Introduction

An algebraic structure is a set together with one or more operations that combine elements. Abstracting away from what the elements are (numbers, strings, permutations, matrices) and keeping only the laws the operation obeys lets one theorem cover all of them. Strings under concatenation form a monoid, which is why automata theory and parsing can use algebraic arguments; the integers modulo nn form groups that underlie hashing, checksums and public-key cryptography; permutations of a set form groups used in counting arguments.

The development follows the hierarchy semigroup, monoid, group, abelian group, then covers subgroups, cyclic groups and Lagrange's theorem. Rings and fields are defined at the end for reference.

Prerequisites

  • Functions: a binary operation is a function S×S→SS \times S \to S; bijections and composition.
  • Modular arithmetic: a mod na \bmod n is the remainder in {0,…,n−1}\{0, \ldots, n-1\}; gcd⁡\gcd.

Core concepts

Binary operations and their properties

A binary operation on a set SS is a function ∗:S×S→S* : S \times S \to S. Requiring the result to lie in SS is the closure property; subtraction is a binary operation on Z\mathbb{Z} but not on N\mathbb{N} (2−5∉N2 - 5 \notin \mathbb{N}).

PropertyStatement
Associative(a∗b)∗c=a∗(b∗c)(a * b) * c = a * (b * c) for all a,b,ca, b, c
Commutativea∗b=b∗aa * b = b * a for all a,ba, b
Identitysome ee with e∗a=a∗e=ae * a = a * e = a for all aa
Inverse (given identity ee)for aa, some a−1a^{-1} with a∗a−1=a−1∗a=ea * a^{-1} = a^{-1} * a = e
Idempotent elementa∗a=aa * a = a

On an nn-element set there are nn2n^{n^2} binary operations (each of the n2n^2 table cells holds one of nn values) and nn(n+1)/2n^{n(n+1)/2} commutative ones (only the diagonal and the cells above it are free).

The hierarchy

StructureClosureAssociativeIdentityInversesCommutative
SemigroupYesYes
MonoidYesYesYes
GroupYesYesYesYes (every element)
Abelian groupYesYesYesYesYes
Hierarchy of algebraic structures from bottom to top: magma, semigroup (associative), monoid (identity element), group (every element has an inverse), abelian group (commutative).
Figure 1. Each structure adds one axiom to the one below it.

Some books call a set with just a closed binary operation an algebraic system or groupoid instead of a magma; the axioms above are what matter.

Examples and non-examples

Set and operationSemigroupMonoidGroupAbelian
(Z,+)(\mathbb{Z}, +)YesYes (e=0e = 0)Yes (a−1=−aa^{-1} = -a)Yes
(N,+)(\mathbb{N}, +), with 0∈N0 \in \mathbb{N}YesYesNo (no −1-1)n/a
(Z+,+)(\mathbb{Z}^{+}, +)YesNo (no 0)Non/a
(Z,×)(\mathbb{Z}, \times)YesYes (e=1e = 1)No (2 has no inverse)n/a
(Q−{0},×)(\mathbb{Q} - \{0\}, \times)YesYesYesYes
(Z,−)(\mathbb{Z}, -)No (not associative)NoNon/a
(Σ∗,⋅)(\Sigma^*, \cdot), strings under concatenationYesYes (e=εe = \varepsilon)No for a nonempty alphabet; yes for the empty alphabetn/a for a nonempty alphabet (not a group); yes for the empty alphabet. Concatenation is commutative exactly when ∣Σ∣≤1\lvert\Sigma\rvert\le1.
(P(A),∪)(\mathcal P(A),\cup)YesYes (e=∅e=\emptyset)No for A≠∅A\ne\emptyset; yes for A=∅A=\emptyset (trivial group)Yes for A=∅A=\emptyset; n/a for A≠∅A\ne\emptyset (not a group)
(P(A),⊕)(\mathcal{P}(A), \oplus) (symmetric difference)YesYes (e=∅e = \emptyset)Yes (X−1=XX^{-1} = X)Yes
(Zn,+n)(\mathbb{Z}_n, +_n), addition mod nnYesYesYesYes
(Zn∗,×n)(\mathbb{Z}_n^{*}, \times_n), units mod nnYesYesYesYes
(Sn,∘)(S_n, \circ), permutations of nn objectsYesYesYesOnly for n≤2n \le 2
Invertible n×nn \times n real matrices under productYesYesYesOnly for n=1n = 1

For these modular-unit examples assume n≥2n\ge2. Here Zn∗={a∈{1,…,n−1}∣gcd⁡(a,n)=1}\mathbb{Z}_n^{*} = \{a \in \{1, \ldots, n-1\} \mid \gcd(a, n) = 1\}, of size φ(n)\varphi(n) (Euler's totient). (Zn−{0},×n)(\mathbb{Z}_n - \{0\}, \times_n) is a group if and only if nn is prime.

