Simple Lie Algebras Over Fields of Positive Characteristic Volume I Structure Theory Helmut Strade PDF Download
Simple Lie Algebras Over Fields of Positive Characteristic Volume I Structure Theory Helmut Strade PDF Download
https://ebookultra.com/download/simple-lie-algebras-over-fields-
of-positive-characteristic-volume-i-structure-theory-helmut-
strade/
https://ebookultra.com/download/simple-lie-algebras-over-fields-of-
positive-characteristic-i-structure-theory-2nd-ed-edition-helmut-
strade/
https://ebookultra.com/download/central-simple-algebras-and-galois-
cohomology-philippe-gille/
Sheaves of Algebras over Boolean Spaces 1st Edition Arthur
Knoebel (Auth.)
https://ebookultra.com/download/sheaves-of-algebras-over-boolean-
spaces-1st-edition-arthur-knoebel-auth/
https://ebookultra.com/download/expansion-in-finite-simple-groups-of-
lie-type-terence-tao/
https://ebookultra.com/download/the-quantum-theory-of-fields-volume-
iii-supersymmetry-1st-edition-steven-weinberg/
https://ebookultra.com/download/the-quantum-theory-of-fields-volume-
ii-modern-applications-1st-edition-steven-weinberg/
https://ebookultra.com/download/higher-dimensional-geometry-over-
finite-fields-y-tschinkel/
Simple Lie Algebras over Fields of Positive Characteristic
Volume I Structure Theory Helmut Strade Digital Instant
Download
Author(s): Helmut Strade
ISBN(s): 9783110197945, 3110197944
Edition: Reprint 2014
File Details: PDF, 2.75 MB
Year: 2008
Language: english
de Gruyter Expositions in Mathematics 38
Editors
by
Helmut Strade
≥
Walter de Gruyter · Berlin · New York
Author
Helmut Strade
Fachbereich Mathematik
Schwerpunkt Algebra und Zahlentheorie
Universität Hamburg
Bundesstrasse 55
20146 Hamburg, Germany
E-mail:
[email protected]
앪
앝 Printed on acid-free paper which falls within the guidelines
of the ANSI to ensure permanence and durability.
ISBN 3-11-014211-2
쑔 Copyright 2004 by Walter de Gruyter GmbH & Co. KG, 10785 Berlin, Germany.
All rights reserved, including those of translation into foreign languages. No part of this book
may be reproduced or transmitted in any form or by any means, electronic or mechanical,
including photocopy, recording, or any information storage or retrieval system, without permission
in writing from the publisher.
Typesetting using the authors’ TEX files: I. Zimmermann, Freiburg.
Printing and binding: Hubert & Co. GmbH & Co. KG, Göttingen.
Cover design: Thomas Bonnie, Hamburg.
Für meine liebe Renate, die mit bewundernswerter
Introduction 1
Notation 521
Bibliography 527
Index 539
Introduction
The theory of finite dimensional Lie algebras over fields F of positive characteristic p
was initiated by E. Witt, N. Jacobson [Jac37] and H. Zasssenhaus [Zas39]. Sometime
before 1937 E. Witt came up with the following example of a simple Lie algebra of
dimension p (for p > 3), afterwards named the Witt algebra W (1; 1). On the vector
p−2
space i=−1 F ei define the Lie product
(j − i)ei+j if − 1 ≤ i + j ≤ p − 2,
[ei , ej ] :=
0 otherwise.
This algebra behaves completely different from those algebras we know in charac-
teristic 0. As an example, it contains a unique subalgebra of codimension 1, namely
i≥0 F ei . It also has sandwich elements, i.e., elements c = 0 satisfying (ad c) = 0
2
(for example, ep−2 ). E. Witt himself never published this example or generalizations
of it, which he presumably knew of. At that time he was interested in the search for new
finite simple groups. When he realized that these new structures had only known au-
tomorphism groups he apparently lost his interest in these algebras. We have only oral
and indirect information of Witt’s work on this field by two publications of H. Zassen-
haus [Zas39] and Chang Ho Yu [Cha41]. Chang determined the automorphisms and
irreducible representations of W (1; 1) over algebraically closed fields. He also men-
tioned that Witt himself gave a realization of W (1; 1) in terms of truncated polynomial
rings. Namely, W (1; 1) is isomorphic with the vector space F [X]/(X p ) endowed with
the product {f, g} := f d/dx(g) − gd/dx(f ) for all f, g ∈ F [X]/(X p ) under the
mapping ei → x i+1 , where x = X + (Xp ).
In [Jac37] N. Jacobson proved a Galois type theorem for inseparable field exten-
sions by substituting the algebra of derivations for the automorphism group of a field
extension. More explicitly, he was able to show that the set of intermediate fields of a
p
field extension F (c1 , . . . , cn ) : F with ci ∈ F is in bijection with the set of those Lie
subalgebras of Der F F (c1 , . . . , cn ), which are F (c1 , . . . , cn )-modules and are closed
under the p-power mapping D → D p . At that early time Jacobson already introduced
the term “restrictable” for those Lie algebras, which admit a p-mapping x → x [p]
satisfying the equation ad x [p] = (ad x)p for all x. Later he preferred to use the term
“restricted Lie algebra” for pairs (L, [p]), when such a p-mapping is fixed. The Lie
algebras of linear algebraic groups over F are all equipped with a natural p-mapping,
hence they carry canonical restricted Lie algebra structures.
2 Introduction
Such Lie algebras are now called Zassenhaus algebras. He also proved the first
classification result, saying that a simple Lie algebra having a 1-dimensional CSA (=
Cartan subalgebra) such that all roots are GF(p)-dependent and all root spaces are
1-dimensional is isomorphic to sl(2) or W (1; 1).
Since then a great number of publications on this new theory of modular Lie alge-
bras have appeared. We were shown how to construct analogues of the characteristic
0 simple Lie algebras [Jac41], [Jac43], [Che56], [M-S57] (these algebras, including
the exceptional ones, are called classical in the modular theory), and in which way
classes of non-classical algebras (called Cartan type) arise from infinite dimensional
algebras of differential operators over C [K-S66], [K-S69], [Wil69], [Kac74], [Wil76].
