Graded Rings and Graded Grothendieck Groups 1st

Edition Roozbeh Hazrat pdf download

https://ebookgate.com/product/graded-rings-and-graded-
grothendieck-groups-1st-edition-roozbeh-hazrat/

Get the full ebook with Bonus Features for a Better Reading Experience on ebookgate.com
Instant digital products (PDF, ePub, MOBI) available
Download now and explore formats that suit you...

Complete Graded Spelling Lists with Spelling and

Vocabulary Exercises Years One and Two British and World
English Kit s Graded Spelling Lists 1st Edition Kit'S
Educational Publishing
https://ebookgate.com/product/complete-graded-spelling-lists-with-
spelling-and-vocabulary-exercises-years-one-and-two-british-and-world-
english-kit-s-graded-spelling-lists-1st-edition-kits-educational-
publishing/
ebookgate.com

Rings Modules Algebras and Abelian Groups 1st Edition

Alberto Facchini

https://ebookgate.com/product/rings-modules-algebras-and-abelian-
groups-1st-edition-alberto-facchini/

ebookgate.com

Graded Chinese Reader 2000 Words Selected Abridged Chinese

Contemporary Short Stories W MP3 English and Chinese
Edition Shi Ji
https://ebookgate.com/product/graded-chinese-reader-2000-words-
selected-abridged-chinese-contemporary-short-stories-w-mp3-english-
and-chinese-edition-shi-ji/
ebookgate.com

A First Course in Abstract Algebra Rings Groups and Fields

Third Edition Marlow Anderson

https://ebookgate.com/product/a-first-course-in-abstract-algebra-
rings-groups-and-fields-third-edition-marlow-anderson/

ebookgate.com
Grothendieck Serre Correspondence Jean-Pierre Serre

https://ebookgate.com/product/grothendieck-serre-correspondence-jean-
pierre-serre/

ebookgate.com

Galois groups and fundamental groups 1st Edition Tamás

Szamuely

https://ebookgate.com/product/galois-groups-and-fundamental-
groups-1st-edition-tamas-szamuely/

ebookgate.com

Rings and Their Modules 1st Edition Paul E. Bland

https://ebookgate.com/product/rings-and-their-modules-1st-edition-
paul-e-bland/

ebookgate.com

Codes and Rings Volume Theory and Practice 1st Edition

Minjia Shi

https://ebookgate.com/product/codes-and-rings-volume-theory-and-
practice-1st-edition-minjia-shi/

ebookgate.com

Near Rings Fuzzy Ideals and Graph Theory 1st Edition

Bhavanari Satyanarayana

https://ebookgate.com/product/near-rings-fuzzy-ideals-and-graph-
theory-1st-edition-bhavanari-satyanarayana/

ebookgate.com
LONDON MATHEMATICAL SOCIETY LECTURE NOTE SERIES
Managing Editor: Professor M. Reid, Mathematics Institute,
University of Warwick, Coventry CV4 7AL, United Kingdom

The titles below are available from booksellers, or from Cambridge University Press at
http://www.cambridge.org/mathematics

