Solvability by Radicals and the Insolubility of the Quintic

⚡ Key Takeaways

One-line statement: A polynomial equation is solvable by radicals if and only if its Galois group is a solvable group; the general quintic is not solvable by radicals because $S_5$ is not a solvable group. Key result in words: There is no general formula expressing the roots of a degree-5 (or higher) polynomial using only the coefficients, the four arithmetic operations, and the extraction of $n$ -th roots. Where it applies: Field theory, Galois theory, the classification of when algebraic equations admit closed-form solutions. Who/when: Conjectured and partially proved by Paolo Ruffini (1799), rigorously established by Niels Henrik Abel (1824), and given its definitive structural explanation by Évariste Galois (1830–32).

💡 The Core Insight

Adjoining a radical $\sqrt[n]{a}$ to a field (containing the needed roots of unity) produces a cyclic Galois extension. A formula built from nested radicals therefore corresponds to a tower of cyclic extensions, whose total Galois group is built by successive abelian "layers" — i.e., it is a solvable group. Conversely, every solvable group can be realized by such a tower. Thus solvability by radicals = solvability of the Galois group. Since the Galois group of the generic quintic is $S_5$ , and $S_5$ is not solvable (its only nontrivial proper normal subgroup $A_5$ is simple and non-abelian), no radical formula can exist.

DefinitionRadical Extension and Solvability by Radicals

A field extension $K/F$ is a radical extension if there is a tower

1

\quad F = K_0 \subseteq K_1 \subseteq \cdots \subseteq K_m = K, \qquad K_{i+1} = K_i(\alpha_i), \quad \alpha_i^{n_i} \in K_i

Derivation hint: each step adjoins a single element whose $n_i$ -th power already lies in the previous field; this is the algebraic encoding of "taking an $n_i$ -th root." A polynomial $f \in F[x]$ is solvable by radicals if its splitting field $E$ over $F$ is contained in some radical extension $K$ of $F$ . Here $F$ is the base field (typically $\mathbb{Q}$ or $\mathbb{Q}(t_1,\dots,t_n)$ ), $\alpha_i$ are the radical generators, and $n_i \in \mathbb{N}$ are the radical degrees.

DefinitionSolvable Group

A finite group $G$ is solvable if it possesses a subnormal series

2

\quad \{e\} = G_0 \triangleleft G_1 \triangleleft \cdots \triangleleft G_r = G, \qquad G_{i+1}/G_i \text{ abelian}

Derivation hint: each quotient of consecutive terms is abelian; equivalently the derived series $G \supseteq G' \supseteq G'' \supseteq \cdots$ terminates at $\{e\}$ . Equivalently, $G$ is solvable iff the derived series $G^{(0)}=G$ , $G^{(k+1)} = [G^{(k)}, G^{(k)}]$ reaches the trivial group in finitely many steps.

TheoremGalois's Solvability Criterion

Let $F$ be a field of characteristic $0$ (or characteristic $p$ with appropriate separability), $f \in F[x]$ , and let $E$ be the splitting field of $f$ over $F$ with Galois group $\mathrm{Gal}(E/F)$ . Then

3

\quad f \text{ is solvable by radicals over } F \iff \mathrm{Gal}(E/F) \text{ is a solvable group}.

Derivation hint: the forward direction refines a radical tower into cyclic steps after adjoining roots of unity; the converse builds a radical tower from a composition series with cyclic factors via Kummer theory.

TheoremInsolubility of the General Quintic — Abel–Ruffini

Let $t_1,\dots,t_5$ be indeterminates and $F = \mathbb{Q}(s_1,\dots,s_5)$ the field of symmetric functions, where $s_i$ are the elementary symmetric polynomials. The general quintic

4

\quad f(x) = x^5 - s_1 x^4 + s_2 x^3 - s_3 x^2 + s_4 x - s_5

Derivation hint: $f$ is the polynomial whose roots are the algebraically independent $t_i$ ; its coefficients are the elementary symmetric functions $s_i = \sigma_i(t_1,\dots,t_5)$ . has Galois group $\mathrm{Gal}(E/F) \cong S_5$ , which is not solvable. Hence the general quintic is not solvable by radicals: no formula in $s_1,\dots,s_5$ using $+,-,\times,\div$ and radicals expresses the roots.