In some sense [Wil76] was a cornerstone. In this paper the previously known finite
dimensional simple Lie algebras had been categorized into the classes of classical Lie
algebras and “generalized” Cartan type Lie algebras for characteristic p > 3. People
began to believe that the list of finite dimensional Lie algebras known so far could pos-
sibly be complete, at least for p > 5. There were some indications that characteristic
5 is a borderline case. In fact, additional examples of simple Lie algebras were known
in characteristics 2 and 3 (G. Brown, M. Frank, I. Kaplansky, A. I. Kostrikin) as early
as from 1967. In 1980 G. M. Melikian published a new family of simple Lie algebras
in characteristic 5 ([Mel80]), now named Melikian algebras. The present Classifica-
tion Theory of Block–Wilson–Strade–Premet indeed proves that the classical, Cartan
type, and Melikian algebras exhaust the class of simple Lie algebras for p > 3. It
could also well be that the list of known simple Lie algebras in characteristic 3 is close
to complete. But, as an example, Yu. Kotchetkov and D. Leites [K-L92] constructed
simple Lie algebras in characteristic 2 from superalgebras. This indicates that a greater
variety of constructions could yield many more examples in characteristic 2.
A more complete history of this search for new simple Lie algebras would have
to mention many other mathematicians who prepared the ground well, whose names,
unfortunately, will remain in the dark during this short introduction.
Let us briefly describe the known simple Lie algebras for p > 3. The construction
of C. Chevalley provides in a finite dimensional simple Lie algebra g over C a basis
B of root vectors with respect to a CSA h such that the multiplication coefficients
are integers of absolute value < 5. The Z-span gZ of B is a Z-form in g closed
under taking Lie brackets. Therefore, gF := F ⊗Z gZ is a Lie algebra over F with
basis 1 ⊗ B and structure constants obtained from those for gZ by reducing modulo
p. For p > 3, the Lie algebra gF fails to be simple if and only if the root system
= (g, h) has type Al where l = mp−1 for some m ∈ N. If has type Amp−1 , then
gF ∼= sl(mp) has the one dimensional center of scalar matrices and the Lie algebra
Introduction 3
gF /z(gF ) ∼= psl(mp) is simple. The simple Lie algebras over F thus obtained are
called classical. All classical Lie algebras are restricted with p-mapping given by
(1 ⊗ eα )[p] = 0 and (1 ⊗ hi )[p] = 1 ⊗ hi for all α ∈ and 1 ≤ i ≤ l. As in
characteristic 0, they are parametrized by Dynkin diagrams of types Al , Bl , Cl , Dl ,
G2 , F4 , E6 , E7 , E8 . We stress that, by abuse of characteristic 0 notation, the classical
simple Lie algebras over F include the Lie algebras of simple algebraic F -groups
of exceptional types. All classical simple Lie algebras are closely related to simple
algebraic groups over F .
In [K-S69] A. I. Kostrikin and I. R. Šafarevič gave a unified description of a
large class of non-classical simple Lie algebras over F . Their construction was mo-
tivated by classical work of E. Cartan [Car09] on infinite dimensional, simple, tran-
sitive pseudogroups of transformations. To define finite dimensional modular ana-
logues of complex Cartan type Lie algebras Kostrikin and Šafarevič replaced algebras
of formal power series over C by divided power algebras over F . Let Nm denote
!mof non-negative integers. For α, β ∈ N define
the m
α additive
α(1)monoid
α(m)of all m-tuples
β = β(1) . . . β(m) and α! = i=1 α(i)!. For 1 ≤ i ≤ m set i = (δi1 , . . . , δim )
and 1 = 1 + · · · + m . Give the graded polynomial algebra F [X1 , . . . , Xm ] its
standard coalgebra structure (with each Xi being primitive) and let O(m) denote the
graded dual of F [X1 , . . . , Xm ], a commutative associative algebra over F . It is well-
known (and easily seen) that O(m) has basis {x (α) | α ∈ Nm } and the product in O(m)
is given by
α (α+β)
x (α) x (β) = x for all α, β ∈ Nm .
β
We write xi for x (i ) ∈ O(m), 1 ≤ i ≤ m. For each m-tuple n ∈ Nm we denote by
O(m; n) the F -span of all x (α) with 0 ≤ α(i) < p ni for i ≤ m. This is a subalgebra
of O(m) of dimension p |n| . Note that O(m; 1) is just the commutative algebra with
p
generators x1 , . . . , xm and relations xi = 0 for all i. Hence it is isomorphic to the
p p
truncated polynomial algebra F [X1 , . . . , Xm ]/(X1 , . . . , Xm ). There is another way
of looking at these algebras. Define in the polynomial ring C[X1 , . . . , Xm ] elements
! Xi
α(i)
X(α) := m i=1 α(i)! . Then PZ := α ZX
(α) is a Z-subalgebra of C[X , . . . , X ]
1 m
∼
and O(m) = F ⊗Z PZ under the mapping x → 1 ⊗ X .
(α) (α)
%
m
D(x (α) ) = x (α−i ) D(xi )
i=1
for all α. For 1 ≤ i ≤ m, the i-th partial derivative ∂i of O(m) is defined as the
special derivation of O(m) with the property that ∂i (x (α) ) = x (α−i ) if α(i) > 0 and
0 otherwise. Each finite dimensional subalgebra O(m; n) is stable under the partial
derivatives ∂1 , . . . , ∂m . Let W (m) denote the space of all special derivations of O(m).
Since each D ∈ W (m) is uniquely determined by its values D(x1 ), . . . , D(xm ), the
Lie algebra W (m) is a free O(m)-module with basis ∂1 , . . . , ∂m .
4 Introduction
The Cartan type Lie algebra W (m; n) is the O(m; n)-submodule of W (m) gen-
erated by the partial derivatives ∂1 , . . . , ∂m . This Lie algebra is canonically embed-
ded into Der O(m; n). If n = 1, it is isomorphic to the full derivation algebra of
p p
F [X1 , . . . , Xm ]/(X1 , . . . , Xm ), the truncated polynomial ring in m variables. Thus
this family generalizes the p-dimensional Witt algebra.
