Professional Documents
Culture Documents
Textbook Algebras Rings and Modules Volume 2 Non Commutative Algebras and Rings 1St Edition Michiel Hazewinkel Ebook All Chapter PDF
Textbook Algebras Rings and Modules Volume 2 Non Commutative Algebras and Rings 1St Edition Michiel Hazewinkel Ebook All Chapter PDF
https://textbookfull.com/product/foundations-of-commutative-
rings-and-their-modules-1st-edition-fanggui-wang/
https://textbookfull.com/product/biota-grow-2c-gather-2c-cook-
loucas/
https://textbookfull.com/product/non-associative-normed-algebras-
volume-2-representation-theory-and-the-zel-manov-approach-1st-
edition-miguel-cabrera-garcia/
https://textbookfull.com/product/chapter-1-burnside-rings-serge-
bouc/
Algebra 1 Groups Rings Fields and Arithmetic 1st
Edition Ramji Lal
https://textbookfull.com/product/algebra-1-groups-rings-fields-
and-arithmetic-1st-edition-ramji-lal/
https://textbookfull.com/product/universal-algebras-joseph-
muscat/
https://textbookfull.com/product/affine-vertex-and-w-algebras-
drazen-adamovic/
https://textbookfull.com/product/simple-relation-algebras-1st-
edition-steven-givant/
https://textbookfull.com/product/algebras-and-representations-
math-3193-2016-alison-parker/
Algebras, Rings and Modules
Non-commutative Algebras
and Rings
Volume 2
Algebras, Rings and Modules
Non-commutative Algebras
and Rings
Volume 2
Michiel Hazewinkel
Dept. of Pure and Applied Mathematics
Centrum Wiskunde & Informatica
Amsterdam, Netherlands
and
Nadiya Gubareni
Institute of Mathematics
¸
Czestochowa University of Technology
¸
Czestochowa, Poland
R
$6&,(1&(38%/,6+(56%22.
CRC Press
Taylor & Francis Group
6000 Broken Sound Parkway NW, Suite 300
Boca Raton, FL 33487-2742
This book contains information obtained from authentic and highly regarded sources. While all reasonable
efforts have been made to publish reliable data and information, neither the author[s] nor the publisher
can accept any legal responsibility or liability for any errors or omissions that may be made. The publish-
ers wish to make clear that any views or opinions expressed in this book by individual editors, authors
or contributors are personal to them and do not necessarily reflect the views/opinions of the publishers.
The information or guidance contained in this book is intended for use by medical, scientific or health-
care professionals and is provided strictly as a supplement to the medical or other professional’s own
judgement, their knowledge of the patient’s medical history, relevant manufacturer’s instructions and the
appropriate best practice guidelines. Because of the rapid advances in medical science, any information
or advice on dosages, procedures or diagnoses should be independently verified. The reader is strongly
urged to consult the relevant national drug formulary and the drug companies’ and device or material
manufacturers’ printed instructions, and their websites, before administering or utilizing any of the drugs,
devices or materials mentioned in this book. This book does not indicate whether a particular treatment
is appropriate or suitable for a particular individual. Ultimately it is the sole responsibility of the medi-
cal professional to make his or her own professional judgements, so as to advise and treat patients ap-
propriately. The authors and publishers have also attempted to trace the copyright holders of all material
reproduced in this publication and apologize to copyright holders if permission to publish in this form has
not been obtained. If any copyright material has not been acknowledged please write and let us know so
we may rectify in any future reprint.
Except as permitted under U.S. Copyright Law, no part of this book may be reprinted, reproduced, trans-
mitted, or utilized in any form by any electronic, mechanical, or other means, now known or hereafter
invented, including photocopying, microfilming, and recording, or in any information storage or retrieval
system, without written permission from the publishers.
For permission to photocopy or use material electronically from this work, please access www.copyright.
com (http://www.copyright.com/) or contact the Copyright Clearance Center, Inc. (CCC), 222 Rosewood
Drive, Danvers, MA 01923, 978-750-8400. CCC is a not-for-profit organization that provides licenses and
registration for a variety of users. For organizations that have been granted a photocopy license by the
CCC, a separate system of payment has been arranged.
Trademark Notice: Product or corporate names may be trademarks or registered trademarks, and are
used only for identification and explanation without intent to infringe.
Visit the Taylor & Francis Web site at
http://www.taylorandfrancis.com
and the CRC Press Web site at
http://www.crcpress.com
Preface
This the second part of a two-part treatise on (selected parts of) non-commutative
algebra and ring theory. The contents of the first part [103] are given below.
Representation theory is a fundamental tool for studying groups, algebras and
rings (and many other things). In this volume we consider representation theory for
finite posets, finite dimensional algebras and semiperfect rings. The description of
modules over some classes of semiperfect rings is reduced to mixed matrix problems
over discrete valuation rings and division rings, which are considered in this book.
One of the main goals in ring theory is to reduce in a certain sense the description of
large classes of rings to simpler classes by the use of some ring theoretic constructions.
The best classical example is the Wedderburn-Artin theorem, however, there are a
number of other general results that can be mentioned.
Section 1.1 represents the definition and main properties of the basic main
construction of rings which are incidence rings of posets over associative rings.
In section 1.2 we consider some properties of a special class of incidence rings of
the form T(S) = I(S, D), where S is a finite partially ordered set and D is a division
ring. This class of rings properly contains the class of hereditary serial rings and all
Artinian rings with quivers that are trees.
A special class of right hereditary rings A(S, O) of a finite poset S over a family of
discrete valuation rings {Oi }i ∈I with a common skew field of fractions is introduced
and studied in section 1.3.
In section 1.4 we introduce and study a special class of incidence rings modulo
the radical of the form I(S, Λ, M) related to a finite poset S over a local Noetherian
domain Λ with Jacobson radical M.
Serial and semidistributive rings of the form I(S, Λ, M), where Λ is a discrete
valuation ring with Jacobson radical M are considered in section 1.5.
The main aim of chapter 2 is to study the properties and structure of different
classes of rings whose lattices of submodules are distributive (or semidistributive).
Such rings are called distributive (or semidistributive) and they can be considered as
a non-commutative generalization of Prüfer domains. The class of distributive rings
is very wide and includes, for example the ring of integers and the ring of polynomials
K[x] over a field K, rings of integral algebraic numbers, and commutative principal
ideal rings. More generally, all commutative Dedekind rings and Prüfer domains are
examples of distributive rings. The first papers which are devoted to distributive rings
appeared in the fifties-sixties of the 20-th century in [26], [27], [29], [155], [156].
vi Algebras, Rings and Modules
The systematic study of these rings began with papers of H. Achkar [2], V. Camillo
[39], H.H. Brungs [33] and W. Stephenson [203].
In section 2.1 we consider the main properties of distributive modules and rings.
