In this paragraph, X denotes a set, F(X) the free group constructed on X (Algebra, Chapter I, § 7, no. 5) and A(X) the free associative algebra constructed on X with its total graduation (An(X))n≥0 (cf. Algebra, Chapter III, § 3, no. 1, Example 3). X is identified with its images in F(X) and A(X).
1. Magnus algebras
Let A^(X) be the product module ∏n≥0An(X). We define on A^(X) a multiplication by the rule
(a⋅b)n=i=0∑nai⋅bn−i
where a=(an) and b=(bn) are in A^(X). We know (Commutative Algebra, Chapter III, § 2, no. 12, Example 1) that A^(X) is an associative algebra and that A(X) is identified with the subalgebra of A^(X) consisting of the sequence all of whose terms are zero except for a finite number.
A^(X) is given the product topology of the discrete topologies on the factors
A^n(X); this topology makes A^(X) into a complete Hausdorff topological algebra, when the ring K has the discrete topology, and A(X) is dense in A^(X)
Let a=(an)∈A^(X); the family (an)n≥0 is summable and a=∑n≥0an.
For every integer m≥0, let A^m(X) denote the ideal consisting of the series a=∑n≥man such that an∈An(X) for all n≥m. This sequence of ideals is a fundamental system of neighbourhoods of 0 in A^(X) and an integral filtration on A^(X). The order function associated with the above filtration is denoted by ω; then ω(0)=+∞ and ω(a)=m if a=∑n≥man with an∈An(X) for all n≥m and am=0 (§ 4, nos. 1 and 2).
A^(X) is called the Magnus algebra of the set X with coefficients in K. If there is any ambiguity over K we write A^K(X).
Proposition 1
Let B be a unital associative algebra with a real filtration (Bα)α∈R such that B is Hausdorff and complete (§ 4, nos. 1 and 2). Let f be a mapping of X into B such that there exists λ>0 for which f(X)⊂Bλ. Then f can be extended in one and only one way to a continuous unital homomorphism f^ of A^(X) into B.
Let f′ be the unique unital algebra homomorphism of A(X) into B extending f (Algebra, Chapter III, § 2, no. 7, Proposition 7). We show that f′ is continuous: f′(An(X))⊂Bnλ whence f′(A^n(X)∩A(X))⊂Bnλ. Therefore f′ can be extended in one and only one way by continuity to a homomorphism f^:A^(X)→B.
We preserve the hypotheses and notation of Proposition 1 and let u∈A^(X). The element f^(u) is denoted by u((f(x))x∈X) and called the result of substituting the f(x) for the x in u. In particular, u((x)x∈X)=u. Now let u=(uy)y∈Y be a family of elements of A^1(X) and let v∈A^(Y). The above allows us to define the element v((uy)y∈Y)∈A^(X). It is denoted by v∘u. As uy((f(x)))∈Bλ, the elements uy((f(x))) can be substituted for the y in v. The mappings v↦(v∘u)((f(x))) and v↦v((uy((f(x)))))) are then two continuous homomorphisms of unital algebras of A^(X) into B taking the same value uy((f(x))) at the element y∈Y. Therefore (Proposition 1)
(v∘u)((f(x)))=v((uy((f(x))))))
for all v∈A^(Y).
2. Magnus group
For all a=(an)n≥0 in A^(X), the element a0 of K will be called the constant term of a and denoted by ε(a). Formula (1) shows that ε is an algebra homomorphism of A^(X) into K.
Lemma 1
For an element a of A^(X) to be invertible, it is necessary and sufficient that its constant term be invertible in K.
If a is invertible in A^(X), ε(a) is invertible in K. Conversely, if ε(a) is invertible in K, there exists u∈A^1(X) such that a=ε(a)(1−u); we write b=(∑n≥0un)ε(a)−1. Then ab=ba=1 and a is invertible.