Give the O(m)-module
!1 (m) := HomO(m) (W (m), O(m))
the canonical W (m)-module structure by setting (Dα)(D ) := D(α(D ))−α([D, D ])
for all D, D ∈ W (m) and α ∈ !1 (m), and define d : O(m) −→ !1 (m) by the rule
(df )(D) = D(f ) for all D ∈ W (m) and f ∈ O(m). Notice that d is a homomorphism
of W (m)-modules and !1 (m) is a free O(m)-module with basis dx1 , . . . , dxm . Let
!(m) = !k (m)
0≤k≤m
be the exterior algebra over O(m) on !1 (m). Then !0 (m) = O(m) and each graded
component !k (m), k ≥ 1, is a free O(m)-module with basis (dxi1 ∧ · · · ∧ dxik | 1 ≤
i1 < · · · < ik ≤ m). The elements of !(m) are called differential forms on O(m).
The map d extends uniquely to a zero-square linear operator of degree 1 on !(m)
satisfying
d(f ω) = (df ) ∧ ω + f d(ω), d(ω1 ∧ ω2 ) = d(ω1 ) ∧ ω2 + (−1)deg(ω1 ) ω1 ∧ d(ω2 )
for all f ∈ O(m) and all ω, ω1 , ω2 ∈ !(m). Each D ∈ W (m) extends to a derivation
of the F -algebra !(m) commuting with d. As in the characteristic 0 case, the three
differential forms below are of particular interest:
ωS := dx1 ∧ · · · ∧ dxm , m ≥ 3,
%
r
ωH := dxi ∧ dxi+r , m = 2r ≥ 2,
i=1
%
r
ωK := dx2r+1 + (xi dxi+r − xi+r dxi ), m = 2r + 1 ≥ 3.
i=1
Each X(m; n) := X(m) ∩ W (m; n) is a graded Lie subalgebra of W (m), viewed with
its grading given by deg(xi ) = 1 for all i if X = W , S, CS, H , CH , and deg(xi ) = 1
(i = 2r + 1), deg(x2r+1 ) = 2 in case X = K.
Suppose p ≥ 3. It is shown in [K-S69] that the Lie algebras S(m; n)(1) , H (m; n)(1)
and K(m; n)(1) are simple for m ≥ 3 and that so is H (2; n)(2) . Moreover, K(m; n) =
K(m; n)(1) unless p|(m + 3). Any graded Lie subalgebra of X(m; n) containing
X(m; n)(∞) for some X ∈ {W, S, CS, H, CH, K} is called a finite dimensional
graded Cartan type Lie algebra, and any filtered deformation of a graded Cartan
type Lie algebra is called a Cartan type Lie algebra.
In characteristic 5 the additional family of Melikian algebras M(n1 , n2 ) occurs.
Set n = (n1 , n2 ) ∈ N2 , let W(2; n) denote a copy of W (2; n), and endow the vector
space
M(n1 , n2 ) := O(2; n) ⊕ W (2; n) ⊕ W
(2; n)
[D, Ẽ] = [D, E] + 2 div(D)Ẽ
[D, f ] = D(f ) − 2 div(D)f
[f1 ∂˜1 + f2 ∂˜2 , g1 ∂˜1 + g2 ∂˜2 ] = f1 g2 − f2 g1
[f, Ẽ] = f E
[f, g] = 2 g∂2 (f ) − f ∂2 (g) ∂˜1 + 2 f ∂1 (g) − g∂1 (f ) ∂˜2
for all D, E ∈ W (2; n), f, g ∈ O(2; n). M(n1 , n2 ) is a Z-graded Lie algebra by
setting
h1 = 2x1 ∂1 , h2 = x1 ∂1 − x2 ∂2 ,
eα2 = x1 ∂2 , e−α2 = x2 ∂1 ,
eα1 = ∂˜1 , eα1 +α2 = ∂˜2 ,
e2α1 +α2 = 1/2,
e3α1 +α2 = ∂1 , e3α1 +2α2 = ∂2
gives an isomorphism of the local algebras i≤0 G2,[i] and i≤0 M(n1 , n2 )[i] .
About 30 years after the first appearance of non-classical Lie algebrasA. I. Kostrikin
and I. R. Šafarevič [K-S66] conjectured that every simple restricted Lie algebra over
an algebraically closed field of characteristic p > 5 is of classical or Cartan type.
An early step towards the Classification had been undertaken by W. H. Mills and
G. B. Seligman [M-S57], who characterized the classical algebras by internal prop-
erties in characteristic > 3. They showed that, if a simple Lie algebra has an abelian
CSA and a root space decomposition with respect to this CSA with the properties we
are familiar with in characteristic 0, then these algebras are classical.
Note, however, that in the characteristic p situation most of the classical methods
fail to work. Generally speaking, no Killing form is available, Lie’s theorem on
solvable Lie algebras is not true, semisimplicity of an algebra does not imply complete
reducibility of its modules, CSAs in simple algebras need neither be abelian nor have
equal dimension, root lattices with respect to a CSA may be full vector spaces over
the prime field. The occurrence of the Cartan type Lie algebras indicates that filtration
methods should by very useful. In another Recognition Theorem, A. I. Kostrikin and
I. R. Šafarevič[K-S69] and V. Kac [Kac70] proved that a simple graded Lie algebra is
of Cartan type, if its gradation has some rather special properties. In particular, it is
required that the 0-component L0 is close to classical.
The Kostrikin–Šafarevič conjecture has been proved for p > 7 by R. E. Block
and R. L. Wilson [B-W88]. Since the known classical methods no longer work in the
modular case, people had to develop a variety of new techniques. Unfortunately, these
techniques often rely on complex detailed arguments and subtle computations. The
most basic idea is to choose a suitable toral subalgebra T in the simple restricted Lie
algebra L (this choice has to be done in a very sophisticated manner), and to determine
the structure of 1-sections i∈GF(p) Liα (T ) and 2-sections i,j ∈GF(p) Liα+jβ (T ).
The investigation of the 2-sections covers the hardest part of the Block–Wilson work.
