Finite fields F_q and the Frobenius automorphism

Intuition Beginner

A finite field is a number system with only finitely many elements, yet you can still add, subtract, multiply, and divide by anything nonzero. The clock arithmetic on the hours $0,1,\dots ,p-1$ for a prime $p$ is the smallest example; mathematicians write it as ${\mathbb{F}}_p$.

The surprise is that there are larger finite fields too. For every prime $p$ and every whole number $n\ge 1$, there is exactly one field with $p^n$ elements, written ${\mathbb{F}}_{p^n}$ (or ${\mathbb{F}}_q$ with $q=p^n$). These, and only these, are all the finite fields that exist.

The deepest feature is a built-in symmetry. In a field of characteristic $p$, the map that sends every element $x$ to $x^p$ reshuffles the elements while preserving all the arithmetic. This reshuffle is the Frobenius, and doing it $n$ times returns every element of ${\mathbb{F}}_{p^n}$ to itself. The Frobenius is the heart of this unit: it is both a symmetry of the field and the generator of its Galois group.

Visual Beginner

The field ${\mathbb{F}}_4$ with four elements, built by adjoining a symbol $\alpha$ with ${\alpha }^2=\alpha +1$. The Frobenius map $x\mapsto x^2$ acts on the four elements as follows.

element $x$	$0$	$1$	$\alpha$	$\alpha +1$
Frobenius $x^2$	$0$	$1$	$\alpha +1$	$\alpha$

The Frobenius swaps $\alpha$ and $\alpha +1$ while fixing $0$ and $1$. Applying it twice returns every element to itself, so its order is $2$, the degree of ${\mathbb{F}}_4$ over ${\mathbb{F}}_2$.

Worked example Beginner

Take the prime $p=7$. The nonzero hours $1,2,3,4,5,6$ form the multiplicative system ${\mathbb{F}}_7^{\times }$. List the powers of $3$ modulo $7$:

$3^1=3$, $3^2=9=2$, $3^3=6$, $3^4=18=4$, $3^5=12=5$, $3^6=15=1$.

Every nonzero value appears exactly once, so $3$ generates the whole cycle. The nonzero part ${\mathbb{F}}_7^{\times }$ is a single cycle of length $6=7-1$. A generator like $3$ is called a primitive root.

Now build a field with four elements from scratch. Start from $\{0,1\}$ and add a new symbol $\alpha$ with the rule ${\alpha }^2=\alpha +1$. The four elements are $0,1,\alpha ,\alpha +1$. Multiplying $\alpha$ by itself gives $\alpha +1$, and multiplying once more gives $\alpha \cdot (\alpha +1)={\alpha }^2+\alpha =1$. So $\alpha$ cycles through all three nonzero elements, exactly like the primitive root $3$ above.

Check your understanding Beginner

Formal definition Intermediate+

Fix a prime $p$. The prime field of characteristic $p$ is ${\mathbb{F}}_p=\mathbb{Z}/p\mathbb{Z}$. A finite field is a field with finitely many elements; any such field has characteristic $p$ for a unique prime $p$ and is a finite-dimensional vector space over ${\mathbb{F}}_p$, hence has $q=p^n$ elements for a unique $n\ge 1$. The notation ${\mathbb{F}}_q$ denotes the field with $q=p^n$ elements ^{[Lang §VIII]}.

The Frobenius endomorphism of a field $K$ of characteristic $p$ is the map

$${\mathrm{Frob}}_p:K\to K,x\mapsto x^p.$$

It is a ring homomorphism because $(xy{)}^p=x^p{y}^p$ and the freshman's dream identity $(x+y{)}^p=x^p+y^p$ holds: every middle binomial coefficient $\left(\genfrac{}{}{0ex}{}{p}{k}\right)$ for $1\le k\le p-1$ is divisible by $p$, so it vanishes in characteristic $p$. On a finite field the Frobenius is injective, hence bijective, so it is a field automorphism ^{[Dummit-Foote §14]}.

