Bourbaki

§ 1. Laws of composition; associativity; commutativity

1. Laws of composition

Definition 1

Let E be a set. A mapping f of E × E into E is called a law of composition on E. The value f(x, y) of f for an ordered pair (x, y) ∈ E × E is called the composition of x and y under this law. A set with a law of composition is called a magma.

Cited by 1

Cited by I § 7. Free monoids, free groups

The composition of x and y is usually denoted by writing x and y in a definite order and separating them by a characteristic symbol of the law in question (a symbol which it may be agreed to omit). Among the symbols most often used are + and ., the usual convection being to omit the latter if desired; with these symbols the composition of x and y is written respectively as x + y and x.y or xy. A law denoted by the symbol + is usually called addition (the composition x + y being called the sum of x and y) and we say that it is written additively; a law denoted by the symbol . is usually called multiplication (the composition x.y = xy being called the product of x and y) and we say that it is written multiplicatively. In the general arguments of paragraphs 1 to 3 of this chapter we shall generally use the symbols ⊔ and ⊓ to denote arbitrary laws of composition.

By an abuse of language, a mapping of a subset of E × E into E is sometimes called a law of composition not everywhere defined on E.

Example

(1) The mappings (X, Y) ↦ X ∪ Y and (X, Y) ↦ X ∩ Y are laws of composition on the set of subsets of a set E.

(2) On the set N of natural numbers, addition, multiplication and exponentiation are laws of composition (the compositions of x ∈ N and y ∈ N under these laws being denoted respectively by x + y, xy or x.y and x^y) (Set Theory, III, § 3, no. 4).

(3) Let E be a set; the mapping (X, Y) ↦ X ⋅ Y is a law of composition on the set of subsets of E × E (Set Theory, II, § 3, no. 3, Definition 6); the mapping (f, g) ↦ f ⋅ g is a law of composition on the set of mappings of E into E (Set Theory, II, § 5, no. 2).

(4) Let E be a lattice (Set Theory, III, § 1, no. 11); if sup(x, y) denotes the least upper bound of the set {x, y}, the mapping (x, y) ↦ sup(x, y) is a law of composition on E. Similarly for the greatest lower bound inf(x, y). Example 1 above is a particular case of this with P(E)\mathfrak{P}(E) ordered by inclusion.

(5) Let (Et)t∈I(E_t)_{t \in I} be a family of magmas. Let τt\tau_t denote the law of composition on EtE_t. The mapping

((xi),(yi))↦((xiτtyi)) ((x_i), (y_i)) \mapsto ((x_i \tau_t y_i))
is a law of composition on the product E=∏t∈IEtE = \prod_{t \in I} E_t, called the product of the laws τt\tau_t. The set E with this law is called the product magma of the magmas EtE_t. In particular, if all the magmas EtE_t are equal to the same magma M, we obtain the magma of mappings of I from M.

Let (x,y)↦xτy(x, y) \mapsto x \tau y be a law of composition on a set E. Given any two subsets X, Y of E, XτYX \tau Y (provided this notation does not lead to confusion†) will denote the set of elements xτyx \tau y in E, where x∈X,y∈Yx \in X, y \in Y (in other words, the image of X×YX \times Y under the mapping (x,y)↦xτy(x, y) \mapsto x \tau y).

If a∈Ea \in E we usually write aτYa \tau Y instead of {a}τY\{a\} \tau Y and XτaX \tau a instead of Xτ{a}X \tau \{a\}. The mapping (X,Y)↦XτY(X, Y) \mapsto X \tau Y is a law of composition on the set of subsets of E.

Cites 4

Cites E III § 3. Equipotent sets. Cardinals, Definition 6, E II § 3, E II § 5. Product of a family of sets, E III § 1. Order relations. Ordered sets

Definition 2

Let E be a magma and τ\tau denote its law of composition. The law of composition (x,y)↦yτx(x, y) \mapsto y \tau x on E is called the opposite of the above. The set E with this law is called the opposite magma of E.

Cites 1

Cites E IV § 1. Structures and isomorphisms

Let E and E' be two magmas; we shall denote their laws by the same symbol τ\tau. Conforming with the general definitions (Set Theory, IV, § 1, no. 5), a bijective mapping f of E onto E' such that

f(xτy)=f(x)τf(y) f(x \tau y) = f(x) \tau f(y)
for every ordered pair (x,y)∈E×E(x, y) \in E \times E is called an isomorphism of E onto E'. E and E' are said to be isomorphic if there exists an isomorphism of E onto E'.

More generally:

Definition 3

A mapping f of E into E' such that relation (1) holds for every ordered pair (x,y)∈E×E(x, y) \in E \times E is called a homomorphism, or morphism, of E into E'; if E=E′E = E', f is called an endomorphism of E.

Cited by 1

Cited by Theorem 1 (Associativity theorem)

The identity mapping of a magma E is a homomorphism, the composition of two homomorphism is a homomorphism.