From the knowledge obtained this way they construct a filtration on L, and deduce
that either the Mills–Seligman axioms or the Recognition Theorem applies for gr L.
In the first case L is classical, in the second L is classical or a filtered deformation of
a graded Cartan type Lie algebra, hence is a Cartan type Lie algebra.
The generalization of the Kostrikin–Šafarevič conjecture for the general case of
not necessarily restricted Lie algebras and p > 7 has been proved by the author (partly
in conjunction with R. L. Wilson) in a series of papers, the result has been announced
Introduction 7
in [S-W91]. In order to achieve this result one embeds the simple Lie algebra L into
a restricted semisimple Lie algebra L[p] , and proves that the essential parts of the
Block–Wilson results on the 2-sections remain valid. The last step of constructing
the filtration and recognizing the algebra, which in the restricted case had been rather
easy compared with the work on the determination of the 2-sections, is incomparably
more complicated in the general case.
About 30 years after the first definition of a non-classical Lie algebra by E. Witt,
the conjecture of A. I. Kostrikin and I. R. Šafarevič had been stated. After another 35
years A. A. Premet and the author have settled the remaining case of the Kostrikin–
Šafarevič conjecture, the case p = 7. Moreover, they completed the classification for
p > 3. The result is the following
Classification Theorem. Every simple finite dimensional Lie algebra over an al-
gebraically closed field of characteristic p > 3 is of classical, Cartan, or Melikian
type.
The strategy of a proof for the small characteristics p = 7, 5 is the same as
before, however because of the small characteristic, is even more subtle. There is
some promising progress for characteristic 3 due to M. Kuznetsov and S. Skryabin,
but in my opinion the classification of the simple Lie algebras in characteristic 2 is far
beyond the range of the presently known methods.
Let us give an outline of the major steps of this classification work. In principle
one
proceeds as in the classical case. Start with a root space decomposition L = H ⊕ Lα
with respect to a CSA H . There is, in general, no Jordan–Chevalley decomposition
of elements available. But this decomposition is a very important tool. In order to
obtain that, one needs to consider p-envelopes. There is an injective homomorphism
ad : L → L ⊂ Der L,
where L is the subalgebra generated by ad L and associative p-th powers. L is a
restricted Lie algebra (a p-envelope of L), but it is no longer simple.
Next one takes a toral subalgebra T ⊂ L of maximal dimension. As in the classical
case one determines the structure of 1-sections and 2-sections with respect to T ,
% %
L(α) = Liα (T ), L(α, β) = Liα+jβ (T ),
i∈GF(p) i,j ∈GF(p)
and puts this information together. In the classical case this procedure already yields
the list of Dynkin diagrams. In characteristic p things are much more involved. To
begin with, even a simple restricted Lie algebra might contain maximal toral subalge-
bras of various dimensions. Even worse, not all tori of maximal dimension are good
for our purpose, as we shall see below. So define the absolute toral rank TR(L) of a
simple Lie algebra L to be the maximum of the dimensions of toral subalgebras in L.
This concept has to be generalized to all finite dimensional Lie algebras. One proves
that k-sections with respect to a toral subalgebra of maximal dimension have absolute
toral rank ≤ k.
8 Introduction
The next obstruction we face is the fact that Lie’s theorem on solvable Lie alge-
bras does no longer hold. However, various substitutes for particular cases have been
proved. Historically, every new result on this problem finally allowed an extension
of the Classification. As examples, R. L. Wilson [Wil77] proved that CSAs act trig-
onalizably on L (provided L is simple and p > 7). This was one major item for
Block and Wilson to achieve their classification result. The present author extended
this result to CSAs of p-envelopes of simple Lie algebras, which are the 0-space for
toral subalgebras of maximal dimension [Str89/2]. This result allowed one to apply
the Block–Wilson classification of semisimple restricted Lie algebras of absolute toral
rank 2 to 2-sections of p-envelopes of simple Lie algebras with respect to toral subal-
gebras of maximal dimension, and so became the starting point for the classification
of not necessarily restricted simple Lie algebras (p > 7). Finally, A. A. Premet clar-
ified the situation for p = 5, 7 and showed that the Melikian algebras are the only
exceptions to this trigonalizability theorem [Pre94]. This result encouraged us to start
the classification for p = 5, 7.
The semisimple quotient of a 1-section L(α)/ rad L(α) with respect to a toral
subalgebra of maximal dimension in L has absolute toral rank at most 1, and from
this one concludes that it is (0), or contains a unique minimal ideal S which has
absolute toral rank 1. If L(α) is solvable, then due to Wilson, Premet, Strade, L(α)(1)
acts nilpotently on L (which is another important substitute for Lie’s theorem). In the
other case, S is simple containing a CSA, for which the root lattice is spanned by a
single root. At least S is then known by a result of Wilson [Wil78] and its extension
to p > 3 by Premet [Pre86].
Next, consider the T -semisimple quotients of 2-sections L(α, β)/ radT L(α, β)
with respect to a toral subalgebra T of maximal
dimension in L. The T -socle of
this algebra is defined to be the direct sum Si of all its minimal T -invariant ideals.
These algebras Si are either simple or, due to Block’s theorem (see below) of the form
S̃i ⊗F O(m; 1), where S̃i is a simple Lie algebra. One can prove that the simple
ingredients of the socle have absolute toral rank ≤ 2. This result implies that one has
to classify the simple Lie algebras M with TR(M) = 2 in order to obtain the necessary
information on the 2-sections. I shall now indicate some principles of a proof for this
case in the work of Premet–Strade.
(A) Choose a T -invariant filtration of M,
At first one has to decide if such a filtration exists for which M(1) = (0). To attack
that problem we construct T -sandwiches, i.e., elements c ∈ M satisfying
One first decides on the existence of an element satisfying (ad x)3 = 0, which is
difficult only in the case p = 5. Then one uses Jordan algebra theory to construct
sandwiches. The result is the following.
Introduction 9
Theorem ([P-S97]). Let M be a simple Lie algebra of absolute toral rank 2 over an
algebraically closed field of characteristic p > 3. Then M is either classical or of
Cartan type H (2; 1; (τ ))(1) , or there is a 2-dimensional toral subalgebra T in the
semisimple p-envelope of M such that M contains T -sandwiches.