312 Foundations of computational mathematics, Minneapolis 2002, F. CUCKER et al (eds)

313 Transcendental aspects of algebraic cycles, S. MÜLLER-STACH & C. PETERS (eds)
314 Spectral generalizations of line graphs, D. CVETKOVIC, P. ROWLINSON & S. SIMIC
315 Structured ring spectra, A. BAKER & B. RICHTER (eds)
316 Linear logic in computer science, T. EHRHARD, P. RUET, J.-Y. GIRARD & P. SCOTT (eds)
317 Advances in elliptic curve cryptography, I.F. BLAKE, G. SEROUSSI & N.P. SMART (eds)
318 Perturbation of the boundary in boundary-value problems of partial differential equations, D. HENRY
319 Double affine Hecke algebras, I. CHEREDNIK
320 L-functions and Galois representations, D. BURNS, K. BUZZARD & J. NEKOVÁŘ (eds)
321 Surveys in modern mathematics, V. PRASOLOV & Y. ILYASHENKO (eds)
322 Recent perspectives in random matrix theory and number theory, F. MEZZADRI & N.C. SNAITH (eds)
323 Poisson geometry, deformation quantisation and group representations, S. GUTT et al (eds)
324 Singularities and computer algebra, C. LOSSEN & G. PFISTER (eds)
325 Lectures on the Ricci flow, P. TOPPING
326 Modular representations of finite groups of Lie type, J.E. HUMPHREYS
327 Surveys in combinatorics 2005, B.S. WEBB (ed)
328 Fundamentals of hyperbolic manifolds, R. CANARY, D. EPSTEIN & A. MARDEN (eds)
329 Spaces of Kleinian groups, Y. MINSKY, M. SAKUMA & C. SERIES (eds)
330 Noncommutative localization in algebra and topology, A. RANICKI (ed)
331 Foundations of computational mathematics, Santander 2005, L.M PARDO, A. PINKUS, E. SÜLI &
M.J. TODD (eds)
332 Handbook of tilting theory, L. ANGELERI HÜGEL, D. HAPPEL & H. KRAUSE (eds)
333 Synthetic differential geometry (2nd Edition), A. KOCK
334 The Navier–Stokes equations, N. RILEY & P. DRAZIN
335 Lectures on the combinatorics of free probability, A. NICA & R. SPEICHER
336 Integral closure of ideals, rings, and modules, I. SWANSON & C. HUNEKE
337 Methods in Banach space theory, J.M.F. CASTILLO & W.B. JOHNSON (eds)
338 Surveys in geometry and number theory, N. YOUNG (ed)
339 Groups St Andrews 2005 I, C.M. CAMPBELL, M.R. QUICK, E.F. ROBERTSON & G.C. SMITH (eds)
340 Groups St Andrews 2005 II, C.M. CAMPBELL, M.R. QUICK, E.F. ROBERTSON & G.C. SMITH (eds)
341 Ranks of elliptic curves and random matrix theory, J.B. CONREY, D.W. FARMER, F. MEZZADRI &
N.C. SNAITH (eds)
342 Elliptic cohomology, H.R. MILLER & D.C. RAVENEL (eds)
343 Algebraic cycles and motives I, J. NAGEL & C. PETERS (eds)
344 Algebraic cycles and motives II, J. NAGEL & C. PETERS (eds)
345 Algebraic and analytic geometry, A. NEEMAN
346 Surveys in combinatorics 2007, A. HILTON & J. TALBOT (eds)
347 Surveys in contemporary mathematics, N. YOUNG & Y. CHOI (eds)
348 Transcendental dynamics and complex analysis, P.J. RIPPON & G.M. STALLARD (eds)
349 Model theory with applications to algebra and analysis I, Z. CHATZIDAKIS, D. MACPHERSON, A. PILLAY
& A. WILKIE (eds)
350 Model theory with applications to algebra and analysis II, Z. CHATZIDAKIS, D. MACPHERSON,
A. PILLAY & A. WILKIE (eds)
351 Finite von Neumann algebras and masas, A.M. SINCLAIR & R.R. SMITH
352 Number theory and polynomials, J. MCKEE & C. SMYTH (eds)
353 Trends in stochastic analysis, J. BLATH, P. MÖRTERS & M. SCHEUTZOW (eds)
354 Groups and analysis, K. TENT (ed)
355 Non-equilibrium statistical mechanics and turbulence, J. CARDY, G. FALKOVICH & K. GAWEDZKI
356 Elliptic curves and big Galois representations, D. DELBOURGO
357 Algebraic theory of differential equations, M.A.H. MACCALLUM & A.V. MIKHAILOV (eds)
358 Geometric and cohomological methods in group theory, M.R. BRIDSON, P.H. KROPHOLLER &
I.J. LEARY (eds)
359 Moduli spaces and vector bundles, L. BRAMBILA-PAZ, S.B. BRADLOW, O. GARCÍA-PRADA &
S. RAMANAN (eds)
360 Zariski geometries, B. ZILBER
361 Words: Notes on verbal width in groups, D. SEGAL
362 Differential tensor algebras and their module categories, R. BAUTISTA, L. SALMERÓN & R. ZUAZUA
363 Foundations of computational mathematics, Hong Kong 2008, F. CUCKER, A. PINKUS & M.J. TODD (eds)
364 Partial differential equations and fluid mechanics, J.C. ROBINSON & J.L. RODRIGO (eds)
365 Surveys in combinatorics 2009, S. HUCZYNSKA, J.D. MITCHELL & C.M. RONEY-DOUGAL (eds)
366 Highly oscillatory problems, B. ENGQUIST, A. FOKAS, E. HAIRER & A. ISERLES (eds)
367 Random matrices: High dimensional phenomena, G. BLOWER
368 Geometry of Riemann surfaces, F.P. GARDINER, G. GONZÁLEZ-DIEZ & C. KOUROUNIOTIS (eds)
369 Epidemics and rumours in complex networks, M. DRAIEF & L. MASSOULIÉ
370 Theory of p-adic distributions, S. ALBEVERIO, A.YU. KHRENNIKOV & V.M. SHELKOVICH
371 Conformal fractals, F. PRZYTYCKI & M. URBANSKI
372 Moonshine: The first quarter century and beyond, J. LEPOWSKY, J. MCKAY & M.P. TUITE (eds)
373 Smoothness, regularity and complete intersection, J. MAJADAS & A. G. RODICIO
374 Geometric analysis of hyperbolic differential equations: An introduction, S. ALINHAC
375 Triangulated categories, T. HOLM, P. JØRGENSEN & R. ROUQUIER (eds)
376 Permutation patterns, S. LINTON, N. RUŠKUC & V. VATTER (eds)
377 An introduction to Galois cohomology and its applications, G. BERHUY
378 Probability and mathematical genetics, N. H. BINGHAM & C. M. GOLDIE (eds)
379 Finite and algorithmic model theory, J. ESPARZA, C. MICHAUX & C. STEINHORN (eds)
380 Real and complex singularities, M. MANOEL, M.C. ROMERO FUSTER & C.T.C WALL (eds)
381 Symmetries and integrability of difference equations, D. LEVI, P. OLVER, Z. THOMOVA &
P. WINTERNITZ (eds)
382 Forcing with random variables and proof complexity, J. KRAJÍČEK
383 Motivic integration and its interactions with model theory and non-Archimedean geometry I, R. CLUCKERS,
J. NICAISE & J. SEBAG (eds)
384 Motivic integration and its interactions with model theory and non-Archimedean geometry II, R. CLUCKERS,
J. NICAISE & J. SEBAG (eds)
385 Entropy of hidden Markov processes and connections to dynamical systems, B. MARCUS, K. PETERSEN &
T. WEISSMAN (eds)
386 Independence-friendly logic, A.L. MANN, G. SANDU & M. SEVENSTER
387 Groups St Andrews 2009 in Bath I, C.M. CAMPBELL et al (eds)
388 Groups St Andrews 2009 in Bath II, C.M. CAMPBELL et al (eds)
389 Random fields on the sphere, D. MARINUCCI & G. PECCATI
390 Localization in periodic potentials, D.E. PELINOVSKY
391 Fusion systems in algebra and topology, M. ASCHBACHER, R. KESSAR & B. OLIVER
392 Surveys in combinatorics 2011, R. CHAPMAN (ed)
393 Non-abelian fundamental groups and Iwasawa theory, J. COATES et al (eds)
394 Variational problems in differential geometry, R. BIELAWSKI, K. HOUSTON & M. SPEIGHT (eds)
395 How groups grow, A. MANN
396 Arithmetic differential operators over the p-adic integers, C.C. RALPH & S.R. SIMANCA
397 Hyperbolic geometry and applications in quantum chaos and cosmology, J. BOLTE & F. STEINER (eds)
398 Mathematical models in contact mechanics, M. SOFONEA & A. MATEI
399 Circuit double cover of graphs, C.-Q. ZHANG
400 Dense sphere packings: a blueprint for formal proofs, T. HALES
401 A double Hall algebra approach to affine quantum Schur–Weyl theory, B. DENG, J. DU & Q. FU
402 Mathematical aspects of fluid mechanics, J.C. ROBINSON, J.L. RODRIGO & W. SADOWSKI (eds)
403 Foundations of computational mathematics, Budapest 2011, F. CUCKER, T. KRICK, A. PINKUS &
A. SZANTO (eds)
404 Operator methods for boundary value problems, S. HASSI, H.S.V. DE SNOO & F.H. SZAFRANIEC (eds)
405 Torsors, étale homotopy and applications to rational points, A.N. SKOROBOGATOV (ed)
406 Appalachian set theory, J. CUMMINGS & E. SCHIMMERLING (eds)
407 The maximal subgroups of the low-dimensional finite classical groups, J.N. BRAY, D.F. HOLT &
C.M. RONEY-DOUGAL
408 Complexity science: the Warwick master’s course, R. BALL, V. KOLOKOLTSOV & R.S. MACKAY (eds)
409 Surveys in combinatorics 2013, S.R. BLACKBURN, S. GERKE & M. WILDON (eds)
410 Representation theory and harmonic analysis of wreath products of finite groups,
T. CECCHERINI-SILBERSTEIN, F. SCARABOTTI & F. TOLLI
411 Moduli spaces, L. BRAMBILA-PAZ, O. GARCÍA-PRADA, P. NEWSTEAD & R.P. THOMAS (eds)
412 Automorphisms and equivalence relations in topological dynamics, D.B. ELLIS & R. ELLIS
413 Optimal transportation, Y. OLLIVIER, H. PAJOT & C. VILLANI (eds)
414 Automorphic forms and Galois representations I, F. DIAMOND, P.L. KASSAEI & M. KIM (eds)
415 Automorphic forms and Galois representations II, F. DIAMOND, P.L. KASSAEI & M. KIM (eds)
416 Reversibility in dynamics and group theory, A.G. O’FARRELL & I. SHORT
417 Recent advances in algebraic geometry, C.D. HACON, M. MUSTAŢĂ & M. POPA (eds)
418 The Bloch–Kato conjecture for the Riemann zeta function, J. COATES, A. RAGHURAM, A. SAIKIA &
R. SUJATHA (eds)
419 The Cauchy problem for non-Lipschitz semi-linear parabolic partial differential equations, J.C. MEYER &
D.J. NEEDHAM
420 Arithmetic and geometry, L. DIEULEFAIT et al (eds)
421 O-minimality and Diophantine geometry, G.O. JONES & A.J. WILKIE (eds)
422 Groups St Andrews 2013, C.M. CAMPBELL et al (eds)
423 Inequalities for graph eigenvalues, Z. STANIĆ
424 Surveys in combinatorics 2015, A. CZUMAJ et al (eds)
425 Geometry, topology and dynamics in negative curvature, C.S. ARAVINDA, F.T. FARRELL &
J.-F. LAFONT (eds)
426 Lectures on the theory of water waves, T. BRIDGES, M. GROVES & D. NICHOLLS (eds)
427 Recent advances in Hodge theory, M. KERR & G. PEARLSTEIN (eds)
428 Geometry in a Fréchet context, C. T. J. DODSON, G. GALANIS & E. VASSILIOU
429 Sheaves and functions modulo p, L. TAELMAN
430 Recent progress in the theory of the Euler and Navier-Stokes equations, J.C. ROBINSON, J.L. RODRIGO,
W. SADOWSKI & A. VIDAL-LÓPEZ (eds)
431 Harmonic and subharmonic function theory on the real hyperbolic ball, M. STOLL
432 Topics in graph automorphisms and reconstruction (2nd Edition), J. LAURI & R SCAPELLATO
433 Regular and irregular holonomic D-modules, M. KASHIWARA & P. SCHAPIRA
434 Analytic semigroups and semilinear initial boundary value problems (2nd Edition), K. TAIRA
435 Graded rings and graded Grothendieck groups, R. HAZRAT
London Mathematical Society Lecture Note Series: 435