ProofRadical solvability implies solvable Galois group

Step 1 (Reduce to a Galois radical tower). Suppose $E \subseteq K$ with $K/F$ radical via the tower [1]. By enlarging $K$ we may assume $K/F$ is itself Galois (take the Galois closure; a compositum of conjugates of a radical extension is again radical).

Step 2 (Adjoin roots of unity). Let $n = \mathrm{lcm}(n_i)$ and let $\zeta$ be a primitive $n$ -th root of unity. Form $F' = F(\zeta)$ and $K' = K(\zeta)$ . The extension $F'/F$ is abelian (cyclotomic), so $\mathrm{Gal}(F'/F)$ is abelian, hence solvable.

Step 3 (Each radical step is cyclic). Over a field containing the relevant roots of unity, adjoining $\alpha$ with $\alpha^{n} \in K_i'$ gives a Kummer extension $K_{i+1}'/K_i'$ that is Galois with cyclic Galois group of order dividing $n$ . This is because any automorphism sends $\alpha \mapsto \zeta^j \alpha$ , and the assignment $\sigma \mapsto j \bmod n$ is an injective homomorphism into $\mathbb{Z}/n\mathbb{Z}$ :

5

\quad \frac{\sigma(\alpha)}{\alpha} \in \mu_n, \qquad \sigma \mapsto \frac{\sigma(\alpha)}{\alpha} \text{ is a group homomorphism } \mathrm{Gal}(K_{i+1}'/K_i') \hookrightarrow \mu_n.

Derivation hint: $\sigma(\alpha)^n = \sigma(\alpha^n) = \alpha^n$ forces $\sigma(\alpha)/\alpha$ to be an $n$ -th root of unity; multiplicativity follows from $\sigma\tau(\alpha)/\alpha = (\sigma\tau(\alpha)/\tau(\alpha))(\tau(\alpha)/\alpha)$ using that $\mu_n \subseteq K_i'$ is fixed.

Step 4 (Assemble the solvable series). The chain $F \subseteq F' \subseteq K_1' \subseteq \cdots \subseteq K_m' = K'$ gives, via the Galois correspondence, a subnormal series of $\mathrm{Gal}(K'/F)$ whose successive quotients are abelian (cyclotomic at the bottom, cyclic above). Hence $\mathrm{Gal}(K'/F)$ is solvable.

Step 5 (Pass to the quotient). Since $E/F$ is Galois (the splitting field) and $E \subseteq K'$ , $\mathrm{Gal}(E/F)$ is a quotient of $\mathrm{Gal}(K'/F)$ :

6

\quad \mathrm{Gal}(E/F) \cong \mathrm{Gal}(K'/F)\,/\,\mathrm{Gal}(K'/E).

Derivation hint: the fundamental theorem of Galois theory; the subgroup fixing $E$ is normal because $E/F$ is normal, and the quotient is the Galois group of $E/F$ . Quotients of solvable groups are solvable. Therefore $\mathrm{Gal}(E/F)$ is solvable. $\blacksquare$

∎

ProofSolvable Galois group implies radical solvability

Step 1. Let $G = \mathrm{Gal}(E/F)$ be solvable with composition-style series [2] having abelian, indeed cyclic, quotients of prime order. Adjoin a primitive $n$ -th root of unity $\zeta$ for $n = |G|!$ (or $\mathrm{lcm}$ of quotient orders) to obtain $F' = F(\zeta)$ ; this enlargement keeps the group solvable and supplies the roots of unity needed for Kummer theory.

Step 2. To each cyclic quotient $G_{i+1}/G_i$ of order $p$ (prime, $p \mid n$ ), the fixed-field correspondence assigns a degree- $p$ cyclic extension $L_{i+1}/L_i$ . Because $\mu_p \subseteq L_i$ , Kummer theory guarantees $L_{i+1} = L_i(\beta)$ with $\beta^p \in L_i$ ; the existence of such a radical generator follows from the nonvanishing of a Lagrange resolvent:

7

\quad \beta = \sum_{k=0}^{p-1} \zeta^{-k} \sigma^k(\theta), \qquad \beta^p \in L_i,

Derivation hint: for a generator $\sigma$ of the cyclic group and suitable $\theta$ , $\sigma(\beta) = \zeta \beta$ , so $\sigma(\beta^p) = \beta^p$ , i.e. $\beta^p$ is fixed and lies in $L_i$ ; the resolvent is nonzero for some $\theta$ by linear independence of characters.