† The following is an example where this principle would lead to confusion and should therefore not be used. Suppose that the law of composition in question is the law (A,B)↦A∪B(A, B) \mapsto A \cup B between subsets of a set E; a law of composition

(A,B)↦F(A,B) (\mathscr{A}, \mathscr{B}) \mapsto F(\mathscr{A}, \mathscr{B})
is derived between subsets of P(E)\mathfrak{P}(E), F(A,B)F(\mathscr{A}, \mathscr{B}) being the set of A∪BA \cup B with A∈A,B∈BA \in \mathscr{A}, B \in \mathscr{B}; but F(A,B)F(\mathscr{A}, \mathscr{B}) should not be denoted by A∪B\mathscr{A} \cup \mathscr{B}, as this notation already has a different meaning (the union of A\mathscr{A} and B\mathscr{B} considered as subsets of P(E)\mathfrak{P}(E)).

For a mapping ff of E into E' to be an isomorphism, it is necessary and sufficient that it be a bijective homomorphism and f−1f^{-1} is then an isomorphism of E' onto E.

2. Composition of an ordered sequence of elements

Recall that a family of elements of a set E is a mapping i↦xii \mapsto x_i of a set I (called an indexing set) into E; a family (xi)i∈I(x_i)_{i \in I} is called finite if the indexing set is finite.

A finite family (xi)i∈I(x_i)_{i \in I} of elements of E whose indexing set I is totally ordered is called an ordered sequence of elements of E.

In particular, every finite sequence (xi)i∈H(x_i)_{i \in H}, where H is a finite subset of the set N\mathbf{N} of natural numbers, can be considered as an ordered sequence if H is given the order relation induced by the relation m≤nm \leq n between natural numbers.

Two ordered sequences (xi)i∈I(x_i)_{i \in I} and (yk)k∈K(y_k)_{k \in K} are called similar if there exists an ordered set isomorphism ϕ\phi of I onto K such that yϕ(i)=xiy_{\phi(i)} = x_i for all i∈Ii \in I.

Every ordered sequence (xα)α∈A(x_\alpha)_{\alpha \in A} is similar to a suitable finite sequence. For there exists an increasing bijection of A onto an interval [0,n][0, n] of N\mathbf{N}.

Definition 4

Let* (xα)α∈A(x_\alpha)_{\alpha \in A} be an ordered sequence of elements in a magma E whose indexing set A is non-empty. The composition (under the law T) of the ordered sequence (xα)α∈A(x_\alpha)_{\alpha \in A}, denoted by ∏α∈Axα\prod_{\alpha \in A} x_\alpha, *is the element of E defined by induction on the number of elements in A as follows:

Cited by 1

Cited by Proposition 1, V § 11

(1) if A={β}A = \{\beta\} then ∏α∈Axα=xβ\prod_{\alpha \in A} x_\alpha = x_\beta;

(2) if A has p>1p > 1 elements, β\beta is the least element of A and A′=A−{β}A' = A - \{\beta\}, then ∏α∈Axα=xβ T(∏α∈A′xα)\prod_{\alpha \in A} x_\alpha = x_\beta \ T \left( \prod_{\alpha \in A'} x_\alpha \right).

It follows immediately (by induction on the number of elements in the indexing sets) that the compositions of two similar ordered sequences are equal; in particular, the composition of any ordered sequence is equal to the composition of a finite sequence. If A={λ,μ,ν}A = \{\lambda, \mu, \nu\} has three elements (λ<μ<ν\lambda < \mu < \nu) the composition ∏α∈Axα\prod_{\alpha \in A} x_\alpha is xλ T (xμ T xν)x_\lambda \ T \ (x_\mu \ T \ x_\nu).

Remark

Note that there is a certain arbitrariness about the definition of the composition of an ordered sequence; the induction we introduced proceeds "from right to left". If we proceeded "from left to right", the composition of the above ordered sequence (xλ,xμ,xν)(x_\lambda, x_\mu, x_\nu) would be (xλ T xμ) T xν(x_\lambda \ T \ x_\mu) \ T \ x_\nu.

As a matter of notation, the composition of an ordered sequence (xα)α∈A(x_\alpha)_{\alpha \in A} is written ∏α∈A\prod_{\alpha \in A} for a law denoted by ⊥\perp; for a law written additively it is usually denoted by ∑α∈Axα\sum_{\alpha \in A} x_\alpha and called the sum of the ordered sequence (xα)α∈A(x_\alpha)_{\alpha \in A} (the xαx_\alpha being called the terms of the sum); for a law written multiplicatively it is usually denoted by ∏α∈Axα\prod_{\alpha \in A} x_\alpha and called the product of the ordered sequence (xα)(x_\alpha) (the xαx_\alpha being called the factors of the product).†

When there is no possible confusion over the indexing set (nor over its ordering) it is often dispensed with in the notation for the composition of an ordered sequence and we then write, for example for a law written additively, ∑αxα\sum_\alpha x_\alpha instead of ∑α∈Axα\sum_{\alpha \in A} x_\alpha; similarly for the other notations.

