- Home
- Documents
*fialowsk/pubs-af/assoc04-Jan5-2013.pdf¢ COMPLEX ASSOCIATIVE ALGEBRAS ALICE FIALOWSKI AND...*

prev

next

out of 26

View

0Download

0

Embed Size (px)

THE MODULI SPACE OF 4-DIMENSIONAL NON-NILPOTENT

COMPLEX ASSOCIATIVE ALGEBRAS

ALICE FIALOWSKI AND MICHAEL PENKAVA

Abstract. In this paper, we study the moduli space of 4-dimensional complex

associative algebras. We use extensions to compute the moduli space, and then give a decomposition of this moduli space into strata consisting of complex projective orbifolds, glued together through jump deformations. Because the

space of 4-dimensional algebras is large, we only classify the non-nilpotent algebras in this paper.

1. Introduction

The classification of associative algebras was instituted by Benjamin Peirce in the 1870’s [19], who gave a partial classification of the complex associative algebras of dimension up to 6, although in some sense, one can deduce the complete clas- sification from his results, with some additional work. The classification method relied on the following remarkable fact:

Theorem 1.1. Every finite dimensional algebra which is not nilpotent contains a

nontrivial idempotent element.

A nilpotent algebra A is one which satisfies An = 0 for some n, while an idempo- tent element a satisfies a2 = a. This observation of Peirce eventually leads to two important theorems in the classification of finite dimensional associative algebras. Recall that an algebra is said to be simple if it has no nontrivial proper ideals, and it is not the trivial 1-dimensional nilpotent algebra over K which is given by the trivial product.

Theorem 1.2 (Fundamental Theorem of Finite Dimensional Associative Alge- bras). Suppose that A is a finite dimensional algebra over a field K. Then A has a maximal nilpotent ideal N , called its radical. If A is not nilpotent, then A/N is a semisimple algebra, that is, a direct sum of simple algebras.

In fact, in the literature, the definition of a semisimple algebra is often given as one whose radical is trivial, and then it is a theorem that semisimple algebras are direct sums of simple algebras. Moreover, when A/N satisfies a property called separability over K, then A is a semidirect product of its radical and a semisim- ple algebra. Over the complex numbers, every semisimple algebra is separable. To apply this theorem to construct algebras by extension, one uses the following characterization of simple algebras.

Date: January 5, 2013. Research of the first author was partially supported by OTKA grant K77757 and the Humboldt

Foundation, the second author by grants from the University of Wisconsin-Eau Claire.

1

2 FIALOWSKI AND PENKAVA

Theorem 1.3 (Wedderburn). If A is a finite dimensional algebra over K, then A is simple iff A is isomorphic to a tensor product M ⊗D, where M = gl(n,K) and D is a division algebra over K.

One can also say that A is a matrix algebra with coefficients in a division algebra over K. An associative division algebra is a unital associative algebra where every nonzero element has a multiplicative inverse. (One has to modify this definition in the case of graded algebras, but we will not address this issue in this paper.) Over the complex numbers, the only division algebra is C itself, so Wedderburn’s theorem says that the only simple algebras are the matrix algebras. In particular, there is exactly one simple 4-dimensional complex associative algebra, gl(2,C), while there is one additional semisimple algebra, the direct sum of 4 copies of C.

According to our investigations, there are two basic prior approaches to the clas- sification. The first is the old paper by Peirce [19] which attempts to classify all the nilpotent algebras, including nonassociative ones. There are some evident mistakes in that paper, for example, it gives a classification of the commutative nilpotent associative algebras which contains nonassociative algebras as well. The second approach [18] classifies the unital algebras only. It turns out that classification of unital algebras is not sufficient.

Let us consider the unital algebra of one higher dimension which is obtained by adjoining a multiplicative identity as the unital enlargement of the algebra. Two nonisomorphic non-nilpotent algebras can have isomorphic unital enlargements, so they cannot be recovered so easily. Nevertheless, let us suppose that there were some efficient method of constructing all unital algebras of arbitrary dimension, and to determine their maximal nilpotent ideals. In that case, we could recover all nilpotent algebras of dimension n from their enlargements. Moreover, to recover all algebras of dimension n, one would only have to consider extensions of nilpotent algebras of dimension k by semisimple algebras of dimension n−k, where 0 ≤ k ≤ n. Our method turns out to be efficient in constructing extensions of nilpotent algebras by semisimple ones.

Thus, even if the construction of unital algebras could be carried out simply, which is by no means obvious from the literature, one would still need our method- ology to construct most of the algebras. So the role of our paper is to explore the construction method which leads to the description of all algebras.

The main goal of this paper is to give a complete description of the moduli space of nonnilpotent 4-dimensional associative algebras, including a computation of the miniversal deformation of every element. We get the description with the help of extensions, which is the novelty of our approach. The nilpotent cases will be classified in another paper. We also give a canonical stratification of the moduli space into projective orbifolds of a very simple type, so that the strata are connected only by deformations factoring through jump deformations, and the elements of a particular stratum are given by neighborhoods determined by smooth deformations.