Graded Rings and Graded

Grothendieck Groups

RO O Z B E H H A Z R AT
Western Sydney University
 University Printing House, Cambridge CB2 8BS, United Kingdom

Cambridge University Press is part of the University of Cambridge.

It furthers the University’s mission by disseminating knowledge in the pursuit of
education, learning, and research at the highest international levels of excellence.

www.cambridge.org
Information on this title: www.cambridge.org/9781316619582
© Roozbeh Hazrat 2016
This publication is in copyright. Subject to statutory exception
and to the provisions of relevant collective licensing agreements,
no reproduction of any part may take place without the written
permission of Cambridge University Press.
First published 2016
Printed in the United Kingdom by Clays, St Ives plc
A catalogue record for this publication is available from the British Library
Library of Congress Cataloguing in Publication data
Names: Hazrat, Roozbeh, 1971–
Title: Graded rings and graded Grothendieck groups / Roozbeh Hazrat,
University of Western Sydney.
Description: Cambridge : Cambridge University Press, 2016. | Series: London
Mathematical Society lecture note series ; 435 | Includes bibliographical
references and index.
Identifiers: LCCN 2016010216 | ISBN 9781316619582 (pbk. : alk. paper)
Subjects: LCSH: Grothendieck groups. | Graded rings. | Rings (Algebra)
Classification: LCC QA251.5 .H39 2016 | DDC 512/.4–dc23 LC record available
at http://lccn.loc.gov/2016010216
ISBN 978-1-316-61958-2 Paperback
Cambridge University Press has no responsibility for the persistence or accuracy
of URLs for external or third-party Internet Web sites referred to in this publication,
and does not guarantee that any content on such Web sites is, or will remain,
accurate or appropriate.
 Contents

Introduction page 1
1 Graded rings and graded modules 5
1.1 Graded rings 6
1.1.1 Basic definitions and examples 6
1.1.2 Partitioning graded rings 10
1.1.3 Strongly graded rings 14
1.1.4 Crossed products 16
1.1.5 Graded ideals 20
1.1.6 Graded prime and maximal ideals 22
1.1.7 Graded simple rings 23
1.1.8 Graded local rings 25
1.1.9 Graded von Neumann regular rings 26
1.2 Graded modules 28
1.2.1 Basic definitions 28
1.2.2 Shift of modules 28
1.2.3 The Hom groups and category of graded modules 30
1.2.4 Graded free modules 33
1.2.5 Graded bimodules 34
1.2.6 Tensor product of graded modules 35
1.2.7 Forgetting the grading 36
1.2.8 Partitioning graded modules 37
1.2.9 Graded projective modules 41
1.2.10 Graded divisible modules 47
1.3 Grading on matrices 49
1.3.1 Graded calculus on matrices 50
1.3.2 Homogeneous idempotents calculus 58
1.3.3 Graded matrix units 59

