How can I show that for any prime $p$ and any $d\ge 2$ there exists a polynomial of degree $d$ in $\mathbb{F}_p[X]$ with no roots? ($\mathbb{F}_p$ is the finite field with $p$ elements). Thanks in advance for any idea.
-
1Take $1+\prod(x-a)$, where the product runs over all the elements of the field. https://yutsumura.com/prove-that-any-algebraic-closed-field-is-infinite/#:~:text=f(x)%3D1%2B,)%E2%88%88F%5Bx%5D.&text=Hence%20the%20finite%20field%20F,closed%20field%20must%20be%20infinite. Edit: I've just realised that the polynomial must be of a prescribed degree. Still, I think the comment provides some help. – Dog_69 Nov 26 '21 at 15:12
-
I need it for every possible degree $d\ge 2$. So for example if $p=7$ and $d=3$, how I can show it? – Logic_Problem_42 Nov 26 '21 at 15:14
-
The proof via mimicking Euclid's argument (cf. 1st comment) has been discussed here many times, e.g. here – Bill Dubuque Jul 14 '22 at 15:02
5 Answers
Even more than that, there exists an irreducible polynomial of degree $d$. The field $\mathbb{F}_{p^d}$ is a field extension of degree $d$ over $\mathbb{F_p}$. The multiplicative group a finite field is cyclic, so let $\alpha$ be a generator of $\mathbb{F}_{p^d}^{\times}$, and then we have $\mathbb{F}_{p^d}=\mathbb{F}_p(\alpha)$. Since $[\mathbb{F}_p(\alpha):\mathbb{F}_p]=d$ it follows that the minimal polynomial of $\alpha$ is an irreducible polynomial in $\mathbb{F}_p[x]$ of degree $d$.

- 39,605
-
-
2
-
@WimC The existence of $\mathbb{F}_{p^d}$ is a known theorem. Take the splitting field of the polynomial $x^{p^d}-x$ over $\mathbb{F}_p$. – Mark Nov 26 '21 at 15:33
-
1I'm sure WimC knows about the existence of all the finite fields. I interpreted the comment to ask for a proof without resorting to the existence of splitting fields etc. That can be done, but it is a bit hairy. I take the splitting field route any day of the week :-) – Jyrki Lahtonen Nov 26 '21 at 16:07
-
1@JyrkiLahtonen The question about construction of $\mathbb F_{p^n}$ was to check if the reasoning wasn’t circular (i.e. if it used an irreducible polynomial of degree $n$). I also just noticed that the question you linked to is basically about the counting arguments in my second answer. I think it adds some value, so I’ll leave it here. The counting approach is simple enough to derive from first principles. – WimC Nov 29 '21 at 19:04
Note that $f(x)=x^d-x$ has at least two roots in $\Bbb Z/p\Bbb Z$, namely $0$ and $1$. Thus, viewed as a map $f\colon\Bbb Z/p\Bbb Z\to \Bbb Z/p\Bbb Z$, it is not injective, hence also not surjective. Pick $a\in \Bbb Z/p\Bbb Z$ that is not in the image. Then $g(x):=x^d-x-a$ has no root in $\Bbb Z/p\Bbb Z$.

- 374,180
You can even find an irreducible polynomial of degree $d$.
The field $F_{p^d}$ has characteristic $p$, hence it can be viewed as an extension of $F_p$ with degree $d$.
By the primitive element theorem, there exists some $\alpha \in F_{p^d}$ such that $F_{p^d}=F_p(\alpha)$. Call $M$ the minimal polynomial of $\alpha$. Then $M$ is irreducible, the ring $F_p[\alpha]$ is isomorphic to $F_p[X]/(M)$, which is a field, hence $$F_p[X]/(M) \simeq F_p[\alpha]=F_p(\alpha)=F_{p^d}.$$ https://en.wikipedia.org/wiki/Primitive_element_theorem
As a result, the polynomial $M$ is irreducible with degree $d$.
Remarks : there exists a formula providing the number of irreducible with degree $d$ on $F_p$.

- 8,223
-
2
-
$F_{p^d}$ is the splitting field of the polynomial $X^{p^d}-X$. View https://en.wikipedia.org/wiki/Splitting_field and https://en.wikipedia.org/wiki/Finite_field – Christophe Leuridan Nov 26 '21 at 16:01
For any $\alpha \in \mathbb F_p$ there are $p^{d-1}$ monic polynomials of degree $d$ that vanish at $\alpha$. This would give a maximum of $\lvert \mathbb F_p \rvert \cdot p^{d-1} = p^d$ polynomials with a root in $\mathbb F_p$, which is too weak a bound. However, we counted polynomials with $k$ roots in $\mathbb F_p$ $k$ times and such a polynomial certainly exist for some $k \geq 2$. Therefore the number of monic polynomials with a root in $\mathbb F_p$ is strictly less than $p^d$. In other words, there must be a polynomial of degree $d$ without roots in $\mathbb F_p$.

- 32,192
- 2
- 48
- 88
This answer presents another approach to show that irreducible polynomials of any degree exist. It is purely based on counting arguments that can be derived without (and in fact predate) the theory of splitting fields.
Fix a prime $p$ and let $A_n$ for $n\geq 1$ be the number of irreducible monic polynomials of degree $n$ in $\mathbb F_p[x]$. The following expression for $A_n$ dates back to Gauss and Dedekind:
$$n A_n = \sum_{d \mid n} \operatorname{\mu}\left(\frac n d \right) p^d$$
where $\mu$ is the Möbius function. Since $\mu$ takes values in $\{-1,0,1\}$ and $\mu(1)=1$ this shows that $A_n>0$ for all $n\geq 1$. The formula above is the Möbius inversion (already known to Dedekind) of $$p^n = \sum_{d \mid n} d A_d.$$ This also suffices to show that $n A_n > 0$ since the summands for $d < n$ are too small to sum to $p^n$ (using the bound $d A_d \leq p^d$). This second identity in turn follows from counting the number of ways in which irreducible polynomials can be multiplied to give all polynomials of a given degree which leads to the “generating” identity $$\prod_{n=1}^{\infty}(1-x^n)^{A_n} \equiv 1-px \pmod {x^m}$$ for any $m \geq 1$.

- 32,192
- 2
- 48
- 88