Step 3. Chaining these radical steps yields a radical tower from $F'$ up to $E(\zeta)$ , and prepending the radical step $F \to F(\zeta) = F'$ (since $\zeta^n = 1$ ) shows $E$ lies in a radical extension of $F$ . Hence $f$ is solvable by radicals. $\blacksquare$

∎

ProofThe general quintic has Galois group

S_5

, which is not solvable

Step 1 (Galois group is $S_5$ ). With $t_1,\dots,t_5$ algebraically independent and $F = \mathbb{Q}(s_1,\dots,s_5)$ , the splitting field is $E = \mathbb{Q}(t_1,\dots,t_5)$ . Every permutation of the $t_i$ fixes the symmetric functions $s_i$ , giving $S_5 \hookrightarrow \mathrm{Gal}(E/F)$ ; conversely $[E:F] \le 5! = 120$ since $E$ is generated by 5 roots of a degree-5 polynomial. Hence $\mathrm{Gal}(E/F) \cong S_5$ .

Step 2 ( $A_5$ is simple). The alternating group $A_5$ of order 60 is simple: its conjugacy classes have sizes $1, 15, 20, 12, 12$ , and no proper nonempty union containing $\{e\}$ (other than $\{e\}$ itself) sums to a divisor of 60 — so it has no nontrivial proper normal subgroup.

Step 3 ( $S_5$ not solvable). The only normal subgroups of $S_5$ are $\{e\}$ , $A_5$ , and $S_5$ . The derived series is

8

\quad S_5 \supsetneq S_5' = A_5 \supsetneq A_5' = A_5 = A_5 = \cdots,

Derivation hint: $[S_5,S_5] = A_5$ because commutators are even; $[A_5,A_5]=A_5$ since $A_5$ is non-abelian and simple, so its commutator subgroup (a normal subgroup) must be all of $A_5$ . The derived series stalls at $A_5 \ne \{e\}$ , so $S_5$ is not solvable.

Step 4 (Conclude). By [3] (the forward direction proven above), since $\mathrm{Gal}(E/F) = S_5$ is not solvable, $f$ is not solvable by radicals. $\blacksquare$

∎

Figure

Lattice correspondence: radical tower of fields (left) vs. solvable series of groups (right). The break for the quintic is the simple block

A_5

.

mathematician · Évariste Galois | 1811–1832

Galois, dead in a duel at twenty, transformed algebra by shifting attention from formulas to the symmetries of roots. Where Ruffini and Abel showed that no quintic formula exists, Galois explained why — and simultaneously provided a criterion deciding, for any specific polynomial, whether radicals suffice. He introduced the group attached to an equation (the "group of the equation"), recognized normal subgroups and quotient groups as the engine of the theory, and coined the notion of a solvable (then "résoluble") group. His ideas, dismissed or lost by Cauchy, Fourier, and Poisson during his lifetime, were finally published by Liouville in 1846 and became the foundation of modern abstract algebra.

🧠 Intuition

Think of solving a polynomial as peeling away symmetry layer by layer. Initially the roots are indistinguishable to you — you only know their symmetric combinations (the coefficients), and the full symmetry group permutes them freely. Each radical you extract is a tool that breaks one cyclic chunk of this symmetry: a clever combination (a resolvent) becomes computable, and the group of remaining ambiguity shrinks by a cyclic factor. For the quadratic, one square root suffices because $S_2$ is already cyclic; for the cubic, a square root plus a cube root peel $S_3 \to A_3 \to \{e\}$ ; for the quartic, the chain $S_4 \to A_4 \to V_4 \to \{e\}$ works because every step is abelian. The quintic stalls because, after peeling off the sign ( $S_5 \to A_5$ ), what remains — $A_5$ — is a rigid, simple block of 60 symmetries with no abelian sub-layer to grab onto. There is no resolvent that breaks it cleanly; the symmetry is "atomic."

ExampleThe cubic is solvable — explicit tower

Given: $f(x) = x^3 + px + q$ over $F = \mathbb{Q}(p,q)$ , with generic Galois group $S_3$ .