For a finite extension ${\mathbb{F}}_{p^n}/{\mathbb{F}}_p$ the norm and trace are

$$N_{{\mathbb{F}}_{p^n}/{\mathbb{F}}_p}(x)=\prod _{i=0}^{n-1}{x}^{p^i}=x^{(p^n-1)/(p-1)},{\mathrm{Tr}}_{{\mathbb{F}}_{p^n}/{\mathbb{F}}_p}(x)=\sum _{i=0}^{n-1}{x}^{p^i},$$

the product and the sum of the Galois conjugates $x,x^p,x^{p^2},\dots ,x^{p^{n-1}}$ of $x$.

Constructions

There are two standard ways to build ${\mathbb{F}}_{p^n}$, and they agree by uniqueness. The splitting-field construction takes the splitting field of $x^{p^n}-x$ over ${\mathbb{F}}_p$; its $p^n$ distinct roots form the field. This is the construction that makes uniqueness immediate.

The quotient construction picks an irreducible polynomial $f\in {\mathbb{F}}_p[x]$ of degree $n$ and forms the quotient ring ${\mathbb{F}}_p[x]/(f)$. Because $f$ is irreducible and ${\mathbb{F}}_p[x]$ is a principal ideal domain, the ideal $(f)$ is maximal, so the quotient is a field; it has $p^n$ elements, with basis $1,\bar{x},{\bar{x}}^2,\dots ,{\bar{x}}^{n-1}$ over ${\mathbb{F}}_p$ where $\bar{x}$ is the image of $x$. This is the polynomial analogue of building ${\mathbb{F}}_p$ as $\mathbb{Z}/p\mathbb{Z}$, and it is the construction the number-theory units invoke when they write ${\mathbb{F}}_q\cong {\mathbb{F}}_p[x]/(f)$ for $f$ irreducible of degree $n$.

The two constructions give isomorphic fields, but the isomorphism is not canonical: it depends on the choice of $f$ and on which root of $f$ the quotient element $\bar{x}$ is matched to. Different irreducible polynomials of the same degree $n$ yield different-looking models of the same field ${\mathbb{F}}_{p^n}$, related by Frobenius-twisted change-of-basis maps.

Counterexamples to common slips

• Not every finite ring is a field. $\mathbb{Z}/n\mathbb{Z}$ for composite $n$ has zero divisors, so it is not a field; finite fields exist only for prime-power orders.
• Not every prime power gives several fields. For each $q=p^n$ there is exactly one field of order $q$ up to isomorphism, and exactly one copy of it inside a fixed algebraic closure $\overline{{\mathbb{F}}_p}$.
• The Frobenius is not the identity in general. ${\mathrm{Frob}}_p$ fixes exactly ${\mathbb{F}}_p$ pointwise; on ${\mathbb{F}}_{p^n}$ it has order $n$, not $1$.

Key theorem with proof Intermediate+

Theorem (Existence, uniqueness, and cyclicity). For every prime $p$ and every $n\ge 1$ there exists a field ${\mathbb{F}}_{p^n}$ with $p^n$ elements. It is unique up to isomorphism (and unique inside a fixed algebraic closure $\overline{{\mathbb{F}}_p}$). Its multiplicative group ${\mathbb{F}}_{p^n}^{\times }$ is cyclic of order $p^n-1$.

Proof. Existence. Let $f(x)=x^{p^n}-x\in {\mathbb{F}}_p[x]$, and let $K$ be its splitting field over ${\mathbb{F}}_p$. The formal derivative is $f^{\prime }(x)=p^n{x}^{p^n-1}-1=-1\ne 0$, so $f$ is separable and has $p^n$ distinct roots in $K$. Let $R\subseteq K$ be the set of roots. The freshman's dream, iterated $n$ times, gives $(a+b{)}^{p^n}=a^{p^n}+b^{p^n}$; thus $a,b\in R$ implies $a+b\in R$. Likewise $(ab{)}^{p^n}=a^{p^n}{b}^{p^n}$, so $ab\in R$, and $(-a{)}^{p^n}=-a^{p^n}$ gives additive inverses, while $a\ne 0$ with $a^{p^n}=a$ gives $a^{p^n-1}=1$, a multiplicative inverse. So $R$ is a subfield of $K$ with $|R|=p^n$, proving existence.