For a law denoted by ⊤\top the composition of a sequence (xi)(x_i) with indexing set a non-empty interval [p,q][p, q] of N\mathbf{N} is denoted by ∏p≤i≤qxi\prod_{p \leq i \leq q} x_i or ∏i=pqxi\prod_{i=p}^q x_i; similarly for laws denoted by other symbols.

Let E and F be two magmas whose laws of composition are denoted by ⊤\top and ff a homomorphism of E into F. For every ordered sequence (xα)α∈A(x_\alpha)_{\alpha \in A} of elements of E

f(∏α∈A)=∏α∈Af(xα). f\left( \prod_{\alpha \in A} \right) = \prod_{\alpha \in A} f(x_\alpha).

3. Associative laws

Definition 5

A law of composition* (x,y)↦x⊤y(x, y) \mapsto x \top y on a set E is called associative if, for all elements x,y,zx, y, z in E,

(x⊤y)⊤z=x⊤(y⊤z). (x \top y) \top z = x \top (y \top z).
A magma whose law is associative is called an associative magma.

The opposite law of an associative law is associative.

Example

(1) Addition and multiplication of natural numbers are associative laws of composition on N\mathbf{N} (Set Theory, III, § 3, no. 3, Corollary to Proposition 5) (2) The laws cited in Examples (1), (3) and (4) of no. 1 are associative.

Cites 2

Cites E III § 3. Equipotent sets. Cardinals, Corollary, E III § 3

Theorem 1 (Associativity theorem)

Let* E be an associative magma whose law is denoted by ⊤\top. Let A be a totally ordered non-empty finite set, which is the union of an ordered sequence of non-empty subsets (Bi)i∈I(B_i)_{i \in I} such that the relations α∈Bi,β∈Bj,i<j\alpha \in B_i, \beta \in B_j, i < j imply α<β\alpha < \beta; let (xα)α∈A(x_\alpha)_{\alpha \in A} be an ordered sequence of elements in E with A as indexing set. Then

∏α∈Axα=∏i∈I(∏α∈Bixα) \prod_{\alpha \in A} x_\alpha = \prod_{i \in I} \left( \prod_{\alpha \in B_i} x_\alpha \right)

Cites 3 · Cited by 5

Cites E II § 5. Product of a family of sets, E III § 3. Equipotent sets. Cardinals, Definition 3

Cited by Exercise 3, § 1, Proposition 2, Theorem 3, Definition 3, I § 2, Proposition 3, I § 7

† The use of this terminology and the notation ∏α∈Axα\prod_{\alpha \in A} x_\alpha must be avoided if there is any risk of confusion with the product of the sets xαx_\alpha defined in the theory of sets (Set Theory, II, § 5, no. 3). However, if the xαx_\alpha are cardinals and addition (resp. multiplication) is the cardinal sum (resp. the cardinal product), the cardinal denoted by ∑α∈Axα\sum_{\alpha \in A} x_\alpha (resp. ∏α∈Axα\prod_{\alpha \in A} x_\alpha) in the above notation is the cardinal sum (resp. cardinal product) of the family (xα)α∈A(x_\alpha)_{\alpha \in A} (Set Theory, III, § 3, no. 3).

§ 1.3

We prove the theorem by induction on the cardinal nn of AA. Let pp be the cardinal of II and hh its least element; let J=I−{h}J = I - \{h\}. If n=1n = 1, as the BiB_i are non-empty, of necessity p=1p = 1 and the theorem is obvious. Otherwise, assuming the theorem holds for an indexing set with at most n−1n - 1 elements, we distinguish two cases:

(a) BhB_h has a single element β\beta. Let C=⋃i∈JBiC = \bigcup_{i \in J} B_i. The left-hand side of (3) is equal, by definition, to xβ⊤(∏α∈Cxα)x_\beta \top \left( \prod_{\alpha \in C} x_\alpha \right); the right-hand side is equal, by definition, to

xβ⊤(∏i∈J(∏α∈Bixα)); x_\beta \top \left( \prod_{i \in J} \left( \prod_{\alpha \in B_i} x_\alpha \right) \right);
the result follows from the fact that the theorem is assumed true for CC and (Bi)i∈J(B_i)_{i \in J}.

(b) Otherwise, let β\beta be the least element of AA (and hence of BhB_h); let A′=A−{β}A' = A - \{\beta\} and let Bi′=A′∩BiB'_i = A' \cap B_i for i∈Ii \in I; then Bi′=BiB'_i = B_i for i∈Ji \in J. The set A′A' has n−1n - 1 elements and the conditions of the theorem hold for A′A' and its subsets Bi′B'_i; hence by hypothesis:

∏α∈A′xα=(∏α∈Bh′xα)⊤(∏i∈J(∏α∈Bixα)). \prod_{\alpha \in A'} x_\alpha = \left( \prod_{\alpha \in B'_h} x_\alpha \right) \top \left( \prod_{i \in J} \left( \prod_{\alpha \in B_i} x_\alpha \right) \right).
Forming the composition of xβx_\beta with each side, we have on the left-hand side side, by definition, ∏α∈Axα\prod_{\alpha \in A} x_\alpha and on the right-hand side, using the associativity,
(xβ⊤(∏α∈Bh′xα))⊤(∏i∈J(∏α∈Bixα)) \left( x_\beta \top \left( \prod_{\alpha \in B'_h} x_\alpha \right) \right) \top \left( \prod_{i \in J} \left( \prod_{\alpha \in B_i} x_\alpha \right) \right)
which is equal, by Definition 3, to the right-hand side of formula (3).