Section 2.2 is devoted to study of semidistributive modules and rings. The structure
of Noetherian distributive and semidistributive rings is discussed in section 2.3.
The properties of right hereditary SPSD-rings are studied in section 2.4. It is
shown that the structure of all such rings is closely connected with right hereditary
rings of the form A(S, O) as described in section 1.3.
Section 2.5 is devoted to the study of the main properties and structure of
semihereditary SPSD-rings.
Some of the more fundamental notions and results of the theory of homological
algebra were studied in different sections of [100], [101]. Chapter 3 gives some
additional facts from homological algebra.
The notions of direct limits and inverse limits for a set of modules were considered
in [103, section 1.4] (see also [100, section 4.7]). In the particular case, when
this set has only two modules these constructions have their own names, pullback
and pushout. They are very important and useful in studying rings and modules.
Therefore for the convenience of the reader these module constructions and their
main properties are considered in more detail in section 3.1.
In the theory of homological algebra and its applications there is very important
the statement which is often known as the “snake lemma” (it is also called “zig-zag
lemma”, or “serpent lemma”). It is valid in every Abelian category and it is an
important tool in the construction of long exact sequences. This lemma consists
of two claims: 1) the construction of an exact sequence, which is often called
“kernel-cokernel sequence”, for any commutative diagram of a special type; and 2)
the construction of long exact sequences of homology groups for any given short
exact sequence of complexes. The second part of this lemma was proved in [100,
theorem 6.1.1]. The proof of the first part of this lemma is given in section 3.2.
The functors Extn as right derived functors for the contravariant left exact functor
Hom using projective resolutions were introduced in [100, section 6.4]. This functor
is closely related to module extensions.
Let A be a ring, and let X, Y ∈ Modr A. In this chapter the interpretation of the
group Ext1A(Y, X) is given in terms of short exact sequences. Section 3.3 is devoted
to studying extensions of modules in terms of short exact sequences.
Following R. Baer the addition of extensions of modules, which makes the set of
equivalence classes of all extensions an Abelian group, is introduced in section 3.4.
Some main properties of the group Ext1A(Y, X) are considered in section 3.5. The
results of section 3.6 show that there is an isomorphism between equivalence classes
of the group of extensions of X by Y and elements of Ext1A(Y, X) as considered in
[100].
In chapter 4 the main results about modules over semiperfect rings are given.
Some basic properties of semiperfect rings and modules over them were considered
in [100, chapter 10]. There are a number of equivalent definitions of semiperfect
rings. One of them is given by H. Bass in terms of projective covers. He proved
Preface vii
that a ring is semiperfect if and only if any finitely generated module has a projective
cover (see [103, theorem 1.8.2]).
In [100, section 10.4] the structure of finitely generated projective modules
over semiperfect rings was discussed. It was proved that any such right module
can be uniquely decomposed into a direct sum of principal right modules. The
generalization of this statement, the important theorem [103, theorem 7.2.7], states
that any projective module over a semiperfect ring is a direct sum of principal modules.
In section 4.1 this result is used to study the structure of finitely generated modules
over semiperfect rings. The main result of this section, which was proved by R.B.
Warfield, Jr. concerns the decompositions of any finitely generated module over a
semiperfect ring. This theorem gives a possibility to introduce the notion of stably
isomorphic modules.
In section 4.2 we prove that all modules over a semiperfect ring can be divided
into the equivalence classes of stably isomorphic modules. Moreover, each stable
isomorphism class of finitely generated modules over a semiperfect ring contains a
unique (up to isomorphism) minimal element. Most of the results of this section were
obtained by R.B. Warfield, Jr. (see [220], [222]).
In [101, section 4.10] we considered the duality in Noetherian rings, which is
given by the covariant functor ∗ = Hom A(−, A). For an arbitrary ring A this functor
induces a duality between the full subcategories of finitely generated projective right
A-modules and left A-modules. In section 4.3 the main properties of this functor and
torsionless modules are studied for the case of modules over semiperfect rings.
Section 4.4 presents an introduction to the duality theory of Auslander and
Bridger [13], and yields a connection between finitely presented right modules and
finitely presented left modules over semiperfect rings. Some main properties of the
Auslander-Bridger transpose, which is closely connected with almost split sequences,
are discussed. These sequences were first introduced and studied by M. Auslander
and I. Reiten in [15] and [20], and they play an important role in the representation
theory of rings and finite dimensional algebras. Section 4.5 is devoted to the study
of some main properties of these sequences. In section 4.6 we study almost split
sequences over semiperfect rings and prove the existence of these sequences for
strongly indecomposable modules. Section 4.7 presents an introduction to the theory
of linkage of modules over semiperfect Noetherian rings using two types of functors:
syzygy and transpose.
Chapter 5 is devoted to finite partially ordered sets (posets) and their
representations, which play an important role in representation theory. They were first
introduced and studied by L.A. Nazarova and A.V. Roiter [163] in 1972 in connection
with problems of representations of finite dimensional algebras. M.M. Kleiner
characterized posets of finite type [129] and described their pairwise non-isomorphic
indecomposable representations [130]. He proved a theorem which gives a criterium
for posets to be of finite representation type.
Recall that a finite poset P is called primitive if it is a cardinal sum of linearly
ordered sets L1, . . . , Lm . It is then denoted by P = L1 ⊔ · · · ⊔ Lm . This chapter
viii Algebras, Rings and Modules
gives the proof of the criterium for primitive posets to be of finite representation type
following [30].
To prove this criterium there are used only the trichotomy lemma which was proved
by P. Gabriel and A.V. Roiter in [82], the Kleiner lemma about the representations
of a pair of finite posets proved by M. Kleiner in [129] and the main construction,
considered in section 5.5. Note that this construction (in some form) was introduced
by L.A. Nazarova and A.V. Roiter in [163].
An important problem in the theory of representations of finite dimensional
algebras (or f.d. algebras, in short) is to obtain the full list of different kinds of
algebras which are of finite representation type (or finite type, or f.r.t., in short). The
first classes of associative f.d. algebras of f.r.t which have been described were the
classes of algebras with zero square radical and hereditary algebras over algebraically
closed fields.
There are different approaches to study the representations of f.d. algebras. One
of them is the approach of P. Gabriel [79], which reduces the study of representations
of algebras to the study of representations of quivers. Another approach was first
considered by L.A. Nazarova and A.V. Roiter [163]. This approach is to solve “matrix
problems”, that is, the reducing of some classes of matrices by means of admissible
transformations to their simplest form. A third approach is due to M. Auslander and
it is connected with the technique of almost split sequences.
Chapter 6 can be considered as an introduction to the theory of representations of
quivers and finite dimensional algebras. This chapter gives some main notions and
some fundamental results of these representations, most of which are given without
proof. In section 6.1 we consider the notions of finite quivers and their representations
and give the main results of this theory. Section 6.2 is devoted to species and their
representations. In section 6.3 we consider some main notions and results of the
representation theory of finite dimensional algebras.