v
vi Contents

1.3.4 Mixed shift 60

1.4 Graded division rings 64
1.4.1 The zero component ring of graded simple ring 71
1.5 Strongly graded rings and Dade’s theorem 72
1.5.1 Invertible components of strongly graded rings 78
1.6 Grading on graph algebras 79
1.6.1 Grading on free rings 79
1.6.2 Corner skew Laurent polynomial rings 81
1.6.3 Graphs 84
1.6.4 Leavitt path algebras 85
1.7 The graded IBN and graded type 94
1.8 The graded stable rank 96
1.9 Graded rings with involution 100
2 Graded Morita theory 104
2.1 First instance of the graded Morita equivalence 105
2.2 Graded generators 109
2.3 General graded Morita equivalence 111
3 Graded Grothendieck groups 122
gr
3.1 The graded Grothendieck group K0 124
3.1.1 Group completions 124
gr
3.1.2 K0 -groups 126
gr
3.1.3 K0 of strongly graded rings 128
3.1.4 The reduced graded Grothendieck group 
gr
K0 130
gr
3.1.5 K0 as a Z[Γ]-algebra 132
gr
3.2 K0 from idempotents 132
3.2.1 Stability of idempotents 137
3.2.2 Action of Γ on idempotents 137
gr
3.2.3 K0 is a continuous functor 138
3.2.4 The Hattori–Stallings (Chern) trace map 139
gr
3.3 K0 of graded ∗-rings 141
gr
3.4 Relative K0 -groups 142
gr
3.5 K0 of nonunital rings 145
3.5.1 Graded inner automorphims 148
gr
3.6 K0 is a pre-ordered module 149
3.6.1 Γ-pre-ordered modules 149
gr
3.7 K0 of graded division rings 152
gr
3.8 K0 of graded local rings 158
gr
3.9 K0 of Leavitt path algebras 160
3.9.1 K0 of Leavitt path algebras 161
 Contents vii
gr
3.9.2 Action of Z on K0 of Leavitt path algebras 163
gr
3.9.3 K0 of a Leavitt path algebra via its 0-
component ring 164
gr
3.10 G0 of graded rings 168
gr
3.10.1 G0 of graded Artinian rings 169
gr
3.11 Symbolic dynamics and K0 172
gr
3.12 K1 -theory 177
4 Graded Picard groups 180
4.1 Picgr of a graded commutative ring 181
4.2 Picgr of a graded noncommutative ring 184
gr
5 Graded ultramatricial algebras, classification via K0 192
5.1 Graded matricial algebras 193
gr
5.2 Graded ultramatricial algebras, classification via K0 198
6 Graded versus nongraded (higher) K-theory 205
gr
6.1 K∗ of positively graded rings 206
6.2 The fundamental theorem of K-theory 214
6.2.1 Quillen’s K-theory of exact categories 214
6.2.2 Base change and transfer functors 215
6.2.3 A localisation exact sequence for graded rings 216
6.2.4 The fundamental theorem 217
gr
6.3 Relating K∗ (A) to K∗ (A0 ) 222
gr
6.4 Relating K∗ (A) to K∗ (A) 223
References 227
Index 232
 Introduction

A bird’s eye view of the theory of graded modules over a graded ring might
give the impression that it is nothing but ordinary module theory with all its
statements decorated with the adjective “graded”. Once the grading is consid-
ered to be trivial, the graded theory reduces to the usual module theory. From
this perspective, the theory of graded modules can be considered as an exten-
sion of module theory. However, this simplistic overview might conceal the
point that graded modules come equipped with a shift, thanks to the possibility
of partitioning the structures and then rearranging the partitions. This adds an
extra layer of structure (and complexity) to the theory. This monograph focuses
gr
on the theory of the graded Grothendieck group K0 , that provides a sparkling
illustration of this idea. Whereas the usual K0 is an abelian group, the shift
gr
provides K0 with a natural structure of a Z[Γ]-module, where Γ is the group
used for the grading and Z[Γ] its group ring. As we will see throughout this
book, this extra structure carries substantial information about the graded ring.
Let Γ and Δ be abelian groups and f : Γ → Δ a group homomorphism.
Then for any Γ-graded ring A, one can consider a natural Δ-grading on A
(see §1.1.2); in the same way, any Γ-graded A-module can be viewed as a Δ-
graded A-module. These operations induce functors

U f : GrΓ -A −→ GrΔ -A,

(−)Ω : GrΓ -A −→ GrΩ -AΩ ,

(see §1.2.8), where GrΓ -A is the category of Γ-graded right A-modules, GrΔ -A
that of Δ-graded right A-modules, and GrΩ -A the category of Ω-graded right
AΩ -module with Ω = ker( f ).
One aim of the theory of graded rings is to investigate the ways in which
these categories relate to one another, and which properties of one category
can be lifted to another. In particular, in the two extreme cases when the group

1
2 Introduction

Δ = 0 or f : Γ → Δ is the identity, we obtain the forgetful functors

U : GrΓ -A −→ Mod-A,
(−)0 : GrΓ -A −→ Mod-A0 .

The category PgrΓ-A of graded finitely generated projective A-modules is

an exact category. Thus Quillen’s K-theory machinery [81] defines graded K-
groups
 
Ki (A) := Ki PgrΓ-A ,
gr

for i ∈ N. On the other hand, the shift operation on modules induces a functor
on GrΓ -A that is an auto-equivalence (§1.2.2), so that these K-groups also carry
a Γ-module structure. One can treat the groups Ki (A) and Ki (A0 ) in a similar
way. Quillen’s K-theory machinery allows us to establish relations between
these K-groups. In particular:
gr
Relating K∗ (A) to K∗ (A) for a positively graded rings §6.1. For a Z-graded
ring with the positive support, there is a Z[x, x−1 ]-module isomor-
phism,
Ki (A)  Ki (A0 ) ⊗Z Z[x, x−1 ].
gr

gr
Relating K∗ (A) to K∗ (A0 ) for graded Noetherian regular rings §6.3. Con-
sider the full subcategory Gr0 -A of Gr-A, of all graded modules M
as objects such that M0 = 0. This is a Serre subcategory of Gr-A.
One can show that Gr-A/ Gr0 -A  Mod-A0 . If A is a (right) regular
Noetherian ring, the quotient category identity above holds for the
corresponding graded finitely generated modules, i.e., gr-A/ gr0 -A 
mod-A0 and the localisation theorem gives a long exact sequence of
abelian groups,
δ gr
· · · −→ Kn+1 (A0 ) −→ Kn (gr0 -A) −→ Kn (A) −→ Kn (A0 ) −→ · · · .

gr
Relating K∗ (A) to K∗ (A) for graded Noetherian regular rings §6.4. For
a Z-graded ring A which is right regular Noetherian, there is a long
exact sequence of abelian groups

gr i gr U
· · · −→ Kn+1 (A) −→ Kn (A) −→ Kn (A) −→ Kn (A) −→ · · · .
gr
The main emphasis of this book is on the group K0 as a powerful invariant in
the classification problems. This group is equipped with the extra structure of
the action of the grade group induced by the shift. In many important examples,
 Introduction 3