Here H (2; 1; (τ ))(1) is a filtered deformation of a graded Hamiltonian algebra.
Every T -sandwich c gives rise to a filtration of the required form, namely let M(0) be
a maximal T -invariant subalgebra of M containing ker(ad c). Then [M, c] ⊂ M(0)
and c ∈ M(1) hold. Here is the place to make a comment on the toral subalgebra. In
W (1; 1) = Der O(1; 1) the “good” toral subalgebra F x∂ respects the natural filtration.
There are F x∂-sandwiches. The toral subalgebra F (1 + x)∂ does not respect the
natural filtration and in fact there are no F (1 + x)∂-sandwiches. One would like to
start with a toral subalgebra, which behaves “well” simultaneously in all 1-sections,
but it is not clear at the beginning whether there are “globally well behaving” toral
subalgebras.
(B) One now has to make very technical choices of T and M(0) . By the above
theorem we may assume that M(1) = (0). Put G := gr M, let M(G) be the maximal
ideal of G in i<0 Gi and set Ḡ := G/M(G). By a result of Weisfeiler [Wei78], Ḡ is
semisimple and has a unique minimal ideal A(Ḡ). This is a graded ideal. In this step
it is our goal to gain information on this minimal ideal and then lift this information
to determine M.
Thus let us look at semisimple Lie algebras. In characteristic p, semisimple
algebras are not necessarily direct sums of simple algebras.
Theorem ([Blo68/1]). Let I be a minimal ideal in a semisimple Lie algebra L. Then
there are m ≥ 0 and a simple Lie algebra S such that
Suppose that S is in the list of the Classification Theorem. It follows from the prop-
erties of M/M(0) and the representation theory of these simple Lie algebras that
M(G) = (0). Then Ḡ = G = gr M ⊂ Der S and hence M is a filtered deformation
of G. This means that there is a Lie algebra Q over the polynomial ring F [t], such
that
Q/(t − λ)Q ∼ = M if λ = 0, Q/tQ ∼ = G ⊃ S.
Since M contains sandwiches, so does G. Then S cannot be classical.
Suppose S is Melikian. Then S has a CSA H , for which H (1) acts non-nilpotently.
Hence G does so, and M does so as well. But then M is a Melikian algebra, since this
is the only algebra having such a CSA.
Suppose S is of Cartan type. Recall that G is Z-graded. This grading defines a
1-dimensional algebraic torus in Aut S. By classical theory of linear algebraic groups
this torus can be mapped under conjugation into a naturally given maximal torus. As
a result, one associates a degree with the generators x1 , . . . , xm , this defines a grading
of S, and then G is obtained up to isomorphisms as a graded subalgebra of Der S. Due
to the technical choice of M(0) the only possible grading is the natural grading. Then
M is of Cartan type.
(D) So we are left with the case that S is a counterexample to the Classification
Theorem. One repeats applying the gr-operator for good choices of the toral subalgebra
T and the maximal subalgebra M(0) , and by this one obtains very strong information
on the number and dimensions of the T -weight spaces of i<0 Si .
The most difficult task now is to describe the extension
0 → rad S0 → S0 → S0 / rad S0 → 0.
Applying the theory of Cartan prolongation, Skryabin [Skr97] proves that one of the
following cases occurs.
= (0),
rad S0 = C(S0 ) is 1-dimensional,
= C(S0 ) is abelian.
In the first case one applies the method mentioned in (B) of determining semi-
simple Lie algebras, in the second case one concludes that (S0 / rad S0 )(2) ∈
{sl(2), W (1; 1), H (2; 1)(2) }. The central extensions of these algebras are known.
In the third case one proves that S−1 is a coinduced S0 -module, and similar to the
method mentioned in (B) there is a simultaneous realization
S−1 ∼
= V ⊗ O(m; 1), m > 0
S0 → gl(V ) ⊗ O(m; 1) + IdV ⊗W (m; 1).
The theory of Cartan prolongation then yields that in this case (with π2 the projection
onto W (m; 1))
Introduction 11
– π2 (S0 ) is O(m; 1)-invariant, whence π2 (S0 ) = (0) or π2 (S0 ) = W (m; 1), and
– the extension splits.
This then gives the required list for S0 . Our detailed knowledge on T -weights of
i<0 Si in combination with the representation theory of S0 finally yields that S0 is
abelian. But then S is classical or of Witt type. Hence there is no graded counterex-
ample. This proves the theorem for the case TR(M) = 2.
Now return to the general case. By the former investigations the simple ingredients
S̃i of the T -semisimple quotients of T -2-sections of any simple Lie algebra L are
known. This information then provides a list of 2-sections.
In a next step one determines some of the simple Lie algebras M with TR(M) = 3.
Suppose all 1-sections of M with respect to a torus of maximal dimension are
solvable. Then there is a 2-section having a semisimple quotient isomorphic to
H (2; 1; (τ ))(1) . The representation theory of this algebra yields information, which
allows to determine the multiplication of M.
Suppose all 1-sections are solvable or classical, and there are 1-sections of either
type. Then one shows that there is a 2-section of type H (2; 1; (τ ))(1) . The represen-
tation theory of this latter algebra yields a contradiction. So this case is impossible.
Suppose all non-zero 1-sections are classical. Then one rather easily shows that
M is generated by elements x satisfying (ad x)3 = 0. For such elements exp(ad x) is
an automorphism of M (since p ≥ 5). The theory of linear algebraic groups shows
that the connected component of Aut M is simple and the Lie algebra of this group is
M. Then M is classical.
Suppose M has a 2-section of Melikian type. Then M contains non-trigonalizable
CSAs, and from this one deduces that the 2-sections have semisimple quotients of
Melikian type or type H (2; (1, 2)) only.
Now one puts all the information together. Start with an arbitrary toral subalgebra
T in L of maximal dimension. The k-sections (k = 1, 2) are known. Every 1-section
is solvable, or classical, or has a distinguished subalgebra of maximal dimension.