Uniqueness. Let $L$ be any field with $p^n$ elements. Its unit group $L^{\times }$ has order $p^n-1$, so every nonzero $a\in L$ satisfies $a^{p^n-1}=1$, hence $a^{p^n}=a$; the element $0$ satisfies this too. Therefore every element of $L$ is a root of $x^{p^n}-x$, so $L$ is contained in the splitting field $K$ and in fact equals the root set $R$. Any two fields of order $p^n$ are thus splitting fields of the same polynomial over ${\mathbb{F}}_p$, hence isomorphic, and there is a unique one inside $\overline{{\mathbb{F}}_p}$.

Cyclicity of ${\mathbb{F}}_q^{\times }$. Write $G={\mathbb{F}}_q^{\times }$, a finite abelian group of order $q-1$. By the structure theorem for finite abelian groups, $G\cong \mathbb{Z}/d_1\times \cdots \times \mathbb{Z}/d_r$ with $d_1\mid d_2\mid \cdots \mid d_r$. Every element of $G$ then satisfies $x^{d_r}=1$, since $d_r$ is a multiple of the order of every element; $d_r$ is the exponent of $G$. So all $q-1$ elements of $G$ are roots of the polynomial $x^{d_r}-1\in {\mathbb{F}}_q[x]$. A nonzero polynomial of degree $d_r$ over a field has at most $d_r$ roots, giving $q-1\le d_r$. But the exponent $d_r$ divides $|G|=q-1$, so $d_r\le q-1$. Therefore $d_r=q-1$, which forces $r=1$, and $G\cong \mathbb{Z}/(q-1)$ is cyclic. $\square$

Bridge. The cyclicity of ${\mathbb{F}}_q^{\times }$ builds toward 21.01.05 primitive roots, where the prime case ${\mathbb{F}}_p^{\times }\cong \mathbb{Z}/(p-1)$ supplies a single generator (a primitive root) for the unit group modulo $p$, and appears again in 21.01.04 Fermat's little theorem, where $a^p\equiv a(\mathrm{mod}p)$ is precisely the statement that the Frobenius fixes ${\mathbb{F}}_p$ pointwise. The foundational reason the unit group is cyclic is that a polynomial of degree $d$ over a field has at most $d$ roots, which forces the exponent of ${\mathbb{F}}_q^{\times }$ up to $q-1$; this is exactly the bridge from the field property to the cyclic-multiplicative-structure fact that underwrites the discrete logarithm. The pattern generalises from ${\mathbb{F}}_p$ to every ${\mathbb{F}}_q$ without the moduli restriction that limits $(\mathbb{Z}/n{)}^{\times }$, and putting these together, finite fields are the universal arena where multiplicative cyclic structure and Galois symmetry coincide.

Exercises Intermediate+

Lean formalization Intermediate+

lean_status: none is recorded because the project still needs a local bridge from Mathlib's FiniteField, frobenius, and IsCyclic (Units K) infrastructure to the downstream curriculum narrative: the primitive-element story, the discrete logarithm, the norm and trace as AlgHom and LinearMap satisfying the transitivity identities, and the subfield lattice stated in the language used here. Mathlib itself has substantial finite-field machinery; the gap is the project-level glue to the number-theory consumers in 21.01.04, 21.01.05, and 04.04.03.

Advanced results Master

The classification is complete and striking: up to isomorphism, the finite fields are exactly the fields ${\mathbb{F}}_{p^n}$, one for each prime $p$ and integer $n\ge 1$, and every finite field is perfect because the Frobenius is an automorphism (injective on a finite set). The algebraic closure $\overline{{\mathbb{F}}_p}$ is the union ${\bigcup }_{n\ge 1}{\mathbb{F}}_{p^n}$, and its subfield lattice is the divisor lattice: the subfields of ${\mathbb{F}}_{p^n}$ are exactly the ${\mathbb{F}}_{p^d}$ with $d\mid n$ ^{[Dummit-Foote §14]}.

