Motivation
The earlier calculation proved simplicity by listing conjugacy classes and checking their possible unions. That approach becomes unwieldy as the degree grows. A different proof reduces a nontrivial element of a normal subgroup to an element moving at most six points. Only a few cycle types then remain, and explicit commutators produce a three-cycle.
Three-cycles generate the alternating group and, from degree five onward, form one alternating-group conjugacy class. A normal subgroup containing one must therefore contain everything. We will prove each part of this chain and then use it to determine all normal subgroups of the symmetric group. The exceptional degree-four behavior will identify exactly why an extra fixed point matters.
Use the cycle-type and centralizer unit, sign, normality and the earlier proof that three-cycles generate . Permutations act from right to left, with commutator . The support of a permutation is the set of points it moves. A simple group is nontrivial and has no proper nontrivial normal subgroup.
Definitions
Definition 12.1.1
(Centralizing a whole alternating group and being perfect)
The centralizer consists of permutations commuting with every element of . This differs from the centralizer of one chosen permutation.
A group is perfect if its commutator subgroup equals the whole group. Its abelianization is , the universal abelian quotient proved earlier. A subgroup of finite index gives a homomorphism through left multiplication on its cosets; its kernel is the intersection of all conjugates of .
A perfect group can still have many proper subgroups. The condition only rules out a nontrivial abelian quotient. Simplicity instead controls normal subgroups, and we will use nonabelian simplicity to obtain perfectness, not treat these as identical definitions.
Theorems
Theorem 12.1.2
(Alternating simplicity and perfectness)
For every , is nonabelian simple and perfect. In particular its derived series never becomes trivial, so it is not solvable.
Every nontrivial normal subgroup of contains a three-cycle. The proof uses a commutator moving at most six points and an explicit reduction of each possible even cycle type on that support.
Theorem 12.1.3
(Symmetric normal subgroups and their consequences)
For , the only normal subgroups of are 1, and . Also
Every proper subgroup of has index at least , and a point stabilizer realizes index . Equivalently, no nontrivial action of on fewer than points exists. These statements concern these groups; they are not a classification of all finite simple groups.
Proof sketch or proof idea
A permutation commuting with every three-cycle fixes every point
Proof
(Three-point subsets pin down every point)
Suppose and commutes with all three-cycles. Conjugation relabels a cycle, so for every three-element subset , must preserve setwise: the support of a three-cycle on remains its support after conjugation.
Fix a point . The intersection of all three-element subsets containing is exactly . To exclude any other point from that intersection, choose two points different from ; they exist because . The three-element set consisting of and those two points omits . Since preserves every such set, belongs to their intersection and equals . Thus is identity. As three-cycles lie in , this proves for .
A nontrivial normal subgroup has a nontrivial element with support at most six
Proof
(A commutator with a noncommuting three-cycle)
Let and choose . The preceding lemma supplies a three-cycle not commuting with . The element
is nonidentity. It belongs to because it is the product of with the conjugate of inverse by , and lies in . Its support lies inside the union of the support of and its -image, hence contains at most six points.
An even nonidentity permutation cannot move one or two points. On support at most six, its possible nontrivial types are a three-cycle, two transpositions, a five-cycle, two three-cycles, or a four-cycle together with a transposition. This list follows by partitioning its moved points into cycle lengths at least two and checking parity. A four-cycle, six-cycle or three disjoint transpositions is odd, and a three-cycle with a transposition is odd. We now handle every remaining type.
Each small-support case produces a three-cycle inside N
Proof
(Explicit commutators for each small-support type)
Relabeling letters in the following calculations is only notation: for the actual element take its actual cycle entries and choose the indicated three-cycle on them. Normality permits commutation with that three-cycle, which is always even.
If is already a three-cycle, nothing remains. For , choose . Point 5 exists because and is fixed by . Its conjugate is , so
a three-cycle in . This is the step that explicitly needs a fifth point.
For , choose . Its conjugate is , giving
Exactly the same calculation works for , since only its action on 1, 2, 3 enters that conjugation.
Finally suppose . Choose . Its conjugate is and
This is a five-cycle in , so the preceding five-cycle case produces a three-cycle from it. All even possibilities have now been handled by actual nonidentity commutators, not merely a claim that some support becomes smaller.
Proof X-ray
Normality turns a local commutator into a global conclusion
Commuting with a three-cycle creates an element supported on at most six points. The finite case analysis forces a three-cycle while keeping every new element inside the normal subgroup. Conjugacy then supplies all three-cycles, and generation supplies . Small support alone would not prove simplicity without those last two steps.
Repair conjugator parity, then use generation
Proof
(Parity repair and generation by three-cycles)
Any two three-cycles are conjugate in . If a chosen conjugator is odd, multiply it on the right by the transposition of two fixed points of the source three-cycle. Such points exist because . That transposition commutes with the source cycle and changes the conjugator's parity without changing its conjugation result. Thus the two cycles are conjugate in .
The normal subgroup containing one three-cycle consequently contains all of them. They generate , so , proving simplicity. The group is nonabelian, since for example and do not commute. Its commutator subgroup is therefore nontrivial; it is normal and must equal by simplicity. This proves perfectness and the derived-series nonsolvability assertion.
Use the trivial centralizer to determine symmetric normal subgroups
Proof
(Intersecting a normal subgroup with )
Let . Its intersection with is normal in , hence is either 1 or . If it contains , index two leaves only or . If the intersection is 1, every commutator of an element of with an element of lies in both normal subgroups. It must therefore be identity. Thus centralizes , and the centralizer lemma gives .