For an associative law denoted by ⊤\top the composition ∏p≤i≤qxi\prod_{p \leq i \leq q} x_i of a sequence (xi)i∈[p,q](x_i)_{i \in [p, q]} is also denoted (since no confusion can arise) by

xp⊤⋯⊤xq. x_p \top \cdots \top x_q.
A particular case of Theorem 1 is the formula
x0⊤x1⊤⋯⊤xn=(x0⊤x1⊤⋯⊤xn−1)⊤xn. x_0 \top x_1 \top \cdots \top x_n = (x_0 \top x_1 \top \cdots \top x_{n-1}) \top x_n.
Consider an ordered sequence of nn terms all of whose terms are equal to the same element x∈Ex \in E. The composition of this sequence is denoted by ∏nx\prod^n x for a law denoted by ⊤\top, ∏nx\prod^n x for a law denoted by ⊥\perp. For a law written multiplicatively the composition is denoted by xnx^n and called the nn-th power of xx. For a law written additively the composition is usually denoted by nxnx. The associativity theorem applied to an ordered sequence all of whose terms are equal gives the equation

n1+n2+⋯+npTx=(n1Tx)T(n2Tx)T⋯T(npTx). \frac{n_1 + n_2 + \cdots + n_p}{T} x = \left( \frac{n_1}{T} x \right) T \left( \frac{n_2}{T} x \right) T \cdots T \left( \frac{n_p}{T} x \right).

In particular, if p=2p = 2,

m+nTx=(mTx)T(nTx) \frac{m+n}{T} x = \left( \frac{m}{T} x \right) T \left( \frac{n}{T} x \right)
and if n1=n2=⋯=np=mn_1 = n_2 = \cdots = n_p = m,
pmTx=pT(mTx). \frac{pm}{T} x = \frac{p}{T} \left( \frac{m}{T} x \right).

If XX is a subset of EE, we sometimes denote, in conformity with the above notation, by pTX\frac{p}{T} X the set X1TX2T⋯TXpX_1 T X_2 T \cdots T X_p, where

X1=X2=⋯=Xp=X; X_1 = X_2 = \cdots = X_p = X;
it is thus the set of all compositions x1Tx2T⋯Txpx_1 T x_2 T \cdots T x_p with x1∈X,x2∈X,…,xp∈Xx_1 \in X, x_2 \in X, \ldots, x_p \in X.

It is important not to confuse this set with the set of pTx\frac{p}{T} x, where xx runs through XX.

4. Stable subsets. Induced laws

Definition 6

A subset AA of a set EE is called stable under the law of composition TT on EE if the composition of two elements of AA belongs to AA. The mapping (x,y)↦xTy(x, y) \mapsto x T y of A×AA \times A into AA is then called the law induced on AA by the law TT. The set AA with the law induced by TT is called a submagma of EE.

In other words, for AA to be stable under a law TT it is necessary and sufficient that ATA⊂AA T A \subset A. A stable subset of EE and the corresponding submagma are often identified.

The intersection of a family of stable subsets of EE is stable; in particular there exists a smallest stable subset AA of EE containing a given subset XX; it is said to be generated by XX and XX is called a generating system of AA or a generating set of AA. The corresponding submagma is also said to be generated by XX.

Proposition 1

Let EE and FF be two magmas and ff a homomorphism of EE into FF. (i) The image under ff of a stable subset of EE is a stable subset of FF. (ii) The inverse image under ff of a stable subset FF is a stable subset of EE. (iii) Let XX be a subset of EE. The image under ff of the stable subset of EE generated by XX is the stable subset of FF generated by f(X)f(X). (iv) If gg is a second homomorphism of EE from FF the set of elements xx of EE such that f(x)=g(x)f(x) = g(x) is a stable subset of EE.

Cited by 2

Cited by Lemma 2, I § 2, Corollary 3, I § 4

Assertions (i), (ii) and (iv) are obvious; we prove (iii). Let X‾\overline{X} be the stable subset of E generated by X and f(X)‾\overline{f(X)} the stable subset of F generated by f(X)f(X). By (i) f(X)‾⊂f(X‾)\overline{f(X)} \subset f(\overline{X}) and by (ii) X‾⊂f−1(f(X)‾)\overline{X} \subset f^{-1}(\overline{f(X)}), whence f(X‾)⊂f(X)‾f(\overline{X}) \subset \overline{f(X)}.