As it turns out the category of representations of finite dimensional algebras is
equivalent to the category of representations of special classes of quivers, which are
called bound quivers. That is why the quivers play a central role in the theory of
finite dimensional associative algebras and their modules.
For right Artinian rings one can also introduce the notion of a ring of finite
representation type. As has been shown by D. Eisenbud and P. Griffith [71] this
notion is left-right symmetric. They proved this fact using the duality theory of
Auslander and Bridger. This result is proved in section 7.1.
For finite dimensional algebras along with the notion of finite representation type
there is also considered the notion of bounded representation type. Recall that a
finite dimensional algebra A is called of bounded representation type if there is a
bound on the length of the indecomposable finite dimensional A-modules. The first
Brauer-Thrall conjecture says that these notions are the same in the case of a finite
dimensional algebra A (as was proved by A.V. Roiter [101, theorem 3.5.1]) and in the
case of Artinian algebras (as was proved by M. Auslander [15]).
A ring A, not necessarily Artinian, is said to be of finite representation type, if it
has a finite number of non-isomorphic indecomposable finitely presented A-modules.
Preface ix
For Artinian rings this definition coincides with the earlier one, because in this
case each finitely generated A-module is finitely presented as well. The main
results concerning Artinian hereditary rings of finite representation type are given in
section 7.4.
For Artinian rings along with the notion of finite representation type there is
considered the notion of bounded representation type. Recall that a right Artinian
ring A is said to be of bounded representation type if there is a bound on the
length of finitely generated indecomposable right A-modules. The first Brauer-Thrall
conjecture asserts that these notions are the same. M. Auslander proved that this
conjecture is true for right Artinian rings (see theorem 7.2.11).
Following R.B. Warfield, Jr. a ring has right bounded representation type if
there is an upper bound on the number of generators required for indecomposable
finitely presented right A-modules. In his paper [222] R.B. Warfield, Jr. puts the
following question:
Question 4. For what semiperfect rings is there an upper bound on the number of
generators required for the indecomposable finitely presented modules?
Chapter 8 describes special classes of semiperfect rings of bounded representation
type, which constitute some sort of answer to this question.
As shown by R.B. Warfield, Jr. there is a serious restriction on the structure
of rings of bounded representation type connected with modules of finite Goldie
dimension. In section 8.1 it is proved that a semiprimary ring of finite right bounded
representation type is right Artinian (see [222]).
O-species and tensor algebras, which are generalizations of k-species as introduced
by P. Gabriel, are considered in section 8.2. The connection between right hereditary
SPSD-rings and special kinds of (D, O)-species is considered in section 8.3.
In section 8.4 we discuss the reduction of representations of (D, O)-species to
mixed matrix problems over discrete valuation rings and their common skew field of
fractions. Some important mixed matrix problems are considered in section 8.5.
Sections 8.6 and 8.7 are devoted to the study of (D, O)-species of bounded
representation type. There is a theorem which gives the structure of these species
in the terms of diagrams which can be considered as generalizations of Dynkin
diagrams.
The right hereditary SPSD-rings of finite right bounded representation type are
described in section 8.8.
The book is written on a level accessible to advanced students who have some
experience with modern algebra. It will be useful for those new to the subject as well
for researchers and serves as a reference volume.
While writing this book the second author was in particular supported by FAPEST
of Brazil in 2010. The author would like to express cordial thanks to the Institute
of Mathematics and Statistics of the University of São Paulo, and especially prof.
M. Dokuchaev and prof. V.M. Futorny, for their warm hospitality during her visit in
2010. The author is also grateful to prof. V.V. Kirichenko for useful discussions and
fruitful remarks.
CONTENTS
Contents of Volume 1
Preface v
1. Preliminaries . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1
1.1 Basic Concepts Concerning rings and Modules . . . . . . . . . . . . . . . . . . . . . . . . . 1
1.2 Categories and Functors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10
1.3 Tensor Product of Modules . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 15
1.4 Direct and Inverse Limits . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17
1.5 Projective, Injective, and Flat Modules . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19
1.6 The Functor Tor . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 23
1.7 The Functor Ext . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 24
1.8 Hereditary and Semihereditary Rings . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 26
1.9 Semiperfect and Perfect Rings . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 28
1.10 Serial and Semidistributive Rings . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 31
1.11 Classical Rings of Fractions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 34
1.12 Quivers of Algebras and Rings . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 35
3. Valuation Rings . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 74
3.1 Valuation Domains . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 75
3.2 Discrete Valuation Domains . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 85
3.3 Invariant Valuation Rings of Division Rings . . . . . . . . . . . . . . . . . . . . . . . . . . 87
3.4 Examples of Non-commutative Non-discretely-valued Valuation Rings . . 93
3.5 Discrete Valuation Rings of Division Rings . . . . . . . . . . . . . . . . . . . . . . . . . . 99
3.6 Total Valuation Rings . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 102
xii Algebras, Rings and Modules
Contents of Volume 2
Preface v
1. Rings Related to Finite Posets . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1
1.1 Incidence Rings . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1
1.2 Incidence Rings I(S, D) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8
1.3 Right Hereditary Rings A(S, O) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13
1.4 Incidence Rings Modulo the Radical . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19
1.5 Serial and Semidistributive Rings I(S, Λ, M) . . . . . . . . . . . . . . . . . . . . . . . . . . 23
1.6 Notes and References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 24
2. Distributive and Semidistributive Rings . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 26
2.1 Distributive Modules and Rings . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 26
2.2 Semiprime Semidistributive Rings . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 36
2.3 Semiperfect Semidistributive Rings . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 42
2.4 Right Hereditary SPSD-Rings . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 44
2.5 Semihereditary SPSD-Rings . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 55
2.6 Notes and References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 59
3. The Group of Extensions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 61
3.1 Module Constructions Pushout and Pullback . . . . . . . . . . . . . . . . . . . . . . . . . 62
3.2 The Snake Lemma . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 71
3.3 Extensions of Modules . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 77
3.4 Baer Sum of Extensions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 82
3.5 Properties of Ext1 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 86
3.6 Ext1 and Extensions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 88
3.7 Additive and Abelian Categories . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 92
3.8 Notes and References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 95
4. Modules Over Semiperfect Rings . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 98
4.1 Finitely Generated Modules over Semiperfect Rings . . . . . . . . . . . . . . . . . . . 99
xiv Algebras, Rings and Modules
One of the main goals in the ring theory is to reduce in a certain sense the description
of large classes of rings to simpler classes using some ring theoretic constructions.
The most classical example is the Wedderburn-Artin theorem, however one can
mention a number of other general results.