The Galois group is always cyclic and canonically generated. Since every $x\in {\mathbb{F}}_{p^n}$ satisfies $x^{p^n}=x$, the Frobenius satisfies ${\mathrm{Frob}}_p^n=\mathrm{id}$ on ${\mathbb{F}}_{p^n}$; its fixed field is ${\mathbb{F}}_p$, so its order is exactly $n$. Thus $\mathrm{Gal}({\mathbb{F}}_{p^n}/{\mathbb{F}}_p)=⟨{\mathrm{Frob}}_p⟩\cong \mathbb{Z}/n$. More generally, the fixed field of ${\mathrm{Frob}}_p^k$ is ${\mathbb{F}}_{p^d}$ with $d=\gcd (k,n)$, so the Galois correspondence reduces exactly to the divisor lattice of $n$ ^{[Lang §VIII]}.

The norm and trace package the conjugates into arithmetic. The trace $\mathrm{Tr}:{\mathbb{F}}_{p^n}\to {\mathbb{F}}_p$ is an ${\mathbb{F}}_p$-linear, surjective map (its image is a nonzero ${\mathbb{F}}_p$-subspace, hence all of ${\mathbb{F}}_p$), and the nondegenerate bilinear form $(x,y)\mapsto \mathrm{Tr}(xy)$ identifies ${\mathbb{F}}_{p^n}$ with its dual as an ${\mathbb{F}}_p$-vector space. The norm $N:{\mathbb{F}}_{p^n}^{\times }\to {\mathbb{F}}_p^{\times }$ is a surjective group homomorphism (Exercise 7). Both are transitive in towers: ${\mathrm{Tr}}_{K/{\mathbb{F}}_p}={\mathrm{Tr}}_{L/{\mathbb{F}}_p}\circ {\mathrm{Tr}}_{K/L}$ and similarly for the norm.

Because the Frobenius is an automorphism, every finite field is perfect: every irreducible polynomial over ${\mathbb{F}}_p$ is separable, so finite-field extensions are always Galois and the pathology of inseparability never arises. The algebraic closure $\overline{{\mathbb{F}}_p}={\bigcup }_{n\ge 1}{\mathbb{F}}_{p^n}$ inherits an automorphism ${\mathrm{Frob}}_p$ of infinite order, and the absolute Galois group $\mathrm{Gal}(\overline{{\mathbb{F}}_p}/{\mathbb{F}}_p)$ is the procyclic completion $\hat{\mathbb{Z}}={\underleftarrow{\lim }}_n\mathbb{Z}/n$, topologically generated by the Frobenius. The finite Galois groups $\mathrm{Gal}({\mathbb{F}}_{p^n}/{\mathbb{F}}_p)\cong \mathbb{Z}/n$ are its continuous quotients. This is the cleanest example of an infinite Galois group equipped with the Krull topology ^{[Lang §VIII]}.

The irreducible polynomials over ${\mathbb{F}}_p$ are counted by factoring $x^{p^n}-x$. Since its roots are precisely the elements of ${\mathbb{F}}_{p^n}$, and a monic irreducible of degree $d$ divides it iff $d\mid n$, degree-counting gives ${\sum }_{d\mid n}d{I}_p(d)=p^n$, where $I_p(d)$ is the number of monic irreducibles of degree $d$. Möbius inversion yields $I_p(n)=\frac{1}{n}{\sum }_{d\mid n}\mu (d)p^{n/d}$, which is positive for every $n\ge 1$, supplying the existence of irreducible polynomials of every degree that the quotient construction requires ^{[Lidl-Niederreiter 1997]}.

For computations one wants good bases. The normal basis theorem states that ${\mathbb{F}}_{p^n}$ admits a basis over ${\mathbb{F}}_p$ consisting of a single Frobenius orbit $\{\alpha ,{\alpha }^p,{\alpha }^{p^2},\dots ,{\alpha }^{p^{n-1}}\}$; with such a basis the Frobenius acts by a cyclic permutation of coordinates, so raising to the $p$-th power becomes a coordinate shift. This underlies efficient hardware implementations of finite-field arithmetic and the optimal normal-basis standards used in elliptic-curve cryptography.