Proposition 2

Let E be an associative magma and X a subset of E. Let X' be the set of x1⊤x2⊤⋯⊤xnx_1 \top x_2 \top \cdots \top x_n, where n≥1n \geq 1 and where xi∈Xx_i \in X for 1≤i≤n1 \leq i \leq n. The stable subset generated by X is equal to X'.

Cites 1

Cites Theorem 1 (Associativity theorem)

It follows immediately by induction on n that the composition of an ordered sequence of n terms belonging to X belongs to the stable subset generated by X; it is therefore sufficient to verify that X' is stable. Now if u and v are elements of X' they are of the form u=x0⊤x1⊤⋯⊤xn−1u = x_0 \top x_1 \top \cdots \top x_{n-1}, v=xn⊤xn+1⊤⋯⊤xn+pv = x_n \top x_{n+1} \top \cdots \top x_{n+p} with xi∈Xx_i \in X for 0≤i≤n+p0 \leq i \leq n + p; then (Theorem 1) u⊤v=x0⊤x1⊤⋯⊤xn+pu \top v = x_0 \top x_1 \top \cdots \top x_{n+p} belongs to X'.

Example

(1) In the set N\mathbf{N} of natural numbers the stable subset under addition generated by {1} is the set of integers ≥1\geq 1; under multiplication the set {1} is stable.

(2) Given a law ⊤\top on a set E, for a subset {h} consisting of a single element to be stable under the law ⊤\top it is necessary and sufficient that h⊤h=hh \top h = h; h is then said to be idempotent. For example, every element of a lattice is idempotent for each of the laws sup and inf.

(3) For an associative law ⊤\top on a set E the stable subset generated by a set {a} consisting of a single element is the set of elements ⊤‾na\overline{\top}^n a, where n runs through the set of integers >0> 0.

5. Permutable elements. Commutative laws

Definition 7

Let E be a magma whose law is denoted by ⊤\top. Two elements x and y of E are said to commute (or to be permutable) if y⊤x=x⊤yy \top x = x \top y.

Definition 8

A law of composition on a set E is called commutative if any two elements of E commute under this law. A magma whose law of composition is commutative is called a commutative magma.

A commutative law is equal to its opposite.

Example

(1) Addition and multiplication of natural numbers are commutative laws on N\mathbf{N} (Set Theory, III, § 3, no. 3, Corollary to Proposition 5).

(2) On a lattice the laws sup and inf are commutative; so, in particular, are the laws ∪\cup and ∩\cap between subsets of a set E.

(3) Let E be a set of cardinal >1> 1. The law (f,g)↦f∘g(f, g) \mapsto f \circ g between mappings of E into E is not commutative as is seen by taking f and g to be distinct constant mappings, but the identity mapping is permutable with every mapping.

(4) Let (x,y)↦x⊤y(x, y) \mapsto x \top y be a commutative law on EE; the law

(X,Y)↦X⊤Y (X, Y) \mapsto X \top Y
between subsets of EE is commutative.

Cites 2

Cites E III § 3. Equipotent sets. Cardinals, Corollary, E III § 3

Definition 9

Let EE be a magma and XX a subset of EE. The set of elements of EE which commute with each of the elements of XX is called the centralizer of XX.

Cited by 1

Cited by Proposition 4, I § 5

Let XX and YY be two subsets of EE and X′X' and Y′Y' their respective centralizers. If X⊂YX \subset Y, then Y′⊂X′Y' \subset X'.

Let (Xi)i∈I(X_i)_{i \in I} be a family of subsets of EE and for all i∈Ii \in I let Xi′X'_i be the centralizer of XiX_i. The centralizer of ⋃i∈IXi\bigcup_{i \in I} X_i is ⋂i∈IXi′\bigcap_{i \in I} X'_i.

Let XX be a subset of EE and X′X' the centralizer of XX. The centralizer X′′X'' of X′X' is called the bicentralizer of XX. Then X⊂X′′X \subset X''. The centralizer X′′′X''' of X′′X'' is equal to X′X'. For X′X' is contained in its bicentralizer X′′′X''' and the relation X⊂X′′X \subset X'' implies X′′′′⊂X′X'''' \subset X'.

Proposition 3

Let EE be an associative magma whose law is denoted by ⊤\top. If an element xx of EE commutes with each of the elements yy and zz of EE, it commutes with y⊤zy \top z.

For

x⊤(y⊤z)=(x⊤y)⊤z=(y⊤x)⊤z=y⊤(x⊤z)=y⊤(z⊤x)=(y⊤z)⊤x. x \top (y \top z) = (x \top y) \top z = (y \top x) \top z \\ = y \top (x \top z) = y \top (z \top x) = (y \top z) \top x.

Corollary

Let EE be an associative magma. The centralizer of any subset of EE is a stable subset of EE.

Cited by 1

Cited by Definition 10

Definition 10

The centralizer of a magma EE is called the centre of EE. An element of the centre of EE is called a central element of EE.