Step 1 — check the group is solvable. $S_3 \triangleright A_3 \triangleright \{e\}$ with quotients $S_3/A_3 \cong \mathbb{Z}/2$ and $A_3 \cong \mathbb{Z}/3$ — both cyclic. So $S_3$ is solvable and $f$ is solvable by radicals.

Step 2 — square root (peel $S_3 \to A_3$ ). Adjoin $\sqrt{\Delta}$ where the discriminant is

9

\quad \Delta = -4p^3 - 27q^2.

Derivation hint: $\Delta = \prod_{i<j}(r_i-r_j)^2$ ; the element $\sqrt\Delta = \prod_{i<j}(r_i-r_j)$ is fixed exactly by even permutations, i.e. by $A_3$ .

Step 3 — cube root (peel $A_3 \to \{e\}$ ). Adjoin $\zeta_3$ and form the Lagrange resolvent, leading to Cardano's formula:

10

\quad r = \sqrt[3]{-\tfrac{q}{2} + \sqrt{\tfrac{q^2}{4} + \tfrac{p^3}{27}}} + \sqrt[3]{-\tfrac{q}{2} - \sqrt{\tfrac{q^2}{4} + \tfrac{p^3}{27}}}.

Derivation hint: set $r = u+v$ with $3uv = -p$ ; then $u^3, v^3$ are roots of a quadratic whose discriminant is $-\Delta/108$ , recovering the inner square root.

Boxed answer: $\boxed{\text{Tower } F \subseteq F(\sqrt\Delta) \subseteq F(\sqrt\Delta,\zeta_3,\sqrt[3]{\cdot}) \supseteq E;\ \text{every cubic is solvable by radicals.}}$

ExampleA specific quintic that is NOT solvable by radicals

Given: $f(x) = x^5 - 6x + 3 \in \mathbb{Q}[x]$ .

Step 1 — irreducibility. By Eisenstein at $p = 3$ ( $3 \mid 6, 3 \mid 3, 3^2 \nmid 3$ ), $f$ is irreducible over $\mathbb{Q}$ , so $5 \mid |\mathrm{Gal}(f)|$ and the group is a transitive subgroup of $S_5$ .

Step 2 — count real roots. $f'(x) = 5x^4 - 6$ vanishes at $x = \pm(6/5)^{1/4} \approx \pm 1.046$ , giving exactly two critical points (one local max, one local min). Evaluating: $f(-2)=-17$ , $f(0)=3$ , $f(1)=-2$ , $f(2)=23$ . Sign changes give exactly 3 real roots and hence 2 complex (non-real) roots.

Step 3 — produce a transposition. Complex conjugation fixes the 3 real roots and swaps the 2 conjugate complex roots — it acts as a transposition in $\mathrm{Gal}(f) \subseteq S_5$ .

Step 4 — group is $S_5$ . A transitive subgroup of $S_5$ containing a transposition and an element of order 5 (Cauchy, from $5 \mid |G|$ ) must be all of $S_5$ . By [8], $S_5$ is not solvable.

Boxed answer: $\boxed{\mathrm{Gal}(x^5-6x+3/\mathbb{Q}) \cong S_5 \Rightarrow \text{the roots are NOT expressible by radicals.}}$

⚠️ Common mistake

"Insolubility means the quintic has no solutions." False. Every quintic with complex coefficients has exactly 5 roots (with multiplicity) by the Fundamental Theorem of Algebra. The roots exist and can be computed numerically to arbitrary precision; what fails is their expression in a closed radical formula. Indeed quintics can be solved using elliptic / hypergeometric functions (Hermite, Kronecker, 1858) — just not by radicals.

"Every quintic is unsolvable by radicals." Also false. Many specific quintics are radically solvable, e.g. $x^5 - 2$ (group a subgroup of the metacyclic Frobenius group $F_{20}$ , which is solvable), or any quintic with abelian/solvable Galois group. The theorem concerns the general quintic (group $S_5$ ).