These structures drive the applications the number-theory units depend on. The cyclicity of ${\mathbb{F}}_p^{\times }$ is the existence of primitive roots 21.01.05; the identity $a^p=a$ is Fermat's little theorem, the statement that the Frobenius fixes ${\mathbb{F}}_p$ 21.01.04; and the discrete logarithm in ${\mathbb{F}}_q^{\times }$ underwrites Diffie-Hellman key exchange and the Digital Signature Algorithm. Going further, the zeta function of ${\mathbb{F}}_q$ and the Weil conjectures count the ${\mathbb{F}}_{q^r}$-points of a variety by reading off the eigenvalues of the Frobenius acting on étale cohomology, with the elliptic-curve case treated in 04.04.03 ^{[Lidl-Niederreiter 1997]}.

Synthesis. Finite fields are the meeting point of three structures that the rest of the curriculum reads off separately: the multiplicative group ${\mathbb{F}}_q^{\times }$ is cyclic (the foundational reason discrete logarithms and primitive roots exist, building toward 21.01.05), the Galois group $\mathrm{Gal}({\mathbb{F}}_{p^n}/{\mathbb{F}}_p)\cong \mathbb{Z}/n$ is generated by a single automorphism (this is exactly the Frobenius, the finite-field shadow of complex conjugation), and the subfield lattice ${\mathbb{F}}_{p^d}\subseteq {\mathbb{F}}_{p^n}\iff d\mid n$ mirrors the divisor lattice of $n$. The pattern generalises to the Weil conjectures and the zeta function ${\zeta }_{{\mathbb{F}}_q}(t)$, where the Frobenius acts on cohomology and the point count is read off from its characteristic polynomial; the central insight is that the Frobenius is simultaneously a Galois element, a linear operator, and a cohomological eigenvalue. Putting these together with the norm and trace, finite fields become the universal testbed for arithmetic geometry: the bridge is that every counting problem over a finite field reduces to the fixed-point structure of a power of the Frobenius, appearing again in 04.04.03 elliptic curves, 21.05.01 Galois representations, and the Langlands program.

Full proof set Master

Proposition (Fixed fields of the Frobenius). Let $K={\mathbb{F}}_{p^n}$ inside $\overline{{\mathbb{F}}_p}$, and let $\sigma ={\mathrm{Frob}}_p$. For any $k\ge 1$, the fixed field $K^{⟨{\sigma }^k⟩}=\{x\in K:x^{p^k}=x\}$ equals ${\mathbb{F}}_{p^d}$ where $d=\gcd (k,n)$. In particular $\sigma$ has order $n$ and $\mathrm{Gal}(K/{\mathbb{F}}_p)=⟨\sigma ⟩\cong \mathbb{Z}/n$.

Proof. An element $x\in K$ is fixed by ${\sigma }^k$ precisely when $x^{p^k}=x$, i.e. when $x$ is a root of $x^{p^k}-x$. Inside $\overline{{\mathbb{F}}_p}$ the roots of $x^{p^k}-x$ are exactly the elements of ${\mathbb{F}}_{p^k}$ (the unique subfield of that order). So $K^{⟨{\sigma }^k⟩}=K\cap {\mathbb{F}}_{p^k}={\mathbb{F}}_{p^n}\cap {\mathbb{F}}_{p^k}$. By Exercise 6, the intersection of ${\mathbb{F}}_{p^n}$ and ${\mathbb{F}}_{p^k}$ is the largest subfield contained in both, namely ${\mathbb{F}}_{p^d}$ with $d=\gcd (k,n)$ (since ${\mathbb{F}}_{p^a}\cap {\mathbb{F}}_{p^b}={\mathbb{F}}_{p^{\gcd (a,b)}}$).

