
Preliminaries
Notation
The symbols N, ω, and nat
are used interchangeably; they all denote the set of natural numbers.
If m is a natural number, we write m:N and say "m has type N." (For the reader unfamiliar with type theory, it's safe in the beginning to think of this as meaning m∈N.)
For m:N, we denote and define m_={0,1,…,m−1}. Let a=(a0,a1,…,am−1) be an mtuple of elements from A.
(As explained in the post on composition of operations, the tuple a may be identified with a function of type m_→A, where a(i)=ai, for each i<m.)
If h:A→A, then h∘a:m_→A is the function whose i-th coordinate is (h∘a)(i)=h(a(i))=h(ai), and we may formally identify the function h∘a:m_→A with its "image tuple" (h(a0),h(a1),…,h(am−1)).
A signature S=(F,ρ) consists of a set F of operation symbols and a function ρ:F→N. We call ρf the arity of the symbol f.
If A is a set and f is a ρf-ary operation on A, then we may write f:Aρf→A. On the other hand, as the natural number ρf denotes the set {0,1,…,ρf−1}, a function a:ρf→A can be identified with its graph, which is simply a ρf-tuple of elements from A; that is, ai:A, for each i:ρf. Then, by identifying the ρf-th power Aρf with the type ρf→A of functions from {0,1,…,ρf−1} to A, we thus identify the function type Aρf→A with the type (ρf→A)→A.
Examples.
a. If g:(m_→A)→A is an m_-ary operation on A and if a:m_→A, then ga=g(a0,a1,…,am−1) has type A.
b. If f:(ρf→B)→B is a ρf-ary operation on B, if a:ρf→A is a ρf-tuple on A, and if h:A→B, then of course h∘a:ρf→B, so f(h∘a) has type B.
Generalized composition. We summarize in a single sentence the final result of our previous post on general composition. If f:(n_→A)→A and if g:Πi:n_(ki_→A)→A and a:Πi:n_(ki_→A), then eval∘fork(g)(a) has type n_→A, which is the domain type of f; therefore, f∘(eval∘fork(g)(a)) has type A, as desired. See the post on general composition for details.
We summarize in a single sentence the final result of our previous post on general composition. If f:(n_→A)→A and if g:Πi:n_(ki_→A)→A and a:Πi:n_(ki_→A), then eval∘fork(g)(a) has type n_→A, which is the domain type of f; therefore, f∘(eval∘fork(g)(a)) has type A, as desired. See the post on general composition post for details.
Elementary facts
Let F be an endofunctor on Set and let (A,fA) and (B,fB) be F-algebras. (The F-algebras post explains what this means.)
-
Let g and h be F-homomorphisms from (A,fA) to (B,fB). That is g∘fA=fB∘Fg (similarly with h in place of g). Let E(g,h)={a:A∣g(a)=h(a)}. Then
a. E(g,h) is a subuniverse of (A,fA).
b. If X⊆A and X generates (A,fA) and g|X=h|X, then g=h.
c. If A and B are finite sets and X generates (A,fA), then |Hom((A,fA),(B,fB))|≤|B||X|.
Proof. Suppose there are m operation symbols in the similarity type of the two algebras, and assume the i-th operation symbol has arity ki. Then FA:∐m−1i=0(ki_→A).
a. Fix arbitrary 0≤i<m and a:ki_→E(g,h). We wish to show that g(fA(inia))=h(fA(inia)), as this will show that E(g,h) is closed under the i-th operation of (A,fA). (Since i was arbitrary this will complete the proof.) But this is trivial since, by definition of an F-algebra homomorphism, we have (g∘fA)(inia)=(fB∘Fg)(inia)=(fB∘Fh)(inia)=(h∘fA)(inia).
b. Suppose the subset X⊆A generates (A,fA) and suppose g|X=h|X. Fix an arbitrary a:A. We show g(a)=h(a). Since X generates (A,fA), there exists a term t and a tuple x:ρt→X of generators such that a=tAx. Therefore, since Fg=Fh on X, we have g(a)=g(tAx)=(tB∘Fg)(x)=(tB∘Fh)(x)=h(tAx)=h(a).
c. By b, a homomorphism is uniquely determined by its restriction to a generating set. If X generates (A,fA), then, since there are exactly |B||X| functions from X to B, we have |Hom((A,fA),(B,fB))|≤|B||X|.