There is a procedure (D. Winter’s toral switching) to pass to another toral subalgebra
T of maximal dimension which stabilizes all these distinguished subalgebras of T -
1-sections (that requires arguing in T -2-sections as well). Looking at 2-sections one
realizes that the sum of all distinguished subalgebras is a T -invariant subalgebra L(0)
of L, which gives rise to a filtration satisfying L(1) = (0). The partial knowledge
on the 3-sections gives information on L(0) , and this information allows to apply the
Mills–Seligman theorem to L(0) /L(1) . Thus one knows gr 0 L, and this enables one to
apply the Recognition Theorem to gr L. Once knowing gr L one can determine L.
So far we mentioned only those authors who announced and proved the respective
final classification results. It should be said that, of course, many other mathematicians
have contributed to these results. Thus a large number of publications has to be read
if someone tries to follow the classification from the beginning to the end. It is rather
hard to do so. Some of these publications are even not easy to accede, most of them
use very specific techniques and rather detailed computations. People often realized
12 Introduction
that, in order to achieve a next result, they had to modify some older concepts and
terminology. It was inevitable that some attempts led into a fruitless direction and
some became superfluous.
This two-volume work will include a complete presentation of the classification
of the simple Lie algebras over an algebraically closed field of characteristic p > 3,
in the sense that a list of simple Lie algebras will be presented, and a proof will be
given that this list is complete. I have included all definitions and almost all proofs.
The prospective reader is supposed to be familiar with major parts of [S-F88]. In fact,
I shall often use [S-F88] as a reference, even for results which originally have been
proved elsewhere. Besides that only some very fundamental results like the Mills–
Seligman characterization of classical algebras, Kac’ Recognition Theorem, and some
basic results on linear algebraic groups will be included without giving proofs.
The original Classification Theorem does not say anything about isomorphism
classes. The present monograph will also include the solution of this isomorphism
problem, as is given in a variety of publications of several authors. There are no
isomorphisms between algebras of the different types of classical, Cartan, and Me-
likian algebras (p > 3). Among the classical algebras there are only the natural
isomorphisms. The Witt and Contact algebras are weakly rigid, this meaning that no
non-trivial filtered deformation of naturally graded Witt or Contact algebras exist. The
isomorphism classes of Witt, Special, and Contact algebras are determined, and so are
those of the Melikian algebras. The isomorphism classes of Hamiltonian algebras are
ruled by the orbits of Hamiltonian differential forms under a subgroup of automor-
phisms. Determining these has been accomplished by Skryabin. It was a challenging
task, and its complete presentation lies beyond the scope of this book. So we include
the result but only part of its proof. We shall use in the Classification Theory only
those parts which are proved in this monograph.
Finally, a list of all presently known simple Lie algebras over algebraically closed
fields of characteristic 3 is included.
The main classification work will be presented in Volume 2, while Volume 1
contains methods and results which are of general interest. More detailed, Volume 1
contains the following.
Chapter 1. The basic concepts of a p-envelope and the absolute toral rank of
an arbitrary Lie algebra are introduced. The universal p-envelope of L is the Lie
subalgebra L̂ of U (L) spanned by L and iterated associative p-th powers. Every
homomorphic image L̂/C with C ⊂ C(L̂), C ∩ L = (0), is called a p-envelope
of L. The absolute toral rank TR(L) of a finite dimensional Lie algebra L is the
maximum of dimensions of toral subalgebras of L̂/C(L̂). Note that in contrast to the
characteristic 0 theory CSAs of simple Lie algebras over algebraically closed fields
of positive characteristic need not be toral subalgebras, but may contain ad-nilpotent
elements. The absolute toral rank substitutes the concept of the rank of a simple Lie
algebra in characteristic 0, and thus is an important measure of the size of a Lie algebra.
Several results on the absolute toral rank of subalgebras and homomorphic images are
proved. In particular, TR(L) ≥ TR(gr L) holds for filtered algebras. Finally, we
Other documents randomly have
different content
of the Church. Although he himself was a Catholic, he wished the
Protestant Church to have a stronger position in Germany, since
Germany was two-thirds Protestant.
The Protestant Church, however, was divided into provincial
churches, and there were various small differences which the
dogmatists took very seriously. For that reason they once in the
past, as we know, fought each other for 30 years; but these
differences did not seem so important to us. There were the
Reformed, the United, and the pure Lutherans—I myself am not an
expert in this field.
Constitutionally, as Prussian Prime Minister, I was, to be sure, in a
certain sense the highest dignitary of the Prussian Church, but I did
not concern myself with these matters very much.
The Führer wanted to achieve the unification of the Protestant
Evangelical Churches by appointing a Reich Bishop, so that there
would be a high Protestant church dignitary as well as a high
Catholic church dignitary. To begin with, he left the choice to the
Evangelical churches, but they could not come to an agreement.
Finally they brought forward one name, exactly the one which was
not acceptable to us. Then a man was made Reich Bishop who had
the Führer’s confidence to a higher degree than any of the other
provincial bishops.
With the Catholic Church the Führer ordered a concordat to be
concluded by Herr Von Papen. Shortly before that agreement was
concluded by Herr Von Papen I visited the Pope myself. I had
numerous connections with the higher Catholic clergy because of my
Catholic mother, and thus—I am myself a Protestant—I had a view
of both camps.
One thing, of course, the Führer and all of us, I, too, stood for
was to remove politics from the Church as far as was possible. I did
not consider it right, I must frankly say, that on one day the priest in
church should humbly concern himself with the spiritual welfare of
his flock and then on the following day make a more or less
belligerent speech in parliament.
A separation was planned by us, that is to say, the clergy were to
concentrate on their own sphere and refrain from becoming involved
in political matters. Owing to the fact that we had in Germany
political parties with strong church leanings, considerable confusion
had arisen here. That is the explanation of the fact that, because of
this political opposition that at first played its role in the political field
in parliament, and in election campaigns, there arose among certain
of our people an antagonistic attitude toward the Church. For one
must not forget that such election disputes and speeches often took
place before the electors between political representatives of our
Party and clergymen who represented those political parties which
were more closely bound to the Church.