The set of elements of A^(X) of constant term 1 is therefore a subgroup of the multiplicative monoid A^(X), called the Magnus group constructed on X (relative to K). In this chapter it will be denoted by Γ(X) or simply Γ. For every integer n≥1, we denote by Γn the set of a∈Γ such that ω(a−1)≥n. By Proposition 2 of § 4, no. 5, the sequence (Γn)n≥1 is an integral central filtration on Γ.
3. Magnus group and free group
Theorem 1
Let r be a mapping of X into A^(X) such that ω(r(x))≥2 for all x∈X. The unique homomorphism g of the free group F(X) into the Magnus group Γ(X) such that g(x)=1+x+r(x) for all x∈X is injective.
We first prove three lemmas.
Lemma 2
Let n be a non-zero rational integer. In the ring of formal power series K[[t]] we write (1+t)n=∑j≥0cj,ntj. There exists an integer j≥1 such that cj,n=0.
If n>0, then cn,n=1 by the binomial formula. Suppose that n<0 and let m=−n. If cj,n=0 for all j≥1, then (1+t)n=1, whence, taking the inverse, (1+t)m=1, which contradicts the formula cm,m=1.
Lemma 3
Let x1,…,xs be elements of X such that s≥1 and xi=xi+1 for 1≤i≤s−1; let n1,…,ns be non-zero rational integers. Then the element ∏i=1s(1+xi)ni of A^(X) is =1.
Let m be a maximal ideal of K and k the field K/m; let p:A^K(X)→A^k(X) be the unique continuous homomorphism of unital K-algebras such that p(x)=x for x∈X (no. 1, Proposition 1). It suffices to prove that p(∏i=1s(1+xi)ni)=1 and the problem is reduced to the case where K is a field.
By Lemma 2, there exist integers ai>0 such that cai,ni=0 (1≤i≤s). By Algebra, Chapter I, § 7, no. 4, Proposition 6, no monomial x1b1⋯xsbs such that bi≥0 and (b1,…,bs)=(a1,…,as) can be equal to x1a1…xsas. It follows that the coefficient of x1a1…xsas in ∏i=1s(1+xi)ni is ca1,n1…cas,ns=0, which implies the result.
Lemma 4
Let σ be the continuous endomorphism of A^(X) such that σ(x)=x+r(x) for x∈X (no. 1, Proposition 1). Then σ is an automorphism and σ(A^m(X))=A^m(X) for all m∈N.
σ(x)≡xmodA^2(X) for x∈X, whence, for n≥1 and x1,…,xn in X,
σ(x1)…σ(xn)≡x1…xnmodA^n+1(X);
it follows by linearity that σ(a)≡a modulo A^n+1(X) for all a∈An(X) and in particular σ(An(X))⊂A^n(X). It follows that σ(An(X))⊂A^n(X) for m≥n, whence σ(A^n(X))⊂A^n(X). In other words, σ is compatible with the filtration (A^m(X)) on A(X) and its restriction to the associated graded ring is the identity. Hence σ is bijective (Commutative Algebra, Chapter III, § 2, no. 8, Corollary 3 to Theorem 1).
Finally we prove Theorem 1. Let w=1 be an element of F(X). By Algebra, Chapter I, § 7, no. 5, Proposition 7, there exist x1,…,xs in X and non-zero rational integers n1,…,ns such that s≥1, xi=xi+1 (1≤i≤s−1) and
w=x1n1…xsns.
In the notation of Lemma 4,
g(w)=∏(1+σ(xi))ni=σ(∏(1+xi)ni),
hence g(w)=1 by Lemmas 3 and 4.
4. Lower central series of a free group
We shall prove the two following theorems:
Theorem 2
Suppose that in the ring K the relation n.1=0 implies n=0 for every integer n. Let r be a mapping of X into A^(X) such that ω(r(x))≥2 for x∈X and let g be the homomorphism of F(X) into the Magnus group Γ(X) such that g(x)=1+x+r(x) for x∈X. For all n≥1, CnF(X) is the inverse image under g of the subgroup 1+A^n(X) of Γ(X).