-
Suppose g and h are homomorphisms from (A,fA) to (B,fB) and from (A,fA) to (C,fC), respectively. Assume g is surjective, and kerg⊆kerh. Then there exists a homomorphism k from (B,fB) to (C,fC) such that h=k∘g.
Proof. Define k:B→C as follows: for each b∈B, choose (by Axiom of Choice!) a0∈g−1{b} and let k(b)=h(a0). (Since g is surjective, such an a0 exists for each b∈B.) Fix a∈A. We show h(a)=kg(a). Let a0 be the element of g−1{g(a)} that we chose when defining k at b=g(a). That is, k(b)=h(a0). Then, g(a0)=b=g(a), so (a0,a)∈kerg⊆kerh, so h(a)=h(a0)=k(b)=kg(a), as desired.
To see that k is a homomorphism, let there be m operation symbols and let 0≤i<m be arbitrary. Fix b:ki_→B and a:n_→A be the respective representatives of the g-kernel classes g−1{b(i)} that we chose when defining k. Then, (fC∘Fk)(b)=(fC∘Fk)((Fg)a)=(fC∘Fh)(a)=(h∘fA)(a)=(kg∘fA)(a). (k∘(g∘fA))(a)=(k∘(fB∘Fg))(a)=(k∘fB)((Fg)(a))=(k∘fB)(b).
Subalgebra generation
Clones
-
Let A be a set and S=(F,ρ) a signature and suppose each f∈F is a (ρf)-ary operation on A. Define \begin{aligned} F_0 &= \operatorname{Proj}}}(A);\ F_{n+1} &= F_n \cup \{ f g \mid f \in F, g \colon \rho f \to (F_n \cap (\rho g \to A)) \}, \text{ for } n < \omega. \end{aligned} Then \operatorname{Clo}}}^A(F) = \bigcup_n F_n.
Terms and Free Algebras
-
Let ρ be a similarity type.
a. Tρ(X) is generated by X.
b. For every algebra (A,fA) of type ρ and every function h:X→A there is a unique homomorphism $g\colon \mathbf{T}\rho(X) \to (A, f^A)suchthat{{ \left.\kern-\nulldelimiterspace g \vphantom{\big|} \right|{X} }} = h$.
Proof.* The definition of Tρ(X) exactly parallels the construction in Theorem 1.14 (Bergman 2012). That accounts for (1). For (2), define g(t) by induction on |t|. Suppose |t|=0. Then t \in X \cup {\mathcal{F}}}_0. If t∈X then define g(t)=h(t). For t \in {\mathcal{F}}}_0, g(t)=t(A,fA). Note that since (A,fA) is an algebra of type ρ and t is a nullary operation symbol, t(A,fA) is defined.
For the inductive step, let |t|=n+1. Then t=f(s1,…,sk) for some f \in {\mathcal{F}}}_k and s1,…,sk each of height at most n. We define g(t)=f(A,fA)(g(s1),…,g(sk)).
By its very definition, g is a homomorphism. Finally, the uniqueness of g follows from Exercise 1.16.6 in Bergman (2012).
-
Let (A,fA) and (B,fB) be algebras of type ρ.
a. For every n-ary term t and homomorphism g:(A,fA)→(B,fB), g(t(A,fA)(a1,…,an))=t(B,fB)(g(a1),…,g(an)).
b. For every term t∈Tρ(Xω) and every \theta \in \operatorname{Con}((A, f^A))}}, (A,fA)≡θ(B,fB)⟹t(A,fA)((A,fA))≡θt(A,fA)((B,fB)).
c. For every subset Y of A, \operatorname{Sg}^{(A, f^A)}(Y)}} = \{ t^{(A, f^A)}(a_1,\dots, a_n) : t \in T(X_n), a_i \in Y, i \leq n < \omega\}.
Proof. The first statement is an easy induction on |t|. The second statement follows from the first by taking (B,fB)=(A,fA)/θ and g the canonical homomorphism. For the third statement, again by induction on the height of t, every subalgebra must be closed under the action of t(A,fA). Thus the right-hand side is contained in the left. On the other hand, the right-hand side is clearly a subalgebra containing the elements of Y (take t=x1) from which the reverse inclusion follows.