Because of this situation and a certain animosity, it is
understandable that a more rabid faction—if I may use that
expression in this connection—did not forget these contentions and
now, on its side, carried the struggle on again on a false level. But
the Führer’s attitude was that the churches should be given the
chance to exist and develop. In a movement and a party which
gradually had absorbed more or less the greater part of the German
nation, and which now in its active political aspect had also absorbed
the politically active persons of Germany, it is only natural that not
all the members would be of the same opinion in every respect,
despite the Leadership Principle. The tempo, the method, the
attitude may be different; and in such large movements, even if they
are ever so authoritatively led, certain groups form in response to
certain problems. And if I were to name the group which still saw in
the Church, if not a political danger, at least an undesirable
institution, then I should mention above all two personages:
Himmler on one side and Bormann—particularly later on much more
radically than Himmler—on the other side.
Himmler’s motives were less of a political and more of a confused
mystical nature. Bormann’s aims were much more clear-cut. It was
clear, too, that from the large group of Gauleiter, one or another
might be more keenly interested in this fight against the Church.
Thus, there were a number of Gaue where everything was in the
best of order as far as the Church was concerned, and there were a
few others where there was a keen fight against the Church.
I did interfere personally on frequent occasions. First of all, in
order to demonstrate my attitude and to create order, I called into
the Prussian State Council, as men in whom I had special
confidence, a high Protestant and a high Catholic clergyman.
I myself am not what you might call a churchgoer, but I have
gone now and then, and have always considered I belonged to the
Church and have always had those functions over which the Church
presides—marriage, christening, burial, et cetera—carried out in my
house by the Church.
My intention thereby was to show those weak-willed persons
who, in the midst of this fight of opinions did not know what they
should do, that, if the second man in the State goes to church, is
married by the Church, has his child christened and confirmed, et
cetera, then they can calmly do the same. From the number of
letters which I received as the result, I can see that I did the right
thing.
But as time went by, in other spheres as well as this, the
situation became more critical. During the early years of the war I
spoke to the Führer about it once more and told him that the main
concern now was, that every German should do his duty and that
every soldier should go to his death, if need be, bravely. If in that
connection his religious belief is a help and a support to him,
whether he belongs to this or that confession, it can be only an
advantage, and any disturbance in this connection could conceivably
affect the soldier’s inward strength. The Führer agreed absolutely. In
the Air Force I deliberately had no chaplains, because I was of the
opinion that every member of the Air Force should go to the
clergyman in whom he had the most confidence.
This was repeatedly told to the soldiers and officers at roll call.
But to the Church itself I said that it would be good if we had a clear
separation. Men should pray in church and not drill there; in the
barracks men should drill and not pray. In that manner, from the
very beginning, I kept the Air Force free from any religious
disturbances and I insured complete liberty of conscience for
everyone.
The situation became rapidly more critical—and I cannot really
give the reasons for this—especially in the last 2 or 3 years of the
war. It may have something to do with the fact that in some of the
occupied territories, particularly in the Polish territory and also in the
Czech territory, the clergy were strong representatives of national
feeling and this led again to clashes on a political level which were
then naturally carried over to religious fields. I do not know whether
this was one of the reasons, but I consider it probable. On the whole
I should like to say that the Führer himself was not opposed to the
Church. In fact, he told me on one occasion that there are certain
things in respect to which even as Führer one cannot entirely have
one’s way if they are still undecided and in need of reform, and that
he believed that at the time much was being thought and said about
the reorganization of the Church. He said that he did not consider
himself destined to be a reformer of the Church and that he did not
wish that any of his political leaders should win laurels in this field.
DR. STAHMER: Now, in the course of years, a large number of
clergy, both from Germany and especially from the occupied
territories—you yourself mentioned Poland and Czechoslovakia—
were taken to concentration camps. Did you know anything about
that?
GÖRING: I knew that at first in Germany a number of clergymen
were taken to concentration camps. The case of Niemöller was
common knowledge. I do not want to go into it in detail, because it
is well known. A number of other clergymen were sent to
concentration camps but not until the later years when the fight
became more critical, for they made political speeches in the pulpit
and criticized measures of the State or the Party; then, according to
the severity of this criticism, the police intervened.
I told Himmler on one occasion that I did not think it was wise to
arrest clergymen. As long as they talked in church they should say
what they wanted, but if they made political speeches outside their
churches then he could proceed against them, just as he would in
connection with any other people who made speeches hostile to the
State. Several clergymen who went very far in their criticism were
not arrested. As far as the arrest of clergy from occupied territories
is concerned, I heard about it; and I said earlier that this did not
occur so much on the religious level just because they were
clergymen, but because they were at the same time nationalists—I
understand that from their point of view—and consequently often
involved in actions hostile to the occupying forces.
DR. STAHMER: The Party program included two points, I believe,
dealing with the question of the Jews. What was your basic attitude
towards this question?
GÖRING: This question, which has been so strongly emphasized
in the Indictment, forces me under all circumstances to interpose
certain statements.
After Germany’s collapse in 1918 Jewry became very powerful in
Germany in all spheres of life, especially in the political, general
intellectual and cultural, and, most particularly, the economic
spheres. The men came back from the front, had nothing to look
forward to, and found a large number of Jews who had come in
during the war from Poland and the East, holding positions,
particularly economic positions. It is known that, under the influence
of the war and business concerned with it—demobilization, which
offered great possibilities for doing business, inflation, deflation—
enormous shifts and transfers took place in the propertied classes.
There were many Jews who did not show the necessary restraint
and who stood out more and more in public life, so that they
actually invited certain comparisons because of their numbers and
the position they controlled in contrast to the German people. In
addition there was the fact that particularly those parties which were
avoided by nationally minded people also had Jewish leadership out
of proportion to the total number of Jews.
That did not apply only to Germany, but also to Austria, which we
have always considered a part of Germany. There the entire Social
Democratic leadership was almost exclusively in Jewish hands. They
played a very considerable part in politics, particularly in the left-
wing parties, and they also became very prominent in the press in all
political directions.
At that time, there thus ensued a continuous uninterrupted
attack on everything national, national concepts and national ideals.