Basic theorems

For a group (G,∗)(G, *) with identity ee:

  1. The identity is unique. (If e,e′e, e' are identities, e=e∗e′=e′e = e * e' = e'.) This holds already in a monoid.
  2. Each element has a unique inverse. (If b,cb, c are inverses of aa, b=b∗(a∗c)=(b∗a)∗c=cb = b * (a * c) = (b * a) * c = c.) Also holds for invertible elements of a monoid.
  3. (a−1)−1=a(a^{-1})^{-1} = a and (a∗b)−1=b−1∗a−1(a * b)^{-1} = b^{-1} * a^{-1} (the order reverses).
  4. Cancellation: a∗b=a∗c⇒b=ca * b = a * c \Rightarrow b = c, and b∗a=c∗a⇒b=cb * a = c * a \Rightarrow b = c.
  5. The equations a∗x=ba * x = b and y∗a=by * a = b have unique solutions x=a−1∗bx = a^{-1} * b, y=b∗a−1y = b * a^{-1}. Hence in the Cayley table (operation table) of a finite group every element appears exactly once in each row and each column (a Latin square). This is necessary but not sufficient: a Latin square table need not be associative.
  6. If a∗a=ea * a = e for every aa, then GG is abelian. (Proof: ab=(ab)−1=b−1a−1=baab = (ab)^{-1} = b^{-1} a^{-1} = ba.)

Order of a group and of an element

  • The order ∣G∣|G| is the number of elements.
  • The order of an element aa, written o(a)o(a), is the least positive kk with ak=ea^k = e (in additive notation, k⋅a=0k \cdot a = 0); if none exists, aa has infinite order. o(e)=1o(e) = 1, o(a)=o(a−1)o(a) = o(a^{-1}).
  • For integer mm, if aa has finite order dd, then am=ea^m=e if and only if d∣md\mid m; if aa has infinite order, am=ea^m=e if and only if m=0m=0.
  • In (Zn,+n)(\mathbb{Z}_n, +_n), o(a)=n/gcd⁡(a,n)o(a) = n / \gcd(a, n).

Subgroups

A non-empty subset H⊆GH \subseteq G is a subgroup (H≤GH \le G) if it is a group under the same operation. Tests:

  • One-step test: H≠∅H \ne \emptyset and a,b∈H⇒a∗b−1∈Ha, b \in H \Rightarrow a * b^{-1} \in H.
  • Finite test: if HH is finite and non-empty, closure under ∗* alone suffices.

The intersection of two subgroups is a subgroup; their union is a subgroup if and only if one contains the other.

Lagrange's theorem

If GG is a finite group and H≤GH \le G, then ∣H∣|H| divides ∣G∣|G|. The cosets aH={a∗h∣h∈H}aH = \{a * h \mid h \in H\} partition GG into blocks of size ∣H∣|H|, and their number is the index [G:H]=∣G∣/∣H∣[G : H] = |G| / |H|.

Consequences:

  • o(a)o(a) divides ∣G∣|G| (apply Lagrange to the subgroup generated by aa), so a∣G∣=ea^{|G|} = e.
  • A group of prime order pp is cyclic, generated by any non-identity element, and has only the trivial subgroups {e}\{e\} and GG.
  • The converse of Lagrange is false: the alternating group A4A_4 has order 12 but no subgroup of order 6.
  • Applied to Zn∗\mathbb{Z}_n^{*}: aφ(n)≡1(modn)a^{\varphi(n)} \equiv 1 \pmod n when gcd⁡(a,n)=1\gcd(a, n) = 1 (Euler), and ap−1≡1(modp)a^{p-1} \equiv 1 \pmod p for prime p∤ap \nmid a (Fermat).

Cyclic groups

GG is cyclic if G=⟨g⟩={gk∣k∈Z}G = \langle g \rangle = \{g^k \mid k \in \mathbb{Z}\} for some generator gg. Every cyclic group is abelian (the converse fails: the Klein four-group is abelian and not cyclic). Every subgroup of a cyclic group is cyclic.

For a finite cyclic group of order nn:

  • the number of generators is φ(n)\varphi(n); in Zn\mathbb{Z}_n, aa generates if and only if gcd⁡(a,n)=1\gcd(a, n) = 1;
  • for each divisor dd of nn there is exactly one subgroup of order dd, so the number of subgroups is the number of divisors of nn;
  • the number of elements of order dd is φ(d)\varphi(d) for each d∣nd \mid n.

Every group of order 1, 2, 3, 5 or 7 (prime orders and 1) is cyclic; of order 4 it is either Z4\mathbb{Z}_4 or the Klein four-group V4=Z2×Z2V_4 = \mathbb{Z}_2 \times \mathbb{Z}_2. Every group of order at most 5 is abelian; the smallest non-abelian group is S3S_3, of order 6.

Rings and fields (definitions for reference)

  • A ring (R,+,⋅)(R, +, \cdot): (R,+)(R, +) is an abelian group, (R,⋅)(R, \cdot) is a semigroup (these notes also require a multiplicative identity 1, as in many modern texts; some books do not), and multiplication distributes over addition on both sides. Example: (Z,+,×)(\mathbb{Z}, +, \times), (Zn,+n,×n)(\mathbb{Z}_n, +_n, \times_n), n×nn \times n matrices.
  • A field is a commutative ring with 1≠01 \ne 0 in which every non-zero element has a multiplicative inverse, so (F−{0},⋅)(F - \{0\}, \cdot) is an abelian group. Examples: Q,R,C\mathbb{Q}, \mathbb{R}, \mathbb{C}, and Zp\mathbb{Z}_p for prime pp. Zn\mathbb{Z}_n is a field if and only if nn is prime.