Birkhoff's Theorem
Let ρ be a similarity type. An identity of type ρ is an ordered pair of terms, written p≈q, from Tρ(Xω). Let (A,fA) be an algebra of type ρ. We say that (A,fA) satisfies p≈q if p(A,fA)=q(A,fA). In this situation, we write (A,fA)⊨p≈q. If K is a class of algebras of type ρ, we write K⊨p≈q if ∀(A,fA)∈K, (A,fA)⊨p≈q. Finally, if Σ is a set of equations, we write K⊨Σ if every member of K satisfies every member of Σ.
Let K be a class of algebras and Σ a set of equations, each of similarity type ρ. We define Id(K)={p≈q:K⊨p≈q} and Mod(Σ)={(A,fA):(A,fA)⊨Σ}. Classes of the form Mod(Σ) are called equational classes, and Σ is called an equational base or an axiomatization of the class. Mod(Σ) is called the class of models of Σ. Dually, a set of identities of the form Id(K) is called an equational theory.
-
For every class K, each of the classes S(K), H(K), P(K), and V(K) satisfies exactly the same identities as does K.
(exercise)
-
K⊨p≈q if and only if for every (A,fA)∈K and every h\in \operatorname{Hom}(\mathbf{T}(X_\omega),(A, f^A))}}, we have h(p)=h(q).
Proof. First assume that K⊨p≈q. Pick (A,fA) and h as in the theorem. Then (A,fA)⊨p≈q⟹p(A,fA)=q(A,fA)⟹p(A,fA)(h(x1),…,h(xn))=q(A,fA)(h(x1),…,h(xn)). Since h is a homomorphism, we get h(p(A,fA)(x1,…,xn))=h(q(A,fA)(x1,…,xn)), i.e., h(p)=h(q).
To prove the converse we must take any (A,fA)∈K and a1,…,an∈A and show that p(A,fA)(x1,…,xn)=q(A,fA)(x1,…,xn). Let h0:Xω→A be a function with h0(xi)=ai for i≤n. By Theorem 4.21 (Bergman, 2012), h0 extends to a homomorphism h from T(Xω) to (A,fA). By assumption h(p)=h(q). Since h(p)=h(p(A,fA)(x1,…,xn))=p(A,fA)(h(x1),…,h(xn))=p(A,fA)(a1,…,an) (and similarly for q) the result follows.
-
Let K be a class of algebras and p≈q an equation. The following are equivalent.
a. K⊨p≈q.
b. (p,q) belongs to the congruence λK on T(Xω).
c. $\mathbf{F}{\mathcal{K}}(X\omega) \models p\approx q$.
Proof. We shall show (a) ⟹ (c) ⟹ (b) ⟹ (a). Throughout the proof we write F for $\mathbf{F}{\mathcal{K}}(X\omega),\mathbf{T}for\mathbf{T}(X_\omega)and\lambdafor\lambda_{\mathcal{K}}$.
Recall that F=T/λ∈SP(K). From (a) and Lemma 4.36 (Bergman, 2012) we get SP(K)⊨p≈q. Thus (c) holds.
From (c), pF(ˉx1,…,ˉxn)=qF(ˉx1,…,ˉxn) where $\bar{x}i = x_i/\lambda.Fromthedefinitionof\mathbf{F},p^{\mathbf{T}}(x_1,\dots, x_n) \equiv\lambda q^{\mathbf{T}}(x_1,\dots, x_n)fromwhich(b)followssincep = p^{\mathbf{T}}(x_1,\dots, x_n)andq = q^{\mathbf{T}}(x_1,\dots, x_n)$.
Finally assume (b). We wish to apply Lemma 4.37. Let (A,fA)∈K and h \in \operatorname{Hom}(\mathbf{T},(A, f^A))}}. Then T/kerh∈S((A,fA))⊆S(K) so kerh⊇λ. Then (b) implies that h(p)=h(q) hence (a) holds, completing the proof.
Remark. The last result tells us that we can determine whether an identity is true in a variety by consulting a particular algebra, namely F(Xω). Sometimes it is convenient to work with algebras free on other generating sets besides Xω. The following corollary takes care of that for us.