I draw attention to all the magazines and articles which dragged
through the mud things which were holy to us. I likewise call
attention to the distortion which was practiced in the field of art in
this direction, to plays which dragged the fighting at the front
through the mud and befouled the ideal of the brave soldier. In fact I
could submit an enormous pile of such articles, books, plays, and so
forth; but this would lead too far afield and I am actually not too
well informed on the subject. Because of all this, a defense
movement arose which was by no means created by National
Socialism but which had existed before, which was already strong
during the war and which came even more strongly to the fore after
the war, when the influence of Jewry had such effects.
Moreover, in the cultural and intellectual sphere also many things
which were not in accordance with German feeling came to be
expressed. Here, too, there was a great split. In addition there was
the fact that in economic matters, if one overlooks the western
industry, there was an almost exclusive domination on the part of
Jewry, which, indeed, consisted of elements which were most
sharply opposed by the old, established Jewish families.
When the movement then drew up its program, which was done
by a few simple people—as far as I know, not even Adolf Hitler
himself took part in the drafting of the program, at least not yet as a
leader—the program included that point which played a prominent
part as a defensive point among large sections of the German
people. Shortly before that there had been the Räte-Republik in
Munich and the murder of hostages, and here, too the leaders were
mostly Jews. It can be understood, therefore, that a program drawn
up in Munich by simple people quite naturally took this up as a
defense point. News also came of a Räte-Republik in Hungary—
again consisting mainly of Jews. All this had made a very strong
impression. When the program became known, the Party—which
was at that time extremely small—was at first not taken seriously
and was laughed at. But then, from the very beginning, a
concentrated and most bitter attack on the part of the entire Jewish
press, or the Jewish-influenced press, was started against the
movement. Everywhere Jewry was in the lead in the fight against
National Socialism, whether in the press, in politics, in cultural life by
making National Socialism contemptible and ridiculous, or in the
economic sphere. Whoever was a National Socialist could not get a
position; the National Socialist businessman could not get supplies or
space for advertisements, and so on. All this naturally resulted in a
strong defensive attitude on the part of the Party and led from the
very beginning to an intensification of the fight, such as had not
originally been the intention of the program. For the program aimed
very definitely at one thing above all—that Germany should be led
by Germans. And it was desired that the leadership, especially the
political shaping of the fate of the German people, should be in the
hands of German persons who could raise up the spirit of the
German people again in a way that people of a different kind could
not. Therefore the main point was at first merely to exclude Jewry
from politics, from the leadership of the State. Later on, the cultural
field was also included because of the very strong fight which had
developed, particularly in this sphere, between Jewry on the one
side and National Socialism on the other.
I believe that if, in this connection, many a hard word which was
said by us against Jews and Jewry were to be brought up, I should
still be in a position to produce magazines, books, newspapers, and
speeches in which the expressions and insults coming from the other
side were far in excess. All that obviously was bound to lead to an
intensification.
Shortly after the seizure of power countless exceptions were
made. Jews who had taken part in the World War and who had been
decorated were treated differently and shown consideration; they
remained unaffected by measures excluding Jews from civil services.
As I have said, the chief aim was to exclude them from the
political sphere, then from the cultural sphere.
The Nuremberg Laws were intended to bring about a clear
separation of races and, in particular, to do away with the notion of
persons of mixed blood in the future, as the term of half Jew or
quarter Jew led to continuous distinctions and confusion as far as
their position was concerned. Here I wish to emphasize that I
personally had frequent discussions with the Führer regarding
persons of mixed blood and that I pointed out to the Führer that,
once German Jews were clearly separated, it was impossible to have
still another category between the two which constituted an
unclarified section of the German people, which did not stand on the
same level as the other Germans. I suggested to him that, as a
generous act, he should do away with the concept of the person of
mixed blood and place such people on the same footing as the other
Germans. The Führer took up this idea with great interest and was
all for adopting my point of view, in fact, he gave certain preparatory
orders. Then came more troubled times, as far as foreign policy was
concerned—the Sudeten crisis, Czechoslovakia, the occupation of the
Rhineland, and afterward the Polish crisis—and the question of
persons of mixed blood stepped into the background; but at the
beginning of the war the Führer told me that he was prepared to
solve this matter in a positive, generous fashion, but only after the
war.
The Nuremberg Laws were to exclude, for the future, that
concept of persons of mixed blood by means of a clear separation of
races. Consequently it was provided in the penal regulations of the
Nuremberg Laws, that never the woman but always the man should
be punishable, no matter whether he was German or Jewish. The
German woman or the Jewess should not be punished. Then quieter
times came, and the Führer was always of the opinion that for the
time being Jews should remain in economy, though not in leading
and prominent positions, until a controlled emigration, gradually
setting in, then intensified, should solve this problem. In spite of
continuous disturbances and difficulties in the economic field, the
Jews on the whole remained unmolested in their economic positions.
The extraordinary intensification which set in later did not really
start in until after the events of 1938, and then to a still greater
extent in the war years. But here, again, there was naturally one
more radical group for whom the Jewish question was more
significantly in the foreground than it was for other groups of the
Movement; just as, as I should like to emphasize at this point, the
idea of National Socialism as a philosophy was understood in various
ways—by one person more philosophically, by another mystically, by
a third in a practical and political sense. This was also true of the
different points of the program. For one person certain points were
more important, for another person less so. One person would see in
the point of the program which was directed against Versailles and
toward a free and strong Germany the main point of the program;
another person, perhaps, would consider the Jewish question the
main point.
THE PRESIDENT: Would that be a convenient time to break off?
Dr. Stahmer, can you inform the Tribunal how much longer you think
the Defendant Göring’s examination will last?
DR. STAHMER: I think that we shall finish in the course of
tomorrow morning.
THE PRESIDENT: That is a very long time.
DR. STAHMER: I shall do my best to shorten it.
[A recess was taken.]
Our website is not just a platform for buying books, but a bridge
connecting readers to the timeless values of culture and wisdom. With
an elegant, user-friendly interface and an intelligent search system,
we are committed to providing a quick and convenient shopping
experience. Additionally, our special promotions and home delivery
services ensure that you save time and fully enjoy the joy of reading.
ebookultra.com