Section 1.1 represents the definition and main properties of a basic construction
of rings. These are called incidence rings of posets over associative rings.
In section 1.2 we consider some properties of a special class of incidence rings of
the form T(S) = I(S, D), where S is a finite partially ordered set and D is a division
ring. This class of rings properly contains the class of hereditary serial rings and all
Artinian rings with quivers that are trees.
A special class of right hereditary rings A(S, O) of a finite poset S over a family of
discrete valuation rings {Oi }i ∈I with a common skew field of fractions is introduced
and studied in section 1.3.
In section 1.4 we introduce and study a special class of incidence rings. These
are called incidence rings modulo the radical. They are denoted by I(S, Λ, M) and
involve a finite poset S over a local Noetherian domain Λ with Jacobson radical M.
Serial and semidistributive rings of the form I(S, Λ, M), where Λ is a discrete
valuation ring with Jacobson radical M, are considered in section 1.5.
This section covers the definition and some properties of incidence rings of partially
ordered sets over associative rings.
Let S be a set with a binary relation denoted by ≼. The relation ≼ is called a
preorder if the following properties satisfy:
Let S be a poset with partial ordering relation ≼ (which is transitive, reflexive and
antisymmetric). Denote by x ≺ y the strict order, i.e. the relation "x ≼ y and x , y".
An element x ∈ S is called maximal if there is no element y ∈ S satisfying x ≺ y.
Dually, x is minimal if there is no element y ∈ S satisfying y ≺ x. An element x ∈ S
is said to be a least element if x ≼ y for all y ∈ S, and x ∈ S is said to be a greatest
element if y ≼ x for all y ∈ S.
A subset C of a poset S is called a chain if it is totally ordered, i.e. for any x, y ∈ C,
either x ≼ y or y ≼ x. Denote a chain of cardinality n by Cn , or (n). In this case
the number n is called the length of the chain Cn . A subset A of a poset S is called
an antichain if all its elements are pairwise incomparable in S, i.e. for any pair of
distinct elements x, y ∈ A, both x y and y x.
In order to visualize a poset S one uses the so called diagram of S. Let x and y
be distinct elements of S. It is said that y covers x if x ≺ y but there is no element z
such that x ≺ z ≺ y. Recall that the diagram H(S) of a poset S is the directed graph
whose vertex set is S and whose set of edges is given by the set of covering pairs
(x, y) of S, moreover, there is a (directed) edge from a vertex x up to a vertex y if and
only if y covers x.
For example, the diagram below
•O a3
• a1 • a2
represents the poset (S, ≼) with 3 elements {a1, a2, a3 } and the one relation a2 ≺ a3 .
Remark 1.1.1. The diagram H(S) of a poset S is often called its Hasse diagram.
Usually it is drawn in the plane in such a way that if y covers x then the point
representing y is drawn higher than the point representing x. In this case the Hasse
diagram is drawn without arrows. For example, the Hasse diagram below
• a3
• a1 • a2
represents the same poset (S, ≼) as above, i.e. S = {a1, a2, a3 } with the one relation
a2 ≺ a3 .
Rings Related to Finite Posets 3
The idea of incidence algebras had its beginning in the works of R. Dedekind
and E.T. Bell. The incidence algebra of a locally finite poset over a field was first
introduced by G.-C. Rota in [185]. This notion can be extended to the case of a
non-commutative ring A.
Definition 1.1.3. Let S be a locally finite poset with a partial ordering relation ≼,
and A an associative (not necessarily commutative) ring with identity. The incidence
ring I(S, A) of S over A is the set of functions f : S × S → A such that f (x, y) = 0 if
x y with operations:
( f + g)(x, y) = f (x, y) + g(x, y),
∑
( f g)(x, y) = f (x, z)g(z, y)
x ≼z ≼y
(a f )(x, y) = a( f (x, y))
for all x, y, z ∈ S and a ∈ A.
Let us restrict our attention (for the moment) to the case of finite posets. Let
S = {α1, . . . , αn } be a finite poset with a partial ordering relation ≼, and let Mn (A) be
the generalized matrix ring of all n × n-matrices with elements from an associative
ring A. In this case the incidence ring of S over A is the subring I(S, A) of Mn (A)
such that the (i, j)-entry of I(S, A) is equal to 0 if αi α j in S. So, letting correspond
to any element αi ∈ S the matrix unit eii ∈ Mn (A), and to any pairs of elements
αi, α j ∈ S such that αi ≼ α j the matrix unit ei j ∈ Mn (A) one obtain a basis (over A)
of I(S, A).
d
•8
888
88
b •8 •c (1.1.5)
88
88
8
•
a
4 Algebras, Rings and Modules
Example 1.1.6. Let S be the poset with Hasse diagram of the following form:
2 •3 • 3
33
33
3
• 1
• 4
then
A A A 0
© ª
0 A 0 0®
I(S, A) = ®
0 0 A 0®
« A A A A¬
As these examples illustrate, transitivity of the ordering relation has much to do
with "I(S, A) is a subring of Mn (A)".
The following statement is obvious:
Lemma 1.1.7. Every interval of a finite poset S is a chain if and only if S contains
no subposets whose diagrams are rhombuses.
Lemma 1.1.8. The non-oriented graph S is a disconnected union of trees if and only
if S contains no subposets whose diagrams are rhombuses.
Lemma 1.1.9. A finite poset S with partial ordering relation ≼ can be labeled as
S = {α1, . . . , αn } where αi ≼ α j implies i ≤ j.
Using this lemma it is easy to show that in the case that S has property just
mentioned the incidence ring I(S, A) is isomorphic to a subring of the ring of the
upper triangular matrix ring Tn (A). In particular, if a poset S is linearly ordered then
I(S, A) Tn (A).
1. I(S, A) is an FDI-ring.
2. I(S, A) is a semiperfect ring if and only if A is semiperfect.
3. The Jacobson radical of I(S, A) is the each set of elements of I(S, A) for which
(i, i)-entry is in R.
4. I(S, A) is a right (left) Noetherian ring if and only if A is right (left) Noetherian.
5. I(S, A) is a right (left) Artinian ring if and only if A is right (left) Artinian.
Proof.
Remark 1.1.11. The notion of an incidence ring has been generalized to the case of
preodered sets in [75], [204].
Let S be a finite preordered set with a preorder ≼, and A an associative (not
necessarily commutative) ring with identity. The incidence ring I(S, A) of S over A is
a ring with the additive structure of a free A-module with basis { fxy : x ≼ y; x, y ∈ S},
where multiplication is given by the linear extension of:
{
fxu if y = z
fxy · fzu = (1.1.12)
0 otherwise,
for all x, y, u, v ∈ V.
In fact, M(B, A) is simply the incidence ring of the preordered set ℜ over the ring A.