in fact this shift is all the difference between the graded Grothendieck group
and the usual Grothendieck group, i.e.,
gr
K0 (A)/[P] − [P(1)]  K0 (A),
where P is a graded projective A-module and P(1) is the shifted module (Chap-
ter 6, see Corollary 6.4.2).
The motivation to write this book came from recent activities that adopt the
graded Grothendieck group as an invariant to classify the Leavitt path alge-
bras [47, 48, 79]. Surprisingly, not much is recorded about the graded ver-
sion of the Grothendieck group in the literature, despite the fact that K0 has
been used on many occasions as a crucial invariant, and there is a substan-
tial amount of information about the graded version of other invariants such
as (co)homology groups, Brauer groups, etc. The other surge of interest in
this group stems from the recent activities on the (graded) representation the-
ory of Hecke algebras. In particular for a quiver Hecke algebra, its graded
Grothendieck group is closely related to its corresponding quantised envelop-
ing algebra. For this line of research see the survey [54].
This book tries to fill this gap, by systematically developing the theory of
graded Grothendieck groups. In order to do this, we have to carry over and
work out the details of known results in the nongraded case to the graded set-
ting, and gather together important results on the graded theory scattered in
research papers.
The group K0 has been successfully used in operator theory to classify cer-
tain classes of C ∗ -algebras. Building on work of Bratteli, Elliott in [36] used
the pointed ordered K0 -groups (called dimension groups) as a complete in-
variant for AF C ∗ -algebras. Another cornerstone of using K-groups for the
classifications of a wider range of C ∗ -algebras was the work of Kirchberg and
Phillips [80], who showed that K0 and K1 -groups together are a complete in-
variant for a certain type of C ∗ -algebras. The Grothendieck group considered
as a module induced by a group action was used by Handelman and Ross-
mann [45] to give a complete invariant for the class of direct limits of finite di-
mensional, representable dynamical systems. Krieger [56] introduced (past) di-
mension groups as a complete invariant for the shift equivalence of topological
Markov chains (shift of finite types) in symbolic dynamics. Surprisingly, we
will see that Krieger’s groups are naturally expressed by graded Grothendieck
groups (§3.11).
We develop the theory for rings graded by abelian groups rather than ar-
bitrary groups for two reasons, although most of the results could be carried
over to nonabelian grade groups. One reason is that using the abelian grad-
ing makes the presentation and proofs much more transparent. In addition, in
4 Introduction

most applications of graded K-theory, the ring has an abelian grading (often a
Z-grading).
In Chapter 1 we study the basic theory of graded rings. Chapter 2 con-
gr
centrates on graded Morita theory. In Chapter 3 we compute K0 for certain
graded rings, such as graded local rings and (Leavitt) path algebras. We study
gr
the pre-ordering available on K0 and determine the action of Γ on this group.
Chapter 4 studies graded Picard groups and in Chapter 5 we prove that for
the so-called graded ultramatricial algebras, the graded Grothendieck group is
a complete invariant. Finally, in Chapter 6, we explore the relations between
gr
(higher) Kn and Kn , for the class of Z-graded rings. We describe a general-
isation of the Quillen and van den Bergh theorems. The latter theorem uses
the techniques employed in the proof of the fundamental theorem of K-theory,
where the graded K-theory appears. For this reason we present a proof of the
fundamental theorem in this chapter.

Conventions Throughout this book, unless it is explicitly stated, all rings have
identities, homomorphisms preserve the identity and all modules are unitary.
Moreover, all modules are considered right modules. For a ring A, the category
of right A-modules is denoted by Mod-A. A full subcategory of Mod-A con-
sisted of all finitely generated A-modules is denoted by mod-A. By Pr-A we
denote the category of finitely
 generated projective
 A-modules.
For a set Γ, we write Γ Z or ZΓ to mean γ∈Γ Zγ , where Zγ = Z for each
γ ∈ Γ. We denote the cyclic group Z/nZ with n elements by Zn .

Acknowledgements I would like to acknowledge Australian Research Council

grants DP150101598 and DP160101481. Part of this work was done at the
University of Bielefeld, where the author was a Humboldt Fellow.
I learned about the graded techniques in algebra from Adrian Wadsworth.
Judith Millar worked with me to study the graded K-theory of Azumaya alge-
bras. Gene Abrams was a source of encouragement that the graded techniques
would be fruitful in the study of Leavitt path algebras. Andrew Mathas told
me how graded Grothendieck groups are relevant in representation theory and
pointed me to the relevant literature. The discussions with Zuhong Zhang, who
kindly invited me to the Beijing Institute of Technology in July 2013 and 2014,
helped to improve the presentation. To all of them, I am grateful.
 1
Graded rings and graded modules

Graded rings appear in many circumstances, both in elementary and advanced

areas. Here are two examples.

1 In elementary school when we distribute 10 apples giving 2 apples to each

person, we have

10 Apples : 2 Apples = 5 People.

The psychological problem caused to many kids as to exactly how the word
“People” appears in the equation can be overcome by correcting it to

10 Apples : 2 Apples / People = 5 People.

This shows that already at the level of elementary school arithmetic, children
work in a much more sophisticated structure, i.e., the graded ring

Z[x1±1 , . . . , xn±1 ]

of Laurent polynomial rings! (see the interesting book of Borovik [23, §4.7]
on this).

2 If A is a commutative ring generated by a finite number of elements of degree

1, then by the celebrated work of Serre [85], the category of quasicoherent
sheaves on the scheme is equivalent to QGr-A  Gr-A/ Fdim-A, where Gr-A
is the category of graded modules over A and Fdim-A is the Serre subcate-
gory of (direct limits of) finite dimensional submodules. In particular when
A = K[x0 , x1 , . . . , xn ], where K is a field, then QCoh-Pn is equivalent to
QGr-A[x0 , x1 , . . . , xn ] (see [85, 9, 79] for more precise statements and rela-
tions with noncommutative algebraic geometry).

This book treats graded rings and the category of graded modules over a

5
6 Graded rings and graded modules

graded ring. This category is an abelian category (in fact a Grothendieck cate-
gory). Many of the classical invariants constructed for the category of modules
can be constructed, mutatis mutandis, starting from the category of graded
modules. The general viewpoint of this book is that, once a ring has a natu-
ral graded structure, graded invariants capture more information than the non-
graded counterparts.
In this chapter we give a concise introduction to the theory of graded rings.
We introduce grading on matrices, study graded division rings and introduce
gradings on graph algebras that will be the source of many interesting exam-
ples.

1.1 Graded rings

1.1.1 Basic definitions and examples

A ring A is called a Γ-graded ring, or simply a graded ring, if A = γ∈Γ Aγ ,
where Γ is an (abelian) group, each Aγ is an additive subgroup of A and Aγ Aδ ⊆
Aγ+δ for all γ, δ ∈ Γ.
If A is an algebra over a field K, then A is called a graded algebra if A is a
graded ring and for any γ ∈ Γ, Aγ is a K-vector subspace.