Cites 1

Cites Corollary

If EE is an associative magma its centre is a stable subset by the Corollary to Proposition 3 and the law induced on its centre is commutative.

Proposition 4

Let EE be an associative magma, XX and YY two subsets of EE. If every element of XX commutes with every element of YY every element of the stable subset generated by XX commutes with every element of the stable subset generated by YY.

Let X′X' and X′′X'' be the centralizer and bicentralizer of XX. They are stable subsets of EE. Now X⊂X′′X \subset X'' and Y⊂X′Y \subset X' and hence X′′X'' (resp. X′X') contains the stable subset of EE generated by XX (resp. YY). As every element of X′′X'' commutes with every element of X′X', the proposition follows.

Corollary 1

If xx and yy are permutable under the associative law ⊤\top so are ⊤mx\top^m x and ⊤ny\top^n y for all integers m>0m > 0 and n>0n > 0; in particular ⊤mx\top^m x and ⊤nx\top^n x are permutable for all xx and all integers m>0m > 0 and n>0n > 0.

Corollary 2

If all pairs of elements of a subset X are permutable under an associative law T, the law induced by T on the stable subset generated by X is associative and commutative.

Cited by 2

Cited by Corollary 2, I § 4, I § 7. Free monoids, free groups

Theorem 2 (Commutativity theorem)

Let T be an associative law of composition on E; let (xα)α∈A(x_\alpha)_{\alpha \in A} be a non-empty finite family of elements of E which are pairwise permutable; let B and C be two totally ordered sets with A as underlying set. Then*

∏α∈Bxα=∏α∈Cxα. \prod_{\alpha \in B} x_\alpha = \prod_{\alpha \in C} x_\alpha.

Cited by 2

Cited by Theorem 3, Proposition 11, I § 7

Since the theorem is true if A has a single element, we argue by induction on the number pp of elements in A. Let pp be an integer > 1 and suppose the theorem is true when Card A < pp. We prove it for Card A = pp. It may be assumed that A is the interval [0,p−1][0, p - 1] in N\mathbf{N}; the composition of the ordered sequence (xα)α∈A(x_\alpha)_{\alpha \in A} defined by the natural order relation on A is ∏i=0p−1xi\prod_{i=0}^{p-1} x_i.

Let A be given another total ordering and let h be the least element of A under this ordering and A' the set of other elements of A (totally ordered by the induced ordering). Suppose first 0<h<p−10 < h < p - 1 and let P = {0,1,…,h−1}\{0, 1, \ldots, h - 1\} and Q = {h+1,…,p−1}\{h + 1, \ldots, p - 1\}; the theorem being assumed true for A', applying the associativity theorem, we obtain (since A′=P∪QA' = P \cup Q)

∏α∈A′xα=(∏i=0h−1xi)∘(∏i=h+1p−1xi) \prod_{\alpha \in A'} x_\alpha = \left( \prod_{i=0}^{h-1} x_i \right) \circ \left( \prod_{i=h+1}^{p-1} x_i \right)

whence, composing xhx_h with both sides and repeatedly applying the commutativity and associativity of T:

∏α∈Axα=xh∘(∏α∈A′xα)=xh∘(∏i=0h−1xi)∘(∏i=h+1p−1xi)=(∏i=0h−1xi)∘xh∘(∏i=h+1p−1xi)=∏i=0p−1xi, \begin{align*} \prod_{\alpha \in A} x_\alpha &= x_h \circ \left( \prod_{\alpha \in A'} x_\alpha \right) = x_h \circ \left( \prod_{i=0}^{h-1} x_i \right) \circ \left( \prod_{i=h+1}^{p-1} x_i \right) \\ &= \left( \prod_{i=0}^{h-1} x_i \right) \circ x_h \circ \left( \prod_{i=h+1}^{p-1} x_i \right) = \prod_{i=0}^{p-1} x_i, \end{align*}

which proves the theorem for this case. If h=0h = 0 or h=p−1h = p - 1, the same result follows, but more simply, the terms arising from P or the terms arising from Q not appearing in the formulae.

Under a commutative associative law on a set E the composition of a finite family (xα)α∈A(x_\alpha)_{\alpha \in A} of elements of E is by definition the common value of the composition of all the ordered sequences obtained by totally ordering A in all possible ways. This composition will still be denoted by ∏α∈Axα\prod_{\alpha \in A} x_\alpha under a law denoted by T; similarly for other notations.

Theorem 3

Let T be an associative law on E and (xα)α∈A(x_\alpha)_{\alpha \in A} a non-empty finite family of elements of E which are pairwise permutable. If A is a union of non-empty subsets (Bi)i∈I(B_i)_{i \in I} which are pairwise disjoint, then

Cites 2 · Cited by 1

Cites Theorem 1 (Associativity theorem), Theorem 2 (Commutativity theorem)

Cited by Definition 3, I § 2

∏α∈Axα=∏i∈I(∏α∈Bixα). \prod_{\alpha \in A} x_\alpha = \prod_{i \in I} \left( \prod_{\alpha \in B_i} x_\alpha \right).