Setting $k=n$: ${\sigma }^n=\mathrm{id}$ on $K$ because every $x\in {\mathbb{F}}_{p^n}$ satisfies $x^{p^n}=x$, so the order of $\sigma$ divides $n$. Setting $k$ with $\gcd (k,n)=1$ gives a fixed field of ${\mathbb{F}}_p$, so no smaller power of $\sigma$ fixes all of $K$; combined with the divisor-lattice description, the order of $\sigma$ is exactly $n$. The extension ${\mathbb{F}}_{p^n}/{\mathbb{F}}_p$ is Galois (it is the splitting field of the separable polynomial $x^{p^n}-x$), so $|\mathrm{Gal}(K/{\mathbb{F}}_p)|=[K:{\mathbb{F}}_p]=n$. The cyclic subgroup $⟨\sigma ⟩$ already has $n$ elements, so $\mathrm{Gal}(K/{\mathbb{F}}_p)=⟨\sigma ⟩\cong \mathbb{Z}/n$. $\square$

Corollary (Enumeration of subfields). The subfields of ${\mathbb{F}}_{p^n}$ are in order-preserving bijection with the divisors of $n$: for each $d\mid n$ there is exactly one subfield, namely ${\mathbb{F}}_{p^d}$, and the Galois correspondence sends it to the unique subgroup $⟨{\sigma }^d⟩\le ⟨\sigma ⟩$ of order $n/d$.

Proposition (Finite subgroups of a field's unit group are cyclic). Let $K$ be any field and $G\subseteq K^{\times }$ a finite subgroup of its multiplicative group. Then $G$ is cyclic.

Proof. Let $|G|=m$ and let $\lambda$ be the exponent of $G$, the least common multiple of the orders of its elements. Every $g\in G$ satisfies $g^{\lambda }=1$, so every element of $G$ is a root of the polynomial $x^{\lambda }-1\in K[x]$. This polynomial has degree $\lambda$ and therefore at most $\lambda$ roots in the field $K$. Since all $m$ elements of $G$ are roots, $m\le \lambda$. On the other hand $\lambda$ is the order of some element of $G$ (in a finite abelian group the exponent is attained), so $\lambda$ divides $m$ by Lagrange, giving $\lambda \le m$. Hence $\lambda =m$, and an element of order $\lambda =m$ generates $G$, so $G$ is cyclic. $\square$

Applied to $K={\mathbb{F}}_q$ with $G={\mathbb{F}}_q^{\times }$ this recovers the cyclicity of order $q-1$; applied to $K={\mathbb{F}}_p$ it is the existence of primitive roots modulo $p$. The argument fails for $(\mathbb{Z}/n\mathbb{Z}{)}^{\times }$ when $n$ is not a prime power precisely because $\mathbb{Z}/n\mathbb{Z}$ is not a field, so a degree-$\lambda$ polynomial may have more than $\lambda$ roots there — the source of the moduli restriction in 21.01.05.

Connections Master

• Fields and Galois theory 01.04.01. A finite field ${\mathbb{F}}_{p^n}$ is the splitting field of the separable polynomial $x^{p^n}-x$ over ${\mathbb{F}}_p$, so the general Galois correspondence of 01.04.01 specializes here to a perfect lattice-reversal between the subfields ${\mathbb{F}}_{p^d}$ and the subgroups of the cyclic group $⟨{\mathrm{Frob}}_p⟩\cong \mathbb{Z}/n$.

• Group 01.02.01. The cyclicity of ${\mathbb{F}}_q^{\times }$ and of $\mathrm{Gal}({\mathbb{F}}_{p^n}/{\mathbb{F}}_p)$ are both statements that certain naturally-arising groups are cyclic; the polynomial-root-count argument is the prototype of "a finite subgroup of the multiplicative group of a field is cyclic," a group-theoretic fact in the language of 01.02.01.

• Fermat, Euler, Wilson 21.01.04. Fermat's little theorem $a^p\equiv a(\mathrm{mod}p)$ is the statement ${\mathrm{Frob}}_p=\mathrm{id}$ on ${\mathbb{F}}_p$, and the freshman's dream $(a+b{)}^p=a^p+b^p$ used in Euler's induction proof is the additivity of the Frobenius; this unit is the precise target those number-theory arguments forward-reference.