The set Ah = γ∈Γ Aγ is called the set of homogeneous elements of A. The
additive group Aγ is called the γ-component of A and the nonzero elements of
Aγ are called homogeneous of degree γ. We write deg(a) = γ if a ∈ Aγ \{0}. We
call the set
 
ΓA = γ ∈ Γ | Aγ  0

the support of A. We say the Γ-graded ring A has a trivial grading, or A is

concentrated in degree zero, if the support of A is the trivial group, i.e., A0 = A
and Aγ = 0 for γ ∈ Γ\{0}.
For Γ-graded rings A and B, a Γ-graded ring homomorphism f : A → B is
a ring homomorphism such that f (Aγ ) ⊆ Bγ for all γ ∈ Γ. A graded homo-
morphism f is called a graded isomorphism if f is bijective and, when such a
graded isomorphism exists, we write A gr B. Notice that if f is a graded ring
homomorphism which is bijective, then its inverse f −1 is also a graded ring
homomorphism.
If A is a graded ring and R is a commutative graded ring, then A is called
a graded R-algebra if A is an R-algebra and the associated algebra homomor-
phism φ : R → A is a graded homomorphism. When R is a field concentrated
in degree zero, we retrieve the definition of a graded algebra above.
 1.1 Graded rings 7

Proposition 1.1.1 Let A = γ∈Γ Aγ be a Γ-graded ring. Then

(1) 1A is homogeneous of degree 0;

(2) A0 is a subring of A;
(3) each Aγ is an A0 -bimodule;
(4) for an invertible element a ∈ Aγ , its inverse a−1 is homogeneous of degree
−γ, i.e., a−1 ∈ A−γ .
Proof (1) Suppose 1A = γ∈Γ aγ for aγ ∈ Aγ . Let b ∈ Aδ , δ ∈ Γ, be an
arbitrary nonzero homogeneous element. Then b = b1A = γ∈Γ baγ , where
baγ ∈ Aδ+γ for all γ ∈ Γ. Since the decomposition is unique, baγ = 0 for all
γ ∈ Γ with γ  0. But as b was an arbitrary homogeneous element, it follows
that baγ = 0 for all b ∈ A (not necessarily homogeneous), and in particular
1A aγ = aγ = 0 if γ  0. Thus 1A = a0 ∈ A0 .
(2) This follows since A0 is an additive subgroup of A with A0 A0 ⊆ A0 and
1 ∈ A0 .
(3) This is immediate.
(4) Let b = δ∈Γ bδ , with deg(bδ ) = δ, be the inverse of a ∈ Aγ , so that
1 = ab = δ∈Γ abδ , where abδ ∈ Aγ+δ . By (1), since 1 is homogeneous of
degree 0 and the decomposition is unique, it follows that abδ = 0 for all δ  −γ.
Since a is invertible, b−γ  0, so b = b−γ ∈ A−γ as required. 
The ring A0 is called the 0-component ring of A and plays a crucial role in
the theory of graded rings. The proof of Proposition 1.1.1(4), in fact, shows
that if a ∈ Aγ has a left (or right) inverse then that inverse is in A−γ . In The-
orem 1.6.9, we characterise Z-graded rings such that A1 has a left (or right)
invertible element.
Example 1.1.2 Group rings
For a group Γ, the group ring Z[Γ] has a natural Γ-grading
Z[Γ] = Z[Γ]γ , where Z[Γ]γ = Zγ.
γ∈Γ

In §1.1.4, we construct crossed products which are graded rings and are gen-
eralisations of group rings and skew groups rings. A group ring has a natural
involution which makes it an involutary graded ring (see §1.9).
In several applications (such as K-theory of rings, Chapter 6) we deal with
Z-graded rings with support in N, the so called positively graded rings.
Example 1.1.3 Tensor algebras as positively graded rings
Let A be a commutative ring and M be an A-module. Denote by T n (M),
8 Graded rings and graded modules

n ≥ 1, the tensor product of n copies of M over A. Set T 0 (M) = A. Then the

natural A-module isomorphism T n (M) ⊗A T m (M) → T n+m (M), induces a ring
structure on
T (M) := T n (M).
n∈N

The A-algebra T (M) is called the tensor algebra of M. Setting

T (M)n := T n (M)
makes T (M) a Z-graded ring with support N. From the definition, we have
T (M)0 = A.
If M is a free A-module, then T (M) is a free algebra over A, generated by
a basis of M. Thus free rings are Z-graded rings with the generators being
homogeneous elements of degree 1. We will systematically study the grading
of free rings in §1.6.1.
Example 1.1.4 Formal matrix rings as graded rings
Let R and S be rings, M a R−S-bimodule and N a S −R-bimodule. Consider
the set

r m 
T :=  r ∈ R, s ∈ S , m ∈ M, n ∈ N .
n s
Suppose that there are bimodule homomorphisms φ : M ⊗S N → R and
ψ : N ⊗R M → S such that (mn)m = m(nm ), where we denote φ(m, n) =
mn and ψ(n, m) = nm. One can then check that T with matrix addition and
multiplication forms a ring with an identity. The ring T is called the formal
matrix ring and denoted also by
R M
T= .
N S
For example, the Morita ring of a module is a formal matrix ring (see §2.3
and (2.6)).
Considering
R 0 0 M
T0 = , T1 = ,
0 S N 0
it is easy to check that T becomes a Z2 -graded ring. In the cases that the images
of φ and ψ are zero, these rings have been extensively studied (see [57] and
references therein).
When N = 0, the ring T is called a formal triangular matrix ring. In this
case there is no need to consider the homomorphisms φ and ψ. Setting further
T i = 0 for i  0, 1 makes T also a Z-graded ring.
 1.1 Graded rings 9

One specific example of such a grading on (subrings of) formal triangular

matrix rings is used in representation theory. Recall that for a field K, a finite
dimensional K-algebra R is called Frobenius algebra if R  R∗ as right R-
modules, where R∗ := HomK (R, K). Note that R∗ has a natural R-bimodule
structure.
Starting from a finite dimensional K-algebra R, one constructs the trivial ex-
 ∗algebra and has a natural Z-graded structure
tension of R which is a Frobenius
as follows. Consider A := R R , with addition defined component-wise and
multiplication defined as
(r1 , q1 )(r2 , q2 ) = (r1 r2 , r1 q1 + q2 r2 ),
where r1 , r2 ∈ R and q1 , q2 ∈ R∗ . Clearly A is a Frobenius algebra with identity
(1, 0). Moreover, setting
A0 = R ⊕ 0,
A1 = 0 ⊕ R∗ ,
Ai = 0, otherwise,
makes A into a Z-graded ring with support {0, 1}. In fact this ring is a subring
of the formal triangular matrix ring
R R∗
T0 = ,
0 R

a q
consisting of elements .
0 a
These rings appear in representation theory (see [46, §2.2]). The graded ver-
sion of this contraction is carried out in Example 1.2.9.
Example 1.1.5 The graded ring A as A0 -module
Let A be a Γ-graded ring. Then A can be considered as an A0 -bimodule. In
many cases A is a projective A0 -module, for example in the case of group rings
(Example 1.1.2) or when A is a strongly graded ring (see §1.1.3 and Theo-
rem 1.5.12). Here is an example that this is not the case in general. Consider
the formal matrix ring T
R M
T= ,
0 0
where M is a left R-module which is not a projective R-module. Then by Ex-
ample 1.1.4, T is a Z-graded ring with T 0 = R and T 1 = M. Now T as a
T 0 -module is R ⊕ M as an R-module. Since M is not projective, R ⊕ M is not a
projective R-module. We also get that T 1 is not a projective T 0 -module.
10 Graded rings and graded modules

1.1.2 Partitioning graded rings

Let A be a Γ-graded ring and f : Γ → Δ be a group homomorphism.
 Then one
can assign a natural Δ-graded structure to A as follows: A = δ∈Δ A δ , where
⎧
⎪
⎪ −1
⎨ γ∈ f −1 (δ) Aγ if f (δ)  ∅;
Aδ = ⎪
⎪
⎩0 otherwise.
In particular, for a subgroup Ω of Γ we have the following constructions.

