{Q}[t,x]$.

My question is: can we drop the "linear" assumption above and have polynomials $p(t)$ of arbitrarily high degree? More precisely, is the following true:

Let $f(t,x)\in\mathbb{Q}[t,x]\setminus\mathbb{Q}[t]$ be an irreducible polynomial. Let $n$ be a positive integer. There exists a polynomial $p(t)\in\mathbb{Q}[t]$ such that $\deg p(t)>n$ and $(p(t),f(t,x))$ is a maximal ideal of $\mathbb{Q}[t,x]$.

Of course $p(t)$ must be irreducible itself, but that's far from enough.


Let $L$ be a finite extension of ${\mathbb Q}$ over which $f(t,x)$ has absolutely irreducible factors, then for any $n$ there exists $K$ of degree $n$ over ${\mathbb Q}$ which is linearly disjoint from $L$, apply Hilbert Irreducibility over $K$ to $f(t,x)$ which is irreducible over $K$, so there exists $a \in K$ such that $K={\mathbb Q}(a)$ and $f(a,x)$ is irreducible over $K$, then $p(t)$ = minimal polynomial of $a$ over ${\mathbb Q}$ has degree $n$ and $(p(t), f(t,x))$ is a maximal ideal of ${\mathbb Q}[t,x]$.

| cite | improve this answer | |
  • $\begingroup$ The problem is that every field contains $\mathbb{Q}$ hence no $K$ is going to be linearly disjoint from $L$. How could we force the element $a$ to be outside $\mathbb{Q}$? If the degree of extension is prime, just one element would do. $\endgroup$ – Adam Przeździecki Jul 8 '18 at 5:49

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.