• Primitive roots and the structure of units 21.01.05. The theorem that ${\mathbb{F}}_p^{\times }\cong \mathbb{Z}/(p-1)$, and hence that primitive roots exist modulo every prime, is the case $n=1$ of the cyclicity of ${\mathbb{F}}_q^{\times }$ proved here; the same polynomial-degree argument lifts the result to every ${\mathbb{F}}_q$ without exception.

• Congruences and the Chinese remainder theorem 21.01.03. The construction ${\mathbb{F}}_q\cong {\mathbb{F}}_p[x]/(f)$ for $f$ irreducible of degree $n$ is the polynomial analogue of the modular arithmetic in 21.01.03, and the factorization of $f$ via the Frobenius decomposition is the polynomial Chinese remainder theorem applied to a single maximal ideal.

Historical & philosophical context Master

Évariste Galois introduced finite fields in his 1830 paper Sur la théorie des nombres, where he constructed the field ${\mathbb{F}}_{p^n}$ as congruences in ${\mathbb{F}}_p[x]$ modulo an irreducible polynomial of degree $n$ and observed that the nonzero elements form a cyclic group ^{[Galois 1830]}. This was the first appearance of what are now called Galois fields (hence the notation $GF(p^n)$), and the Frobenius symmetry $x\mapsto x^p$ is implicit in his construction. Galois's insight was that irreducibility of the modulus is exactly the condition that makes division possible, the polynomial analogue of "$p$ prime makes $\mathbb{Z}/p\mathbb{Z}$ a field."

The classification theorem — that every finite field is a Galois field $\mathrm{GF}(p^n)$, unique for each prime $p$ and degree $n$ — was completed by Eliakim Hastings Moore in 1893 ^{[Moore 1893]}. Moore's theorem closed the question of which finite fields exist: the order of a finite field is necessarily a prime power, and every prime power occurs uniquely. The American school then systematized the subject, culminating in Leonard Eugene Dickson's 1901 treatise Linear Groups: With an Exposition of the Galois Field Theory, the first book-length development of finite fields and the classical groups over them ^{[Dickson 1901]}.

The modern abstract treatment, phrasing finite fields as splitting fields of $x^{p^n}-x$ and the Frobenius as the canonical generator of the Galois group, is due to Emil Artin and the mid-twentieth-century algebra textbooks. The subject gained a second life through applications: Reed-Solomon codes (1960), finite-field cryptography (Diffie-Hellman 1976), and the Weil conjectures, where the Frobenius acts on cohomology and governs the point counts of varieties over finite fields.

Bibliography Master

@book{LangAlgebra,
  author = {Lang, Serge},
  title = {Algebra},
  edition = {3},
  publisher = {Springer GTM 211},
  year = {2002},
}

@book{DummitFoote2004,
  author = {Dummit, David S. and Foote, Richard M.},
  title = {Abstract Algebra},
  edition = {3},
  publisher = {Wiley},
  year = {2004},
}

@book{LidlNiederreiter1997,
  author = {Lidl, Rudolf and Niederreiter, Harald},
  title = {Finite Fields},
  edition = {2},
  publisher = {Cambridge University Press},
  year = {1997},
}

@book{ArtinAlgebra,
  author = {Artin, Michael},
  title = {Algebra},
  edition = {2},
  publisher = {Pearson},
  year = {2011},
}

@article{Galois1830,
  author = {Galois, Évariste},
  title = {Sur la théorie des nombres},
  journal = {Bulletin des sciences math\'ematiques de F\'erussac},
  volume = {13},
  pages = {428--435},
  year = {1830},
}

@article{Moore1893,
  author = {Moore, Eliakim H.},
  title = {A doubly-infinite system of simple groups},
  journal = {Bull. Amer. Math. Soc.},
  year = {1893},
}

@book{Dickson1901,
  author = {Dickson, Leonard E.},
  title = {Linear Groups with an Exposition of the Galois Field Theory},
  publisher = {Teubner, Leipzig},
  year = {1901},
}