Theorem 3
For all x∈X, let c(x) be the canonical image of x in F(X)/(F(X),F(X)). Let g be the graded Lie Z-algebra associated with the filtration (CnF(X))n≥1 of F(X) (§ 4, no. 6). The unique homomorphism of the free Lie Z-algebra LZ(X) into g which extends c is an isomorphism.
Loosely speaking, the graded Lie Z-algebra associated with the free group F(X) (with the lower central series) is the free Lie Z-algebra LZ(X).
We write F(X)=F,Γ(X)=Γ,A^(X)=A^,A^Z(X)=A^Z,CnF(X)=Cn,Γn=1+A^n(X) and let α:LZ(X)→g be the homomorphism introduced in the statement of Theorem 3.
(A) Preliminary reductions.
Let γ denote the homomorphism of F into Γ defined by γ(x)=1+x for x∈X. By Lemma 4 there exists an automorphism σ of the algebra A^ compatible with the filtration on A^ and such that σ(1+x)=g(x) for all x∈X; then σ(Γn)=Γn for all n. As the homomorphisms g and σ∘γ of F into Γ coincide on X,g=σ∘γ and hence g−1(Γn)=γ−1(Γn). Under the hypotheses of Theorem 2, Z can be identified with a subring of K; the Magnus algebra A^Z is therefore identified with a subring of A^ and the filtration on A^Z is induced by that on A^. As γ maps F into A^Z, we see that it suffices to prove Theorems 2 and 3 under the supplementary hypotheses K=Z,r=0 and hence g=γ, hypotheses which we shall henceforth make.
(B) Surjectivity of α.
As X generates the group F=C1, the set c(X) generates the Z-module g1=C1/C2. But g1 generates the Lie Z-algebra g (§ 4, no. 6, Proposition 5) and hence c(X) generates g, which proves that α is surjective.
(C) We identify the graded algebra gr(A^) with A(X) under the canonical isomorphisms An(X)→A^n/A^n+1. For every integer n≥1, we write Fn=γ−1(Γn); we know (§ 4, no. 5) that (Fn)n≥1 is an integral central filtration on F. Let g′ denote the associated graded Lie Z-algebra (§ 4, no. 4). Let f be the Lie algebra homomorphism of g′ into A(X) associated with γ (§ 4, no. 5, Proposition 3). Now Cn⊂Fn for every integer n≥1 (§ 4, no. 6, Proposition 4) and hence there is a canonical homomorphism ε of g=⨁n≥1Cn/Cn+1 into g′=⨁n≥1Fn/Fn+1
LZ(X)αgεg′fA(X).
We write β=f∘ε; we give β explicitly as follows: if u is the class modulo Cn+1 of an element w of Cn, then γ(w)−1 is of order ≥n in A^ and β(u) is the homogeneous component of γ(w)−1 of degree n. In particular,
β(c(x))=xfor all x∈X.
(D) Proof of Theorems 2 and 3.
The Lie algebra homomorphism β∘α:LZ(X)→A(X) restricted to X is the identity by (3) and hence is the canonical injection (§ 3, no. 1). Therefore α is injective and hence bijective by (B); this proves Theorem 3. As β∘α=f∘ε∘α is injective and α is bijective, ε is injective. For all n≥1,
εn:Cn/Cn+1→Fn/Fn+1
is injective and hence
Cn∩Fn+1=Cn+1.
C^1 = F = F^1; if C^n = F^n, then C^n \cap F^{n+1} = F^{n+1} whence C^{n+1} = F^{n+1} which proves Theorem 2 by induction on n \geqslant 1.