Worked examples

Example 1: identify the structure

On Z\mathbb{Z}, define a∗b=a+b−3a * b = a + b - 3. Closure: yes. Associative: (a∗b)∗c=a+b+c−6=a∗(b∗c)(a * b) * c = a + b + c - 6 = a * (b * c). Identity: a∗e=aa * e = a gives e=3e = 3. Inverse: a∗a′=3a * a' = 3 gives a′=6−aa' = 6 - a. Commutative: yes. So (Z,∗)(\mathbb{Z}, *) is an abelian group.

On Z\mathbb{Z}, a∗b=max⁡(a,b)a * b = \max(a, b): associative and commutative, but no identity (it would have to be −∞-\infty), so a semigroup only. On N\mathbb{N} (with 0), max⁡\max has identity 0, giving a commutative monoid that is not a group.

Example 2: element orders in Z12\mathbb{Z}_{12}

o(a)=12/gcd⁡(a,12)o(a) = 12/\gcd(a, 12): o(0)=1o(0) = 1; o(6)=2o(6) = 2; o(4)=o(8)=3o(4) = o(8) = 3; o(3)=o(9)=4o(3) = o(9) = 4; o(2)=o(10)=6o(2) = o(10) = 6; o(1)=o(5)=o(7)=o(11)=12o(1) = o(5) = o(7) = o(11) = 12. Generators: 1,5,7,111, 5, 7, 11, that is φ(12)=4\varphi(12) = 4. Subgroups: one for each divisor 1,2,3,4,6,121, 2, 3, 4, 6, 12, six in all. The counts 1,1,2,2,2,41, 1, 2, 2, 2, 4 of elements of each order equal φ(d)\varphi(d) and add to 12.

Example 3: the group Z10∗\mathbb{Z}_{10}^{*}

Z10∗={1,3,7,9}\mathbb{Z}_{10}^{*} = \{1, 3, 7, 9\}, φ(10)=4\varphi(10) = 4. Powers of 3 mod 10: 3,9,7,13, 9, 7, 1, so 3 generates the group, which is cyclic of order 4. Inverses: 3−1=73^{-1} = 7 (since 21≡121 \equiv 1), 9−1=99^{-1} = 9. Lagrange check: o(9)=2o(9) = 2 divides 4.

Example 4: a non-abelian group

In S3S_3 with composition (apply the right permutation first), let σ=(1 2)\sigma = (1\,2) and τ=(1 2 3)\tau = (1\,2\,3). Then σ∘τ\sigma \circ \tau sends 1→2→11 \to 2 \to 1, 2→3→32 \to 3 \to 3, 3→1→23 \to 1 \to 2, giving (2 3)(2\,3), while τ∘σ=(1 3)\tau \circ \sigma = (1\,3). Since these differ, S3S_3 is not abelian. Element orders in S3S_3: one of order 1, three of order 2, two of order 3; all divide 6, but no element has order 6, so S3S_3 is not cyclic.

Important properties and observations

  • A monoid's invertible elements form a group (the group of units).
  • A finite semigroup with both cancellation laws is a group.
  • Associativity cannot be read off a Cayley table at a glance; check it (or recognise the operation).
  • In additive notation, the identity is 0, the inverse of aa is −a-a, and aka^k becomes kak a.
  • Groups of order p2p^2 (pp prime) are abelian, though not necessarily cyclic (V4V_4 has order 4).

GATE examination perspective

  • Structure identification. Given a set and an operation (possibly a custom one such as a∗b=a+b−aba * b = a + b - ab), check closure, associativity, identity, inverses and commutativity in that order. Check closure and the existence of inverses for every element carefully.
  • Orders and generators. Use o(a)=n/gcd⁡(a,n)o(a) = n/\gcd(a, n) in Zn\mathbb{Z}_n, φ(n)\varphi(n) generators, one subgroup per divisor.
  • Lagrange. Possible subgroup and element orders must divide ∣G∣|G|; remember the converse is false.
  • Trap: (Zn−{0},×n)(\mathbb{Z}_n - \{0\}, \times_n) is a group only for prime nn.
  • Trap: (ab)−1=b−1a−1(ab)^{-1} = b^{-1} a^{-1}, not a−1b−1a^{-1} b^{-1}, in a non-abelian group.

Summary

Sources

  1. J. A. Gallian, Contemporary Abstract Algebra, 8th ed., Brooks/Cole Cengage Learning, 2013: Chapters 2 (Groups), 3 (Finite Groups; Subgroups), 4 (Cyclic Groups) and 7 (Cosets and Lagrange's Theorem).
  2. K. H. Rosen, Discrete Mathematics and Its Applications, 7th ed., McGraw-Hill, 2012: Sections 4.1 (Divisibility and Modular Arithmetic) and 4.4 (Solving Congruences).

Next steps