In this section we have (so far) considered incidence rings of posets and preordered
sets. They are special cases of a general construction - the incidence rings of an
arbitrary relation which is not required to be a preorder. Such rings were introduced
by G. Abrams in [1] and were called generalized incidence rings.
for all x, y, u, v ∈ S.
x ≼ u ∈ S ⇐⇒ y ≼ v ∈ S.
Remark 1.1.19. Note that a relation which is both reflexive and transitive is always
balanced, but there are balanced relations which are not transitive. Some of these are
shown in Example 1.1.23 below.
Proof.
1. If x, y, z, w, u, v ∈ S then the expressions ( fxy fzw ) fuv and fxy ( fzw fuv ) are both
zero unless y = z, w = u and x ≼ v. In this case, the first product is equal to
fxv precisely when x ≼ w, while the second product is equal to fxv precisely
when z ≼ v, i.e. this is the case if and only if ≼ is a balanced relation.
2. From (1.1.16) it immediately follows that the elements fxx = ex for all x ∈ S
are idempotents in I(S, ≼, A) and
{ {
fyz if x = y fyz if x = z
ex fyz = , fyz ex =
0 otherwise 0 otherwise
∑
for all x, y, z ∈ S. This implies that the element ex is a multiplicative
x ∈S
identity of I(S, ≼, A).
The definition of a generalized incidence ring can be given in terms of graphs (see
[4]).
Rings Related to Finite Posets 7
Definition 1.1.21. Let S = (V, E) be a graph with set of vertices V = {1, 2, . . . , n} and
set of edges E ⊆ V × V. Then the associated generalized incidence ring I(S, A) of
S over an associative ring A is defined as the free left A-module with basis consisting
of all edges in E where multiplication is defined by the linear extension of
{
(x, v) if y = u and (x, v) ∈ E
(x, y)(u, v) =
0 otherwise,
for all x, y, u, v ∈ V.
(x, u) ∈ E ⇐⇒ (y, v) ∈ E.
Example 1.1.23. We omit loops in all diagrams represented in these examples. The
following directed graphs are balanced and nontransitive.
a) ; u O • FFF b) ; u • FFF
xxxx FF xxxx FF
x # x #
y • cF /•v y • cF
FF x; FF x;• v
F xxxx F xxxx
•x •x
/• /•
c) ; u • cFFF d) •
xxxx FF x y u
x
y • cF •v
FF xxx;
F xx
•x
Proposition 1.1.20 and these examples show, in particular, that there are associative
generalized incidence rings which are not incidence rings of preordered sets. Note
that Example 1.1.23 is in terms of directed graphs but not Hasse diagrams.
Remarks 1.1.24.
Definition 1.1.26. Let S be a set with a binary relation ≼. A ring I = I(S) with identity
is called a generalized incidence ring if there exists a maximal set of idempotents
E = {eu : u ∈ S, eu ∈ I} such that:
Example 1.1.27. Let A be a semiperfect ring. Then there exists a finite set E =
{e1, e2, . . . , en } of pairwise orthogonal local idempotents of A such that 1 = e1 + e2 +
· · · + en . Define the binary relation on E such that ei ≼ e j if and only if ei Ae j , 0.
Then A with respect to this binary relation is a generalized incidence ring in sense of
definition 1.1.26.
Remark 1.1.29. If S and S ′ are isomorphic then of course the (various) incidence
rings I(S, A) and I(S ′, A) are isomorphic. One may wonder to what extent the reverse
is true. Some results and references pertaining to this are given in section 1.6 below.
In this section we consider the particular example of incidence rings I(S, D) of a finite
poset S over a division ring D.
Rings Related to Finite Posets 9
Let S = {α1, α2, . . . , αn } be a finite poset with a partial order relation ≼, and
D a division ring. Denote by Mn (D) the ring of all (n × n)-matrices over D with
matrix units ei j , where dei j = ei j d for all d ∈ D. Consider the subring T(S, D) ⊂
Mn (D) which is generated by those ei j ∈ Mn (D) for which αi ≼ α j in S. This is
equivalent to the following: eii T(S, D)e j j = D if and only if αi ≼ α j in S, otherwise
eii T(S, D)e j j = 0.
Obviously, T(S, D) = I(S, D). We will often denote this ring T(S) if a division
ring D is fixed and known from the context.
D D D D
© ª
0 D 0 D®
T(S) = ®. (1.2.2)
0 0 D D®
« 0 0 0 D¬
As was shown in [103, Example 9.1.10], the ring T(S) is a piecewise domain but it is
neither right nor left hereditary.
Proposition 1.2.4. There is a labeling of a poset S such that the ring T(S) is a
triangular ring, and so is isomorphic to a subring of Tn (D), where n is the number
of elements of the poset S.
Proof. Since S = {α1, . . . , αn } is a finite poset, S can be labeled in such a way that
αi ≼ α j if i ≤ j. Then for this numbering T(S) is obviously a triangular ring and
T(S) ⊆ Tn (D). Moreover, one has a strong equality if (and only if) S is a linearly
ordered set.
Proposition 1.2.6.
Recall that a semiperfect ring A with Jacobson radical is a basic ring if A/R is a
direct sum of division rings.
r = diag(r1, r2, . . . , rn ) + x,
or ⊕ t
eii R/eii R2 U ji j ,
j
where the U j are right simple T/R-modules, then in the quiver Q(T) the vertex i is
connected with the vertex j by ti j arrows.
Proposition 1.2.9. The quiver Q(T) of the ring T = T(S, D) coincides with the Hasse
diagram H(S) of the poset S.
Proof. One can assume that T = T(S, D) is an indecomposable ring. Then by [100,
theorem 11.1.9], the quiver Q(T) is connected. Let {ei j } be the set of all matrix units
of Mn (D) and 1 = e11 + e22 + · · · + enn the natural decomposition of the identity of
T(S) into a sum of pairwise orthogonal idempotents.
Rings Related to Finite Posets 11
Consider the Hasse diagram H(S) of the poset S and the following cases.
1. Assume that there is an arrow i → j in the diagram H(S), which means that
eii Re j j = D and eii Rekk = 0 or ekk Re j j = 0 for any integer 1 ≤ k ≤ n. Then
∑n
eii R2 e j j = k=1 eii Rekk ekk Re j j = 0. Therefore there is exactly one arrow
i → j in the quiver Q(T).
2. Assume that αi ≼ α j and there is no arrow of the form i → j in H(S).
This means that there is an integer 1 ≤ k ≤ n such that eii Rekk = D or
∑n
ekk Re j j = D. Then eii R2 e j j = k=1 eii Rekk ekk Re j j = D and therefore
eii Re j j /eii R2 e j j = 0, i.e. there is no arrow i → j in the quiver Q(T).