Let H be a Hall set relative to X (§ 2, no. 10). Let M be the magma defined by the law of composition (x, y) \mapsto (x, y) = x^{-1}y^{-1}xy on F(X) and let ϕ be the homomorphism of M(X) into M whose restriction to X is the identity. The elements of ϕ(H) are called the basic commutators of F(X) associated with the Hall set H. For every integer n \geqslant 1, let H_n be the subset of H consisting of the elements of length n; we know (§ 2, no. 11, Theorem 1) that the canonical mapping of H_n into L_\mathbf{Z}(X) is a basis of the Abelian group L_\mathbf{Z}^n(X). Moreover, ϕ(Hn)⊂Cn; for all m \in H_n, let ϕn(m) denote the class mod. C^{n+1} of ϕ(m)∈Cn. Theorem 3 then shows that ϕn is a bijection of Hn onto a basis of the Abelian group* C^n/C^{n+1}. It follows immediately that, for all w \in F(X) and all i \geqslant 1, there exists a unique element αi of Z(Hi) such that, for n \geqslant 1,
w=i=1∏nm∈Hi∏ϕ(m)αi(m)mod.Cn+1,
where the product is calculated according to the total ordering given on H.
Example
Suppose that X is a set with two elements x, y and let H_1 = \{x, y\}, H_2 = \{xy\}. Every element w of F(X) can therefore be written
w≡xayb(x,y)cmod.C3with a,b,c in Z.
For w = (xy)^n, a = b = n and c = n(1 - n)/2 (cf. Exercise 9), whence
(xy)n≡xnyn(x,y)n(1−n)/2mod.C3.
5. p-Filtration of Free Groups
In this no., p denotes a prime number and we assume that K = \mathbf{F}_p. Let γ be the homomorphism of F(X) into Γ(X) defined by γ(x)=1+x for x in X; we write Fn(p)(X)=γ−1(1+A^n(X)). The sequence (Fn(p)(X))n⩾1 is an integral central filtration on F(X), which is separated since γ is injective (no. 3, Theorem 1). It is called the p-filtration on F(X).
Proposition 2
Suppose that X is finite. For every integer n \geqslant 1, the group F(X)/F_n^{(p)}(X) is a finite p-group of nilpotency class \leqslant n.
Arguing by induction on n, it suffices to prove that Fn(p)(X)/Fn+1(p)(X) is a finite commutative p-group for all n⩾1. For all w∈Fn(p)(X), the element γ(w)−1 of A^(X) is of order ⩾n; we denote by δn(w) the homogeneous component of γ(w)−1 of degree n. The mapping δn:Fn(p)(X)→An(X) is a homomorphism with kernel Fn+1(p)(X) (§ 4, no. 5, Proposition 3) and hence Fn(p)(X)/Fn+1(p)(X) is isomorphic to a subgroup of An(X). Since X is finite, An(X) is a finite-dimensional vector space over Fp and hence a finite commutative p-group and so is Fn(p)(X)/Fn+1(p)(X).
Proposition 3
For all w=1 in F(X), there exist a finite p-group G and a homomorphism f of F(X) into G such that f(w)=1.
There exist elements x1,…,xr of X and integers n1,…,nr such that w=x1n1…xrnr. Let Y={x1,…,xr}. The canonical injection of Y into X extends to a homomorphism α:F(Y)→F(X); on the other hand, let β be the homomorphism of F(X) into F(Y) whose restriction to Y is the identity and which maps X−Y to {1}. Then β(α(y))=y for y∈Y and hence β∘α is the identity automorphism of F(Y). There obviously exists w′ in F(Y) such that w=α(w′); then β(w)=w′=1; now ⋂n⩾1Fn(p)(Y)={1} and there therefore exists an integer n⩾1 such that β(w)∈/Fn(p)(Y). By Proposition 2, the group G=F(Y)/Fn(p)(Y) is a finite p-group. If f is the composition of β and the canonical homomorphism of F(Y) onto G, then f(w)=1.
Corollary
The intersection of the normal subgroups of finite index in F(X) is {1}.
Exercises
In the following exercises we assume the hypotheses and notation of § 5. F denotes the free group F(X) and g the unique homomorphism of F into the Magnus group Γ(X) such that g(x)=1+x for all x∈X (cf. Theorem 1).