The authors thank the referees for their useful comments.

2. Construction of algebras by extensions

In [7], the theory of extensions of an algebra W by an algebra M is described. Consider the exact sequence

0→M → V →W → 0

NON-NILPOTENT 4-DIMENSIONAL ALGEBRAS 3

of associative K-algebras, so that V = M ⊕W as a K-vector space, M is an ideal in the algebra V , and W = V/M is the quotient algebra. Suppose that δ ∈ C2(W ) and µ ∈ C2(M) represent the algebra structures on W and M respectively. We can view µ and δ as elements of C2(V ). Let T k,l be the subspace of T k+l(V ) given recursively by

T 0,0 = K

T k,l =M ⊗ T k−1,l ⊕ V ⊗ T k,l−1

Let Ck,l = Hom(T k,l,M) ⊆ Ck+l(V ). If we denote the algebra structure on V by d, we have

d = δ + µ+ λ+ ψ,

where λ ∈ C1,1 and ψ ∈ C0,2. Note that in this notation, µ ∈ C2,0. Then the condition that d is associative: [d, d] = 0 gives the following relations:

[δ, λ] + 12 [λ, λ] + [µ, ψ] = 0, The Maurer-Cartan equation

[µ, λ] = 0, The compatibility condition

[δ + λ, ψ] = 0, The cocycle condition

Since µ is an algebra structure, [µ, µ] = 0. Then if we define Dµ by Dµ(ϕ) = [µ, ϕ], then D2µ = 0. Thus Dµ is a differential on C(V ). Moreover Dµ : C

k,l → Ck+1,l. Let

Zk,lµ = ker(Dµ : C k,l → Ck+1,l), the (k, l)-cocycles

Bk,lµ = Im(Dµ : C k−1,l → Ck,l), the (k, l)-coboundaries

Hk,lµ = Z k,l µ /B

k,l µ , the Du (k, l)-cohomology

Then the compatibility condition means that λ ∈ Z1,1. If we define Dδ+λ(ϕ) = [δ + λ, ϕ], then it is not true that D2δ+λ = 0, but Dδ+λDµ = −DµDδ+λ, so that

Dδ+λ descends to a map Dδ+λ : H k,l µ → H

k,l+1 µ , whose square is zero, giving rise

to the Dδ+λ-cohomology H k,l µ,δ+λ. Let the pair (λ, ψ) give rise to a codifferential d,

and (λ, ψ′) give rise to another codifferential d′. Then if we express ψ′ = ψ + τ , it is easy to see that [µ, τ ] = 0, and [δ + λ, τ ] = 0, so that the image τ̄ of τ in H0,2µ is

a Dδ+λ-cocycle, and thus τ determines an element {τ̄} ∈ H 0,2 µ,δ+λ.

If β ∈ C0,1, then g = exp(β) : T (V ) → T (V ) is given by g(m,w) = (m + β(w), w). Furthermore g∗ = exp(− adβ) : C(V )→ C(V ) satisfies g

∗(d) = d′, where d′ = δ + µ+ λ′ + ψ′ with λ′ = λ+ [µ, β] and ψ′ = ψ + [δ + λ+ 12 [µ, β], β]. In this case, we say that d and d′ are equivalent extensions in the restricted sense. Such equivalent extensions are also equivalent as codifferentials on T (V ). Note that λ and λ′ differ by a Dµ-coboundary, so λ̄ = λ̄

′ in H1,1µ . If λ satisfies the MC equation

for some ψ, then any element λ′ in λ̄ also gives a solution of the MC equation, for the ψ′ given above. The cohomology classes of those λ for which a solution of the MC equation exists determine distinct restricted equivalence classes of extensions.

Let GM,W = GL(M)×GL(W ) ⊆ GL(V ). If g ∈ GM,W then g ∗ : Ck,l → Ck,l,

and g∗ : Ck(W ) → Ck(W ), so δ′ = g∗(δ) and µ′ = g∗(µ) are codifferentials on T (M) and T (W ) respectively. The group Gδ,µ is the subgroup of GM,W consisting of those elements g such that g∗(δ) = δ and g∗(µ) = µ. Then Gδ,µ acts on the restricted equivalence classes of extensions, giving the equivalence classes of general

4 FIALOWSKI AND PENKAVA

extensions. Also Gδ,µ acts onH k,l µ , and induces an action on the classes λ̄ of λ giving

a solution to the MC equation. Next, consider the group Gδ,µ,λ consisting of the automorphisms h of V of the

form h = g exp(β), where g ∈ Gδ,µ, β ∈ C 0,1 and λ = g∗(λ) + [µ, β]. If d =

δ + µ+ λ+ ψ + τ , then h∗(d) = δ + µ+ λ+ ψ + τ ′ where

τ ′ = g