Subgroup grading The ring AΩ := ω∈Ω Aω forms a Ω-graded ring. In par-
ticular, A0 corresponds to the trivial subgroup of Γ.
Quotient grading Considering

A= AΩ+α ,
Ω+α∈Γ/Ω

where
AΩ+α := Aω+α ,
ω∈Ω

makes A a Γ/Ω-graded ring. (Note that if Γ is not abelian, then for

this construction, Ω needs to be a normal subgroup.) Notice that with
this grading, A0 = AΩ . If ΓA ⊆ Ω, then A, considered as a Γ/Ω-graded
ring, is concentrated in degree zero.
This construction induces a forgetful functor (or with other inter-
pretations, a block, or a coarsening functor) from the category of Γ-
graded rings RΓ to the category of Γ/Ω-graded rings RΓ/Ω , i.e.,

U : RΓ → RΓ/Ω .

If Ω = Γ, this gives the obvious forgetful functor from the category

of Γ-graded rings to the category of rings. We give a specific example
of this construction in Example 1.1.8 and others in Examples 1.1.20
and 1.6.1.

Example 1.1.6 Tensor product of graded rings

Let A be a Γ-graded and B a Ω-graded ring. Then A ⊗Z B has a natural Γ × Ω-
graded ring structure as follows. Since Aγ and Bω , γ ∈ Γ, ω ∈ Ω, are Z-modules
then A ⊗Z B can be decomposed as a direct sum

A ⊗Z B = Aγ ⊗ Bω
(γ,ω)∈Γ×Ω

(to be precise, Aγ ⊗ Bω is the image of Aγ ⊗Z Bω in A ⊗Z B).

 1.1 Graded rings 11

Now, if Ω = Γ and

f : Γ × Γ −→ Γ,
(γ1 , γ2 ) −→ γ1 + γ2 ,

then we get a natural Γ-graded structure on A ⊗Z B. Namely,

A ⊗Z B = (A ⊗ B)γ ,
γ∈Γ

where
 
(A ⊗ B)γ = ai ⊗ bi | ai ∈ Ah , bi ∈ Bh , deg(ai ) + deg(bi ) = γ .
i

We give specific examples of this construction in Example 1.1.7. One can re-
place Z by a field K, if A and B are K-algebras and Aγ , Bγ are K-modules.

Example 1.1.7 Let A be a ring with identity and Γ be a group. We consider

A as a Γ-graded ring concentrated in degree zero. Then, by Example 1.1.6,

A[Γ]  A ⊗Z Z[Γ]

has a Γ-graded structure, i.e., A[Γ] = γ∈Γ Aγ. If A itself is a (nontrivial) Γ-

graded ring A = γ∈Γ Aγ , then by Example 1.1.6, A[Γ] has also a Γ-grading
 γ γ

A[Γ] = γ∈Γ A , where A = γ=ζ+ζ Aζ ζ . (1.1)

A specific example is when A is a positively graded Z-graded ring. Then

A[x]  A ⊗ Z[x] is a Z-graded ring with support N, where

A[x]n = Ai x j .
i+ j=n

This graded ring will be used in §6.2.4 when we prove the fundamental the-
orem of K-theory. Such constructions are systematically studied in [72] (see
also [75, §6]).

Example 1.1.8 Let A be a Γ × Γ-graded ring. Define a Γ-grading on A as

follows. For γ ∈ Γ, set

Aγ = Aγ−α,α .
α∈Γ

It is easy to see that A = γ∈Γ Aγ is a Γ-graded ring. When A is Z × Z-graded,
then the Z-grading on A is obtained from considering all the homogeneous
components on a diagonal together, as Figure 1.1 shows.
12 Graded rings and graded modules
y

A−2,3 A−1,3 A0,3 A1,3

3

A−2,2 A−1,2 A0,2 A1,2 A2,2

2

A−2,1 A−1,1 A0,1 A1,1 A2,1 A3,1

1

A−1,0 A0,0 A1,0 A2,0 A3,0 A4,0

x
−3 −2 −1 1 2 3
A0,−1 A1,−1 A2,−1 A3,−1
−1

A1,−2 A2,−2 A3,−2

−2
A−1 A0 A1 A2

−3

Figure 1.1

In fact this example follows from the general construction given in §1.1.2.
Consider the homomorphism Γ × Γ → Γ, (α, β) → α + β. Let Ω be the kernel of
this map. Clearly (Γ × Γ)/Ω  Γ. One can check that the (Γ × Γ)/Ω-graded ring
A gives the graded ring constructed in this example (see also Remark 1.1.26).

Example 1.1.9 The direct limit of graded rings

Let Ai , i ∈ I, be a direct system of Γ-graded rings, i.e., I is a directed partially
ordered set and for i ≤ j there is a graded homomorphism φi j : Ai → A j
which is compatible with the ordering. Then A := lim Ai is a Γ-graded ring
−−→
with homogeneous components Aα = lim Aiα . For a detailed construction of
−−→
such direct limits see [24, II, §11.3, Remark 3].
As an example, the ring A = Z[xi | i ∈ N], where A = lim Z[x1 , . . . , xi ],
−−→i∈N
with deg(xi ) = 1 is a Z-graded ring with support N. We give another specific
example of this construction in Example 1.1.10.
We will study in detail one type of these graded rings, i.e., graded ultra-
matricial algebras (Chapter 5, Definition 5.2.1) and will show that the graded
Grothendieck group (Chapter 3) classifies these graded rings completely.
 