CounterexampleA solvable specific quintic versus the insolvable general case

Consider $x^5 - 2 = 0$ . Its splitting field is $\mathbb{Q}(\sqrt[5]{2}, \zeta_5)$ and its Galois group is the Frobenius group $F_{20} = \mathbb{Z}/5 \rtimes \mathbb{Z}/4$ of order 20, which has the solvable series

11

\quad F_{20} \triangleright \mathbb{Z}/5 \triangleright \{e\}, \qquad F_{20}/(\mathbb{Z}/5) \cong \mathbb{Z}/4,

Derivation hint: the normal Sylow-5 subgroup is generated by the cyclic permutation of the five fifth roots of 2; the quotient $\mathbb{Z}/4$ is the cyclotomic group of $\zeta_5$ . Both quotients are cyclic, so $F_{20}$ is solvable and the roots $\zeta_5^k\sqrt[5]{2}$ are (obviously) radical. This shows the solvability obstruction is a property of the group, not the degree: degree 5 alone does not forbid radicals — only the appearance of an unsolvable group like $S_5$ or $A_5$ does.

connection

The solvability criterion sits at the confluence of several theories. It is the capstone application of the Fundamental Theorem of Galois Theory, which supplies the dictionary between intermediate fields and subgroups used throughout the proofs. The cyclic-extension half rests on Kummer Theory and Cyclotomic Extensions, which characterize abelian extensions adjoining roots of unity. The group-side input — that $A_n$ is simple for $n \ge 5$ — belongs to Simple Groups and Composition Series, and the very notion of solvable group feeds forward into the Classification of Finite Simple Groups. Finally, the philosophy of attaching symmetry groups to structure generalizes into Inverse Galois Problem and, via differential analogues, into Differential Galois Theory (Picard–Vessiot), where one asks when differential equations are solvable in "closed form."

🔬 Where this appears

⏳ History

- 1545 — Cardano's Ars Magna publishes the cubic (Tartaglia/del Ferro) and quartic (Ferrari) formulas, completing radical solutions through degree 4 and prompting the centuries-long quest for the quintic.

- 1770–1771 — Lagrange's Réflexions sur la résolution algébrique des équations introduces resolvents and the permutation theory of roots, the conceptual seed of group theory; he notices the methods fail at degree 5.

- 1799 — Paolo Ruffini publishes a 500-page argument that the general quintic is insolvable, essentially correct but with a gap and ignored by contemporaries.

- 1815–1820s — Cauchy develops permutation-group theory (cycle notation, the theory of substitutions), supplying tools Ruffini lacked.

- 1824 — Niels Henrik Abel gives the first rigorous proof (the Abel–Ruffini theorem), self-published in a terse pamphlet at his own expense.

- 1829–1832 — Évariste Galois writes the memoirs introducing the group of an equation, normal subgroups, and the solvability criterion; submitted to the Académie and lost/rejected before his death in 1832.

- 1846 — Joseph Liouville edits and publishes Galois's manuscripts in his Journal, finally making the theory accessible.

- 1858 — Hermite, Kronecker, and Brioschi solve the general quintic using elliptic modular functions, showing closed-form solution is possible beyond radicals.

- 1870 — Camille Jordan's Traité des substitutions systematizes Galois theory and group theory, cementing them as central mathematics.

open-problem

The Inverse Galois Problem asks the converse-style question: is every finite group the Galois group of some extension of $\mathbb{Q}$ ? Solvable groups are known to be realizable (Shafarevich, 1954), and many simple groups have been realized, but the general problem remains open. This matters here because realizing $S_n$ and $A_n$ over $\mathbb{Q}$ is classical (Hilbert's irreducibility theorem), confirming that genuinely insolvable polynomials over $\mathbb{Q}$ exist in abundance — not as artifacts of generic indeterminates but as concrete equations. Further subtleties: (i) In characteristic $p$ , separability and Artin–Schreier extensions (replacing Kummer theory for degree- $p$ steps) complicate the radical-tower correspondence. (ii) Solvability in radicals vs. "solvability in real radicals" differs — the casus irreducibilis shows some real cubics need complex radicals even when all roots are real. (iii) Computationally, deciding the Galois group of a high-degree polynomial is feasible but subtle; resolvent-based algorithms have been tabulated through degree 15 and beyond, yet effective Galois-group computation for very large degrees remains an active area.

Solvability by Radicals and the Insolubility of the Quintic

Formal Statement

Why It Holds / Derivation

Building Intuition

Worked Examples

Common Misconceptions

Connections

Real-World Applications

Historical Development

Subtleties & Limits

See Also