P j −→ eii R −→ 0
Recall that the endomorphism ring End A(P) of a finitely generated projective
A-module P is called a minor of the ring A (see [100, Section 13.3]). If P can
be decomposed into a sum of n indecomposable modules then End A(P) is called a
minor of order n. So that the ring End(eA) = eAe is a minor for any idempotent e
of a ring A. If e is a sum of n primitive idempotents then End(eA) = eAe is a minor
of order n. It was shown that the minors of right (left) Noetherian rings are right
(left) Noetherian, the minors of right (left) Artinian rings are right (left) Artinian (see
[100, Theorem 3.6.1]). From [100, Theorem 10.3.8] and [100, Corollary 10.3.11] it
follows that any minor of a semiperfect ring is semiperfect. If e is an idempotent of
a hereditary (semihereditary) ring then eAe is also hereditary (semihereditary) (see
[103, proposition 4.4.6(4)] and [103, theorem 8.2.16]).
Let
A11 A12 · · · A1n
© ª
A21 A22 · · · A2n ®
A= . .. .. .. ®®
.. . . . ®
« A n1 A n2 · · · A nn ¬
be a generalized matrix ring, where Aii are rings with identity 1i and Ai j are
Aii -A j j -bimodules. Then the identity of the ring A can be decomposed into the
sum 1 = e1 + e2 + · · · + en , where ei is an element of A with entry equal to 1i at (i, i)
place and 0 otherwise. So that ei Ae j = Ai j . Obviously, each ei is an idempotent of
A. Therefore eAe is a minor of A if e is a sum of some distinct idempotents ei whose
index is in a subset X ⊆ {1, 2, . . . , n}.
Proposition 1.2.10. A ring T(S, D) is hereditary if and only if the Hasse diagram
H(S) of S is a disconnected union of trees, i.e. the poset S contains no subposet
whose diagram is a rhombus.
12 Algebras, Rings and Modules
Proof. Let T = T(S, D) be a hereditary ring. Assume that the Hasse diagram of S
contains a subposet whose Hasse diagram is a rhombus. This means that T contains
an idempotent e such that eT(S)e = B, where
D D D D
© ª
0 D®®
D 0
B= ®.
0 0 D D®®
«0 0 0 D¬
This is impossible, since T(S) is a hereditary ring and for any non-zero idempotent
e ∈ T(S) the ring B = eT(S)e must be also a hereditary ring by [103, proposition
4.4.6(4)], but the ring B is not hereditary.
Conversely, suppose the two-sided Peirce decomposition of T(S, D) does not
contain minors of the form eT(S)e = B. Then the Hasse diagram of S is a
disconnected union of acyclic simply laced quivers with no extra arrows such that its
underlying graph S (obtained from S by deleting the orientation of the arrows) is a
disconnected union of trees. From proposition 1.2.9 it follows that the ring T(S, D)
can be considered as the path algebra corresponding to the graph S over a division
ring D. Therefore T(S, D) is a hereditary ring, by [101, theorem 2.3.4].
Proof. Since a ring T(S, D) is indecomposable if and only if the poset S is connected,
one can assume that T(S, D) is indecomposable. If S is a chain, then the statement
follows from example 1.2.3.
Rings Related to Finite Posets 13
Let O be a discrete valuation ring (not necessarily commutative) with division ring
of fractions D and Jacobson radical M = radO = πO, where π ∈ O is a generator
of M. By [100, corollary 10.2.2] O is a local Noetherian hereditary ring which is a
right and left principal ideal domain (PID) and M is its unique maximal right and left
ideal. Recall (see [100, section 8.4]) that each non-zero element x ∈ D is uniquely
representable in the form x = επ k , where ε is an invertible element of the ring O and
k ∈ Z. Each proper O-module X ⊂ D is cyclic and can be represented in the form
X = π k O = Oπ k , where k ∈ Z. Write
O O ··· O
© ª
M O ··· O®
Hn (O) = . .. .. .. ®® (1.3.1)
.. . . .®
«M M ··· O¬
which is a subring of the matrix ring Mn (D). Clearly, Hn (O) is a Noetherian serial
prime hereditary ring. And so, by the Goldie theorem, it has a classical ring of
fractions, which is Mn (D).
Let {Oi }i=1,...,k be a family of discrete valuation rings (not necessarily
commutative) with Jacobson radical Mi and a common skew field of fractions
D. Let S = {α1, α2, . . . , αn } be a finite poset with a partial order ≼, and
S0 = {α1, α2, . . . , αk } the subset of minimal points of S (k 6 n). Then S = S0 ∪ S1 ,
where S1 = {αk+1, αk+2, . . . , αn }. Corresponding to this partition of S consider a
poset S with weights so that the point i has the weight Hni (Oi ) for i = 1, 2, . . . , k;
ni ∈ N; and all other points have the weight D.
Construct a ring A = A(S, S0, S1 ; O1, . . . , Ok ; D; n1, n2, . . . , nk ) (or shortly,
A(S, O)) as a generalized matrix ring of the form
Proposition 1.3.3. Let {Oi }i=1,...,k be a family of discrete valuation rings with
common division ring of fractions D. Then A(S, O) is a right Noetherian semiperfect
ring.
Proof. By [103, theorem 1.1.23] the ring A is right Noetherian. Since the identity of
A = A(S, O) decomposes into a finite sum of pairwise orthogonal local idempotents,
A is semiperfect by [103, theorem 1.9.3]. Since all the Hni (Oi ) are Noetherian rings,
T(S1, D) is an Artinian ring, and all Mi are finite dimensional vector spaces over D,
A is a right Noetherian ring by [103, theorem 1.1.23].
Let N be the prime radical of a semiperfect ring A = A(S, O). Then N has the
following form
I1 ... 0 M1
© ª
.. . . .. .. ®
. . . . ®
N= ®
O ... I Mk ®
k ®
« 0 ... 0 N1 ¬
where Ii is the prime radical of Hni (Oi ) for (i = 1, 2, . . . , k), and N1 is the prime
radical of T(S1, D). Since all the Hni (Oi ) are prime rings, Ii = 0. By [103, proposition
1.1.25] the prime radical of the Artinian ring T(S1, D) coincides with its Jacobson
radical R1 = rad T(S1, D). Therefore
0 ... 0 M1
© ª
... .. .. .. ®
. . . ®
N = ®
®
0 ... 0 Mk ®
®
0 ... 0 R1
« ¬
Another random document with
no related content on Scribd:
ähnlichen Verwüstungen, wie man sie hier täglich zu Dutzenden
sieht, vorzubeugen. Beim Europäer mit seiner weißen Haut ist das
Auffinden des Sandflohes übrigens weit leichter, als es den
Schwarzen gemacht wird, von deren Haut sich der dunkle Punkt
kaum abhebt. Die vier oder fünf Sandflöhe, die mich trotz steten
Tragens hoher, geschlossener Schnürschuhe bisher zu ihrem Sitz
auserkoren haben, hat mir der vielgewandte Knudsen
herausgehoben; ein Auswaschen der Höhlung mit Sublimat
erscheint mir dabei immer ganz angebracht. Die Neger haben ein
anderes Desinficiens, sie füllen die Öffnungen mit
Wurzelgeschabsel; in einem winzigen Makuadorf am Steilabhang
des Plateaus südlich von Newala sah ich eine Frau, die den Raum
unter den Nägeln prophylaktisch mit Wurzelpulver ausstopfte. Ob es
der Alten etwas nützen wird, wer weiß es.