The sign map kills all commutators, so . A commutator of two overlapping transpositions is a three-cycle. The commutator subgroup is normal in , hence contains all conjugate three-cycles and their generated group . This proves equality. Sign then identifies the abelianization with . In particular is not solvable for , since its first derived term is the nontrivial perfect .
A small action would embed too large a simple group
Proof
(The order bound for small actions)
An action of on points gives a homomorphism into . Its kernel is normal. If the action is nontrivial, simplicity makes the kernel trivial, giving an injection. But for ,
so no injection exists. Every such action must be trivial. For a proper subgroup of index , the coset action is transitive on more than one point and cannot be trivial. Therefore . The natural stabilizer of one point has order and index , proving the minimum is attained. This argument includes , whose point stabilizer is , without requiring that stabilizer itself to be simple.
Worked examples
Worked example 12.1.4
(A double transposition cannot generate a small normal subgroup of )
If a normal subgroup of contains , commute it with . The result is , as computed in the proof. Normality and the single three-cycle conjugacy class then force all three-cycles into the subgroup and hence all of . The three double transpositions supported on four letters, together with the identity, do form a Klein-four subgroup, but it is not normal in . Conjugation can move the fifth fixed point and leave that chosen subgroup.
Worked example 12.1.5
(Two three-cycles in lead through a five-cycle)
Starting with in a normal subgroup, commute with to obtain . In the five-cycle calculation, use its first three successive entries: commuting with gives . Thus two explicitly chosen commutators produce a three-cycle. The argument does not need an class table or any assertion about its exceptional automorphisms.
Worked example 12.1.6
(Every homomorphism from to factors through parity)
For , a homomorphism cannot be injective because exceeds six. Its kernel is therefore or by the normal-subgroup theorem. Consequently it factors through the sign quotient . The image of the generator of must have order dividing two, so it is identity or one of the three transpositions of . This gives exactly four homomorphisms. The three nontrivial maps have the same kernel ; choosing different target transpositions changes the map, not its source kernel.
Counterexample mode
Degree four has a genuine normal Klein-four subgroup
In the identity and the three double transpositions form a normal subgroup . They are preserved as a set under conjugation by cycle relabeling, and is nontrivial proper. The double-transposition commutator used above calls for a fifth point and is unavailable. Three-cycles also split into two classes because only one point is fixed, preventing the same odd-centralizer parity repair. These are concrete failures of proof premises, not exceptions to a theorem stated for .
Common mistakes
Common mistake
(Simple does not mean having no proper subgroups)
Simple does not mean having no proper subgroups; only proper nontrivial normal subgroups are excluded. The conjugator and every three-cycle used in the normal-subgroup argument must belong to . A formal relabeling does not give permission to conjugate a merely -normal subgroup by an odd permutation.
Counterexample mode
Perfect does not imply simple
One might expect the converse of the perfectness step: a perfect group should be simple. Take . Commutators are computed coordinatewise, so , and is perfect. Yet , the kernel of the second projection, is a proper nontrivial normal subgroup. The valid implication runs one way only: a nonabelian simple group is perfect.
Common mistake
(A smaller support must stay nonidentity and inside )
The initial commutator has support at most six, even if the starting element moves many more points. Showing a smaller support is helpful only when the new element is nonidentity and remains in ; the proof checks both. Finally is simple as a cyclic prime-order group but is abelian, while the nonabelian-simple and perfect conclusions here begin at degree five.
Summary
A trivial centralizer guarantees a noncommuting three-cycle. Its commutator reduces any nontrivial normal subgroup to five possible small-support types, each yielding a three-cycle. Conjugacy and generation prove alternating simplicity and perfectness. Intersection with then gives the symmetric normal-subgroup list, parity abelianization and the minimum index of a proper alternating subgroup.
Exercises
Quick check
(Verify the five-cycle commutator identity point by point, remembering the right-to-left convention.)
Solution · Answer
For and , the conjugate is and is . Their product sends 1 to 4, 4 to 2 and 2 to 1; it fixes 3 and 5 and every other point. Hence it is exactly , a three-cycle. Applying permutations in the opposite order would give a different displayed cycle and would not verify the stated convention.
Quick check
(Prove directly from simplicity and the order bound that no proper subgroup of has index less than seven, and identify one of index seven.)
Solution · Answer
A proper subgroup of index gives a nontrivial transitive coset action . Simplicity makes it injective. For , the target order is at most , while , impossible. The stabilizer of one point is of order 360 and index seven. This identifies the minimum without classifying every subgroup.
Quick check
(Embed in by acting on the first points and fixing the others, with and . Compute its centralizer in .)
Solution · Answer
A commuting permutation sends each -orbit to an -orbit of the same size: the first points form the unique orbit of size , while each remaining point is fixed. It therefore preserves the first -point set. Its restriction there centralizes and is identity by the proved lemma. On the remaining points it may be any permutation, all of which commute with because of disjoint support. Thus the centralizer is precisely the symmetric group on those fixed points, isomorphic to , with the trivial conventions for zero or one remaining point.
Quick check
(List all possible abstract quotient groups of for . Explain why itself is not a quotient in this list.)
Solution · Answer
A quotient kernel must be 1, or , giving respectively , or the trivial group. For none is isomorphic to : its order differs from , two and one. occurs as a normal subgroup and as the first derived subgroup, but those facts do not make it an abstract quotient of .