This follows from Theorem 2 if AA and II are totally ordered so that the BiB_i satisfy the conditions of Theorem 1.

We single out two important special cases of this theorem:

1. If (xαβ)(α,β)∈A×B(x_{\alpha \beta})_{(\alpha, \beta) \in A \times B} is a finite family of permutable elements of an associative magma whose indexing set is the product of two non-empty finite sets A,BA, B (a "double family"), then

∏(α,β)∈A×Bxαβ=∏α∈A(∏β∈Bxαβ)=∏β∈B(∏α∈Axαβ) \prod_{(\alpha, \beta) \in A \times B} x_{\alpha \beta} = \prod_{\alpha \in A} \left( \prod_{\beta \in B} x_{\alpha \beta} \right) = \prod_{\beta \in B} \left( \prod_{\alpha \in A} x_{\alpha \beta} \right)

as follows from Theorem 3 by considering A×BA \times B as the union of the sets {α}×B\{\alpha\} \times B on the one hand and of the sets A×{β}A \times \{\beta\} on the other.

For example, if BB has nn elements and for each α∈A\alpha \in A all the xαβx_{\alpha \beta} have the same value xαx_\alpha, then

∏α∈A(∏β=1nxα)=∏α=1n(∏α∈Axα). \prod_{\alpha \in A} \left( \prod^n_{\beta=1} x_\alpha \right) = \prod^n_{\alpha=1} \left( \prod_{\alpha \in A} x_\alpha \right).

If BB has two elements, we obtain the following results: let (xα)α∈A,(yα)α∈A(x_\alpha)_{\alpha \in A}, (y_\alpha)_{\alpha \in A} be two non-empty families of elements of EE. If the xαx_\alpha and the yβy_\beta are pairwise permutable, then

∏α∈Axα∏α∈Ayα=(∏α∈Axα)∏α∈Ayα. \prod_{\alpha \in A} x_\alpha \prod_{\alpha \in A} y_\alpha = \left( \prod_{\alpha \in A} x_\alpha \right) \prod_{\alpha \in A} y_\alpha.

Because of formula (7) the composition of a double sequence (xij)(x_{ij}) whose indexing set is the product of two intervals {p,q}\{p, q\} and {r,s}\{r, s\} in N\mathbf{N} is often denoted for a commutative associative law written additively by

∑i=pq∑j=rsxijor∑j=rs∑i=pqxij \sum_{i=p}^q \sum_{j=r}^s x_{ij} \quad \text{or} \quad \sum_{j=r}^s \sum_{i=p}^q x_{ij}

and similarly for laws denoted by other symbols.

2. Let nn be an integer >0> 0 and let AA be the set of ordered pairs of integers (i,j)(i, j) such that 0≤i≤n,0≤j≤n0 \leq i \leq n, 0 \leq j \leq n and i<ji < j; the composition of a family (xij)(i,j)∈A(x_{ij})_{(i, j) \in A} (under a commutative associative law) is also denoted by ∏0≤i<j≤nxij\prod_{0 \leq i < j \leq n} x_{ij} (or simply ∏i<jxij\prod_{i < j} x_{ij}, if no confusion arises); Theorem 3 here gives the formulae

∏0≤i<j≤nxij=∏i=0n−1(∏j=i+1nxij)=∏j=1n(∏i=0j−1xij). \prod_{0 \leq i < j \leq n} x_{ij} = \prod_{i=0}^{n-1} \left( \prod_{j=i+1}^n x_{ij} \right) = \prod_{j=1}^n \left( \prod_{i=0}^{j-1} x_{ij} \right).

There are analogous formulae to (7) for a family whose indexing set is the product of more than two sets and analogous formulae to (10) for a family whose indexing set is the set SpS_p of strictly increasing sequences (ik)1≤k≤p(i_k)_{1 \leq k \leq p} of pp integers such that 0≤ik≤n0 \leq i_k \leq n (p≤n+1p \leq n + 1): in the latter case the composition of the family (xi1i2…ip)(i1,…,ip)∈Sp(x_{i_1 i_2 \ldots i_p})_{(i_1, \ldots, i_p) \in S_p} is denoted by

∏0≤i1<i2<⋯<ip<nxi1i2…ipor simply∏i1<i2<⋯<ipxi1i2…ip. \prod_{0 \leq i_1 < i_2 < \cdots < i_p < n} x_{i_1 i_2 \ldots i_p} \quad \text{or simply} \quad \prod_{i_1 < i_2 < \cdots < i_p} x_{i_1 i_2 \ldots i_p}.

Proposition 5

Let E and F be magmas whose laws are denoted by T and let f and g be homomorphisms of E into F. Let f T g be the mapping x ↦ f(x) T g(x) of E into F. If F is associative and commutative, f T g is a homomorphism.