-
Let K be a class of algebras, p and q n-ary terms, Y a set and y1,…,yn distinct elements of Y. Then K⊨p≈q if and only if $p^{\mathbf{F}{\mathcal{K}}(Y)}(y_1, \dots, y_n) = q^{\mathbf{F}{\mathcal{K}}(Y)}(y_1, \dots, y_n).Inparticular,\mathcal{K} \models p \approx qifandonlyif\mathbf{F}_{\mathcal{K}}(X_n)\models p \approx q$.
Proof. Since $\mathbf{F}{\mathcal{K}}(Y)\in \mathbf{S}\mathbf{P}(\mathcal{K}),theleft−to−rightdirectionusesthesameargumentasinTheorem4.38(Bergman,2012)(Result3inthissection).Soassumethatp^{\mathbf{F}{\mathcal{K}}(Y)}(y_1, \dots, y_n) = q^{\mathbf{F}_{\mathcal{K}}(Y)}(y_1, \dots, y_n).Toshowthat\mathcal{K} \models p \approx q,let(A, f^A) \in \mathcal{K}anda_1,\dots,a_n \in A.Wemustshowp^{(A, f^A)}(a_1, \dots, a_n) = q^{(A, f^A)}(a_1, \dots, a_n)$.
There is a homomorphism $h\colon \mathbf{F}{\mathcal{K}}(Y) \to (A, f^A)suchthath(y_i) = a_ifori \leq n$. Then \begin{align*} p^{(A, f^A)}(a_1, \dots, a_n) &= p^{(A, f^A)}(h (y_1), \dots, h (y_n)) = h(p^{\mathbf{F}{\mathcal{K}}(Y)}(y_1, \dots, y_n))\ &= h(q^{\mathbf{F}_{\mathcal{K}}(Y)}(y_1, \dots, y_n)) = q^{(A, f^A)}(h(y_1), \dots, h(y_n))\ &= q^{(A, f^A)}(a_1, \dots, a_n).\end{align*} It follows from Lemma 4.36 (Result 1 of this section) that every equational class is a variety. The converse is Birkhoff’s Theorem.
Theorem (Bergman, Thm 4.41) Every variety is an equational class.
Proof. Let W be a variety. We must find a set of equations that axiomatizes W. The obvious choice is to use the set of all equations that hold in W. To this end, take Σ=Id(W). Let ¯W=Mod(Σ). Clearly, W⊆¯W. We shall prove the reverse inclusion.
Let (A,fA)∈¯W and Y a set of cardinality max(|A|,|ω|). Choose a surjection h0:Y→A. By Theorem [thm:4.21], h0 extends to a (surjective) homomorphism h:T(Y)→(A,fA). Furthermore, since $\mathbf{F}{\mathcal{W}}(Y) = \mathbf{T}(Y)/\Theta{\mathcal{W}},thereisasurjectivehomomorphismg \colon \mathbf{T}(Y) \to \mathbf{F}_{\mathcal{W}}$.
We claim that kerg⊆kerh. If the claim is true then by Lemma [ex:1.26.8] there is a map $f\colon \mathbf{F}{\mathcal{W}}(Y) \to (A, f^A)suchthatf {\circ}}g = h.Sincehissurjective,soisf.Hence(A, f^A) \in \mathbf{H}(\mathbf{F}{\mathcal{W}}(Y)) \subseteq \mathcal{W}$ completing the proof.
Let u,v∈T(Y) and assume that g(u)=g(v). Since T(Y) is generated by Y, by Theorem [thm:4.21], there is an integer n, terms p,q∈T(Xn), and y1, …, yn∈Y such that u=pT(Y)(y1,…,yn) and v=qT(Y)(y1,…,yn), by Theorem [thm:4.32]. Applying the homomorphism g, $$p^{\mathbf{F}{\mathcal{W}}(Y)}(y_1,\dots, y_n) = g(u) = g(v) = q^{\mathbf{F}{\mathcal{W}}(Y)}(y_1,\dots, y_n).ThenbyResult4above(Corollary4.39,Bergman,2012),$W⊨p≈q$,hence$(p≈q)∈Σ$.Since$(A,fA)∈¯W=Mod(Σ)$,weget$(A,fA)⊨p≈q$.Therefore,h(u) = p^{(A, f^A)}(h_0(y_1), \dots, h_0(y_n)) = q^{(A, f^A)}(h_0(y_1), \dots, h_0(y_n)) = h(v),$$ as desired.