Der Rest der vielen kleinen Hindernisse, die uns hier das Dasein
erschweren, wirkt mehr komisch als ernsthaft. In Ermangelung von
etwas anderem Rauchbaren greifen Knudsen und ich jetzt zu dem
Inhalt einer vom Inder in Lindi bezogenen Zigarrenkiste. Diese ist
sehr schön beklebt und aufgemacht, aber wehe dem Unglücklichen,
der sich, wie wir, mit ihrem Inhalt befaßt! Ob diese schwelenden
Giftnudeln Opium oder ein anderes Narkotikum enthalten, von uns
beiden weiß es niemand zu sagen, denn nach dem zehnten Zuge
sind wir beide „matt“; dreiviertel betäubt und hundeelend liegen dann
Wiking und Deutscher in sich zusammengesunken da. Langsam
erholt man sich — was geschieht? Nach einer halben Stunde greift
man doch wieder zu dem scheußlichen Kraut; so unstillbar ist hier in
den Tropen der Drang zum Rauchen!
Auch meine jetzigen Fieberanfälle sind kaum geeignet, noch
ernst genommen zu werden. Ich habe ihrer hier in Newala nicht
weniger als drei gehabt, aber alle mit unglaublich kurzem Verlauf.
Emsig fragend, schreibend und notierend quäle ich mich mit meinen
„Gelehrten“ herum, der starke Mittagskaffee hat die Lebensgeister
mächtig angeregt; das Gehirn arbeitet außerordentlich intensiv, so
daß die Arbeit rasch vorwärtsschreitet. Eine wohltuende Wärme
durchrieselt den ganzen Körper, macht jedoch mit einem Male einem
heftigen Kältegefühl Platz, das mich jetzt, beim wärmsten
Sonnenschein, nachmittags 3½ Uhr, bereits zwingt, den Überzieher
anzulegen. Jetzt arbeitet auch das Gehirn nicht mehr so scharf und
logisch, besonders bei syntaktischen Feststellungen des schwierigen
Imakuāni, der Sprache der Makua, an die ich mich zum Überfluß
auch noch herangewagt habe. Da halte ich es denn doch allmählich
für angezeigt, meine Temperatur zu messen, der Einfachheit halber
gleich im Sitzen und ruhig weiterarbeitend; 38,6° ist das Ergebnis!
Nun aber hinaus, meine Herren, heißt es im gleichen Augenblick!
Wenige Minuten später steht mein Bett in der Barasa; unmittelbar
darauf liege ich auch schon darin und beginne mich mit heißem
Zitronenwasser innerlich zu behandeln. Drei Stunden später zeigt
das Thermometer gegen 40°; ich lasse mich jetzt, beim Einsetzen
des Abendwindes, mitsamt meinem Bett ins Zelt zurücktragen —
würde ich meinen furchtbar schwitzenden Körper der eisigen
Abendtemperatur aussetzen, so könnte das meinen Tod bedeuten
—, liege dort noch eine kleine Weile und finde dann zu meiner
Beruhigung, daß das Fieber nicht mehr steigt, sondern anfängt
zurückzugehen. Das ist ungefähr 7½ Uhr; als ich kurz nach 8 Uhr
noch einmal messe, ist die Kurve zu meinem maßlosen Erstaunen
auf unter 37° heruntergegangen; mir ist absolut wohl; ich lese noch
ein paar Stunden und könnte sehr wohl rauchen, wenn ich etwas
Ordentliches hätte. Aber Inderzigarren? Pfui Teufel!
Wie ist so etwas denkbar? muß ich mich selbst als Laie fragen.
Das kann doch unmöglich Malaria sein; näher liegt die Vermutung,
daß diese rasch verlaufenden, hohen Fieberanfälle die Folge einer
zu intensiven Sonnenbestrahlung sind, eine Art Insolationsfieber
oder Sonnenstich. Wenn ich mein Fiebernotizbuch nachsehe, wird
mir dies immer wahrscheinlicher, denn regelmäßig treten diese
Anfälle im Anschluß an größere Strapazen und langen Aufenthalt in
praller Sonne ein. Für mich haben diese kurzen Unpäßlichkeiten
wenigstens das Gute, daß sie mich nur stundenweise von der Arbeit
abhalten, denn am nächsten Morgen bin ich regelmäßig wieder
vollkommen frisch und gesund.
Nicht so gut geht es leider meiner Perle von Koch und dem
Knaben Moritz; jener leidet an einer ungeheuren Hydrozele, die ihm
kaum erlaubt aufzustehen, Moritz aber hat Dunkelarrest wegen
seiner entzündeten Augen. Leider versteht Knudsens Koch, ein bis
vor wenig Wochen gänzlich unbeleckter Wilder von irgendwo aus
dem Busch, noch weniger als mein Omari. Folge: Nils Knudsen ist
selbst zum Koch avanciert. Er hat diese seine neue Tätigkeit
sogleich mit einer großen Tat begonnen; da wir nichts Ordentliches
mehr zu essen haben, hat er die vier von Matola erstandenen
Ferkel, hübsch säuberlich in einen großen Tragkorb gepackt, von
Chingulungulu heraufholen lassen und kaltblütig das größte von
ihnen gemordet. Den ersten Schweinebraten haben wir
leichtsinnigerweise doch Knudsens wildem Koch anvertraut; er war
infolgedessen ungenießbar; den Rest des Tieres haben dann wir zu
einem Gelee verarbeitet, das uns nach den langen Wochen der
Unterernährung herrlich mundet und von dem wir mittags und
abends geradezu fabelhafte Portionen vertilgen. Wenn nur nicht die
ewigen Teltower Rübchen dabei wären! O du gesegnete Stadt auf
märkischem Sande, wer hätte je geahnt, daß du so nachhaltig in die
Ernährung eines stillen, deutschen Gelehrten eingreifen würdest!