Cited by 2

Cited by I § 7. Free monoids, free groups, I § 8. Rings

For all elements x and y of E:

(fTg)(xTy)=f(xTy)Tg(xTy)=f(x)Tf(y)Tg(x)Tg(y)(f T g)(x T y) = f(x T y) T g(x T y) = f(x) T f(y) T g(x) T g(y)
=f(x)Tg(x)Tf(y)Tg(y)=((fTg)(x))T((fTg)(y)).= f(x) T g(x) T f(y) T g(y) = ((f T g)(x)) T ((f T g)(y)).

6. Quotient laws

Definition 11

Let E be a set. A law of composition T and an equivalence relation R on E are said to be compatible if the relations x ≡ x' (mod R) and y ≡ y' (mod R) (for x, x', y, y' ∈ E) imply x T y ≡ x' T y' (mod R); the law of composition on the quotient set E/R which maps the equivalence classes of x and y to the equivalence class of x T y is called the quotient law of the law T with respect to R. The set E/R with the quotient law is called the quotient magma of E with respect to R.

Cites 1

Cites E II § 6. Equivalence relations

To say that an equivalence relation R on E is compatible with the internal law of composition f: E × E → E on E means that the mapping f is compatible (in the sense of Set Theory, II, § 6, no. 5) with the product equivalence relation R × R on E × E and the equivalence relation R on E. (Set Theory, II, § 6, no. 8). This also means that the graph of R is a submagma of E × E.

If the law T is associative (resp. commutative) so is the quotient law (more briefly we say that associativity, or commutativity, is preserved when passing to the quotient).

The canonical mapping from the magma E to the magma E/R is a homomorphism.

For a mapping g of E/R into a magma F to be a homomorphism it is necessary and sufficient that the composition of g with the canonical mapping of E onto E/R be a homomorphism.

The two following propositions are immediate from the definitions:

Proposition 6

Let E and F be two magmas and f a homomorphism of E into F. Let R[x, y] denote the relation f(x) = f(y) between elements x, y of E. Then R is an equivalence relation on E compatible with the law on E and the mapping of E/R onto f(E) derived from f by passing to the quotient is an isomorphism of the quotient magma E/R onto the submagma f(E) of F.

Proposition 7

Let E be a magma and R an equivalence relation on E compatible with the law on E. For an equivalence relation S on E/R to be compatible with the quotient law it is necessary and sufficient that S be of the form T/R where T is an equivalence relation on E implied by R and compatible with the law on E. The canonical mapping of E/T onto (E/R)/(T/R) (Set Theory, II, § 6, no. 7) is then a magma isomorphism.

Cites 1

Cites E II § 6. Equivalence relations

Proposition 8

Let E be a magma, A a stable subset of E and R an equivalence relation on E compatible with the law on E. The saturation B of A with respect to R (Set

Cites 1

Cites E III § 1. Order relations. Ordered sets

Theory, II, , no. 5) is a stable subset. The equivalence relations RAR_A and RBR_B induced by RR on AA and BB respectively are compatible with the induced laws and the mapping derived from the canonical injection of AA into BB by passing to the quotients is a magma isomorphism of A/RAA / R_A onto B/RBB / R_B

Let ⊤\top denote the law on EE. If xx and yy are two elements of BB there exist two elements x′x' and y′y' of AA such that x≡x′(modR)x \equiv x' \pmod{R} and y≡y′(modR)y \equiv y' \pmod{R}; then x⊤y≡x′⊤y′(modR)x \top y \equiv x' \top y' \pmod{R} and x′⊤y′∈Ax' \top y' \in A, whence x⊤y∈Bx \top y \in B. Thus BB is a stable subset of EE and the other assertions are obvious.

Let MM be a magma and ((uα,vα))α∈I((u_\alpha, v_\alpha))_{\alpha \in I} a family of elements of M×MM \times M. Consider all the equivalence relations SS on MM which are compatible with the law on MM and such that uα≡vα(modS)u_\alpha \equiv v_\alpha \pmod{S} for all α∈I\alpha \in I. The intersection of the graphs of these relations is the graph of an equivalence relation RR which is compatible with the law on MM and such that uα≡vα(modR)u_\alpha \equiv v_\alpha \pmod{R}. Hence RR is the finest (Set Theory, III, § 1, nos. 3 and 7) equivalence relation with these two properties. It is called the equivalence relation compatible with the law on MM generated by the (uα,vα)(u_\alpha, v_\alpha).

Proposition 9

Preserving the above notation, let ff be a homomorphism of MM into a magma such that f(uα)=f(vα)f(u_\alpha) = f(v_\alpha) for all α∈I\alpha \in I. Then ff is compatible with RR.

Cited by 2

Cited by Proposition 2, I § 7, Proposition 3, I § 7

Let TT be the equivalence relation associated with ff. Then uα≡vα(modT)u_\alpha \equiv v_\alpha \pmod{T} for all α∈I\alpha \in I and TT is compatible with the law on MM, hence TT is coarser than RR; this proves the proposition.

Exercises

See the exercises for § 1.

Transcribed from the printed edition, 1998, Springer. Not translated.