Dieser boshafte Dr. Jaeger! Er war ein Mann von Zeit und Muße; ihm
halste daher die Landeskundliche Kommission die Besorgung aller
Nahrungsmittel für seine und meine Expedition auf. Feierlich
überweist mir eines schönen Tages in Daressalam der mit der
Verpackung dieser Sachen betraute Handlungsbeflissene meinen
Anteil. Seitdem leide ich unter einer ständigen Rübenfurcht; ich habe
das Gericht an sich ganz gern, aber nur einmal im Jahre, ungern
häufiger. Doch wie ergeht es mir hier? Ich trete an die Kiste heran,
die gerade leergegessen werden muß; der Deckel fliegt hoch; ein
Griff hinein, eine Konservenbüchse kommt zum Vorschein; ein Blick
auf die Etikette: Teltower Rübchen. Puh! Die Dose verschwindet; ein
zweiter Griff; dasselbe Ergebnis; ein dritter, nichts anderes. Nach
langem Suchen erst kommt dann ein anderes Gemüse zutage; oder
auch nicht, denn diese anderen sind allmählich zu Ende gegangen,
nur die Teltower sind geblieben! „Denn helpt dat nich“, sage ich mit
Fritz Reuter; aber zehn Jahre lang esse ich zu Hause keine Teltower
mehr!
Bei all diesem kleinen Leid, das aber nun einmal dazu gehört, um
Afrika schmackhaft zu machen, gibt es wenigstens e i n erfreuliches
Moment: Nils Knudsen hat mit der Geschicklichkeit eines
Feinmechanikers meinen 9 × 12-Apparat wieder in Ordnung
gebracht oder ihn doch wenigstens so weit wieder hergestellt, daß
ich ihn mit einiger List gebrauchen kann. Wie der Mann ohne
Fingernägel mit dieser kniffligen Arbeit hat fertig werden können, bei
der er den ungemein komplizierten Momentverschluß nur mit Hilfe
eines plumpen Schraubenziehers auseinandernehmen und wieder
zusammensetzen mußte, ist mir noch heute schleierhaft, aber er hat
es geschafft. Der Mangel an Fingernägeln hingegen zeigt den guten
Nils von einer Seite, die mit seiner bei der Apparatreparatur
bewiesenen Intelligenz merkwürdig kontrastiert, die andererseits
allerdings auch aufs innigste mit seinem zehnjährigen
Hinterwäldlertum zusammenhängt. Wäscht er da eines Tages in
Lindi irgendeinen Köter. Dieser muß wohl eines schärferen
Reinigungsmittels bedürftig gewesen sein, denn Nils hat ein Gefäß
mitbekommen, dessen Inhalt stark und kräftig riecht. Gewissenhaft
nimmt unser Freund die Reinigung vor, wundert sich ein wenig, daß
sie dem Hunde sehr schlecht bekommt, ist dann aber sehr erstaunt
darüber, daß ihm seine eigenen zehn Fingernägel im Laufe weniger
Tage wegeitern. „Wie kann ich aber auch wissen, daß man
Karbolineum verdünnen muß“, knurrt er oftmals noch jetzt entrüstet,
wenn er seine schrecklich zugerichteten Fingerenden sorgenvoll
mustert!
Weit und breit haben wir die Umgegend durchschweift, seitdem
wir in Newala hausen; zunächst alter Gewohnheit gemäß, sodann
aber, weil der Akide Sefu mit der Zusammenstellung seines
Gelehrtenkollegiums durchaus nicht so rasch fertig geworden ist, wie
er sich zuerst anheischig gemacht hatte. Aber das schadet weiter
nicht, denn auch bloß von außen gesehen, sind Land und Leute
interessant genug.
Das Makondeplateau gleicht einer großen, rechtwinkligen, an
den Ecken abgerundeten Tafel; es ist, vom Indischen Ozean bis
Newala gemessen, etwa 120 Kilometer lang und im Mittel zwischen
dem Lukuledi und dem Rovuma gegen 80 Kilometer breit; es umfaßt
also gegen zwei Drittel der Fläche des Königreichs Sachsen. Nun ist
diese Fläche nicht horizontal, sondern von ihrem Südwestrande
flach, aber ganz gleichmäßig gegen den Ozean hin geneigt. Von der
Schwelle, auf der Newala liegt, kann man viele Meilen über den
Makondebusch nach Osten und Nordosten schauen, ohne einem
Hindernis zu begegnen; es ist ein grünes Meer, aus dem nur hie und
da dichte Rauchwolken in langer Erstreckung emporwirbeln und -
wallen, zum Zeichen dafür, daß auch hier Menschen wohnen und
daß sie ihre Feldkultur ganz nach der Weise so vieler anderer
Naturvölker vorwaltend auf die Verbrennung des
niedergeschlagenen Holzbestandes gründen. Dessen Asche ist
zugleich die einzige Düngung. Selbst am strahlend hellen Tropentag
ist so ein Brand ein großartiges Schauspiel.
Ungleich weniger wirkungsvoll ist der Eindruck, den gegenwärtig
die große Ebene vom Plateaurand aus erweckt. Sooft es mir meine
Zeit gestattet, unternehme ich den kleinen Ausflug an diesen Rand,
bald hierhin, bald dahin, stets in der stillen Hoffnung, endlich einmal
eine klare Luft mit weiter Aussicht vorzufinden; immer aber
vergebens: wohin man dort unten schaut, allerorten steigen
Rauchwolken hoch, der lebhafteste Beweis für die unausgesetzte
Tätigkeit des Waldbrennens; rauchig und dunstig ist auch die ganze
Luft. Schade drum, das Panorama von hier bis weit hinten an die
Madjedjeberge muß unter günstigeren Umständen wirklich großartig
sein. Jetzt haben photographische Aufnahmen eigentlich kaum einen
Zweck, die Profilzeichnung aber gibt nur einen sehr schwachen
Begriff der ganzen Szenerie.
Bei einem dieser Ausflüge habe ich mich absichtlich selbst
einmal am Makondebusch versucht. Der Plateaurand von heute ist
das Ergebnis einer ungeheuer tiefgreifenden Zerstörung durch
Erosion und Abrutschung; überall greifen kurze, aber Hunderte von
Metern tiefe Täler in die Makondeschichten ein. Eine Folge des
lockeren Gefüges dieser Formation ist es, daß nicht nur die
Seitenwände dieser Täler fast senkrecht abstürzen, sondern daß die
Täler auch mit einer ebenso steilen Rückwand enden; dergestalt ist
der Westrand des Makondeplateaus von lauter Talkesseln umsäumt.
Um von einer Seite eines solchen Kessels auf die andere zu
gelangen, habe ich mich eines Tages mit einem Dutzend meiner
Leute durch den Busch geschlagen. Es war eine sehr lichte Stelle,
mit mehr Gras als Buschwuchs; aber welche Mühe hat dieser Weg
von ein paar hundert Metern gekostet, und wie sahen wir alle
nachher aus! Die dünnen Kattunstoffe meiner Leute in Fetzen, sie
selbst aus hundert kleinen Wunden blutend; sogar unsere derben
Khakistoffe hatten den Dornen dieser Vegetationsformation nicht
standgehalten.