Example 1.1.10 Let A = γ∈Γ Aγ and B = Bγ be Γ-graded rings.
γ∈Γ 
Then A × B has a natural grading given by A × B = γ∈Γ (A × B)γ where
(A × B)γ = Aγ × Bγ .
 1.1 Graded rings 13

Example 1.1.11 Localisation of graded rings

Let S be a central multiplicative closed subset of the Γ-graded ring A, con-
sisting of homogeneous elements. Then S −1 A has a natural Γ-graded structure.
Namely, for a ∈ Ah , define deg(a/s) = deg(a) − deg(s) and for γ ∈ Γ,
 
(S −1 A)γ = a/s | a ∈ Ah , deg(a/s) = γ .
It is easy to see that this is well-defined and makes S −1 A a Γ-graded ring.
Many rings have a “canonical” graded structure, among them are crossed
products (group rings, skew group rings, twisted group rings), edge algebras,
path algebras, incidence rings, etc. (see [53] for a review of these ring con-
structions). We will study some of these rings in this book.
Remark 1.1.12 Rings graded by a category
The use of groupoids as a suitable language for structures whose operations
are partially defined has now been firmly recognised. There is a generalised
notion of groupoid graded rings as follows. Recall that a groupoid is a small
category with the property that all morphisms are isomorphisms. As an exam-
ple, let G be a group and I a nonempty set. The set I × G × I, considered as
morphisms, forms a groupoid where the composition is defined by
(i, g, j)( j, h, k) = (i, gh, k).
One can show that this forms a connected groupoid and any connected groupoid
is of this form ([62, Ch. 3.3, Prop. 6]). If I = {1, . . . , n}, we denote I × G × I by
n × G × n.
Let Γ be a groupoid and A be a ring. A is called a Γ-groupoid graded ring
if A = γ∈Γ Aγ , where γ is a morphism of Γ, each Aγ is an additive subgroup
of A and Aγ Aδ ⊆ Aγδ if the morphism γδ is defined and Aγ Aδ = 0 otherwise.
For a group Γ, considering it as a category with one element and Γ as the set
of morphisms, we recover the Γ-group graded ring A (see Example 2.3.1 for an
example of a groupoid graded ring).
One can develop the theory of groupoid graded rings in parallel and similarly
to the group graded rings. See [65, 66] for this approach. Since adjoining a zero
to a groupoid gives a semigroup, a groupoid graded ring is a special case of
rings graded by semigroups (see Remark 1.1.13). For a general notion of a ring
graded by a category see [1, §2], where it is shown that the category of graded
modules (graded by a category) is a Grothendieck category.
14 Graded rings and graded modules

Remark 1.1.13 Rings graded by a semigroup

In the definition of a graded ring (§1.1.1), one can replace the group grad-
ing with a semigroup. With this setting, the tensor algebras of Example 1.1.3
are N-graded rings. A number of results on group graded rings can also be es-
tablished in the more general setting of rings graded by cancellative monoids
or semigroups (see for example [24, II, §11]). However, in this book we only
consider group graded rings.

Remark 1.1.14 Graded rings without identity

For a ring without identity, one defines the concept of the graded ring exactly
as when the ring has an identity. The concept of the strongly graded ring is
defined similarly. In several occasions in this book we construct graded rings
without an identity. For example, Leavitt path algebras arising from infinite
graphs are graded rings without an identity, §1.6.4. See also §1.6.1, the graded
free rings. The unitisation of a (nonunital) graded ring has a canonical grading.
This is studied in relation with graded K0 of nonunital rings in §3.5 (see (3.25)).

1.1.3 Strongly graded rings

Let A be a Γ-graded ring. By Proposition 1.1.1, 1 ∈ A0 . This implies A0 Aγ = Aγ
and Aγ A0 = Aγ for any γ ∈ Γ. If these equalities hold for any two arbitrary
 of Γ, we call the ring a strongly graded ring. Namely, a Γ-graded ring
elements
A= γ∈Γ Aγ is called a strongly graded ring if Aγ Aδ = Aγ+δ for all γ, δ ∈ Γ.
A graded ring A is called crossed product if there is an invertible element in
every homogeneous component Aγ of A; that is, A∗ ∩ Aγ  ∅ for all γ ∈ Γ,
where A∗ is the group of all invertible elements of A. We define the support of
invertible homogeneous elements of A as

Γ∗A = { γ ∈ Γ | A∗γ  ∅ }, (1.2)

where A∗γ := A∗ ∩ Aγ . It is easy to see that Γ∗A is a group and Γ∗A ⊆ ΓA (see
Proposition 1.1.1(4)). Clearly A is a crossed product if and only if Γ∗A = Γ.

Proposition 1.1.15 Let A = γ∈Γ Aγ be a Γ-graded ring. Then

(1) A is strongly graded if and only if 1 ∈ Aγ A−γ for any γ ∈ Γ;

(2) if A is strongly graded then the support of A is Γ;
(3) any crossed product ring is strongly graded;
(4) if f : A → B is a graded homomorphism of graded rings, then B is strongly
graded (resp. crossed product) if A is so.
 1.1 Graded rings 15

Proof (1) If A is strongly graded, then 1 ∈ A0 = Aγ A−γ for any γ ∈ Γ. For the
converse, the assumption 1 ∈ Aγ A−γ implies that A0 = Aγ A−γ for any γ ∈ Γ.
Then for σ, δ ∈ Γ,
Aσ+δ = A0 Aσ+δ = (Aσ A−σ )Aσ+δ = Aσ (A−σ Aσ+δ ) ⊆ Aσ Aδ ⊆ Aσ+δ ,
proving Aσδ = Aσ Aδ , so A is strongly graded.
(2) By (1), 1 ∈ Aγ A−γ for any γ ∈ Γ. This implies Aγ  0 for any γ,
i.e., ΓA = Γ.
(3) Let A be a crossed product ring. By definition, for γ ∈ Γ, there exists
a ∈ A∗ ∩ Aγ . So a−1 ∈ A−γ by Proposition 1.1.1(4) and 1 = aa−1 ∈ Aγ A−γ .
Thus A is strongly graded by (1).
(4) Suppose A is strongly graded. By (1), 1 ∈ Aγ A−γ for any γ ∈ Γ. Thus
1 ∈ f (Aγ ) f (A−γ ) ⊆ Bγ B−γ .
Again (1) implies B is strongly graded. The case of the crossed product follows
easily from the definition. 
The converse of (3) in Proposition 1.1.15 does not hold. One can prove that
if A is strongly graded and A0 is a local ring, then A is a crossed product
algebra (see [75, Theorem 3.3.1]). In §1.6 we give examples of a strongly
graded algebra A such that A is crossed product but A0 is not a local ring. We
also give an example of a strongly Z-graded ring A such that A0 is not local
and A is not crossed product (Example 1.6.22). Using graph algebras we will
produce large classes of strongly graded rings which are not crossed product
(see Theorems 1.6.15 and 1.6.16).
If Γ is a finitely generated group, generated by the set {γ1 , . . . , γn }, then (1)
in Proposition 1.1.15 can be simplified to the following: A is strongly graded
if and only if 1 ∈ Aγi A−γi and 1 ∈ A−γi Aγi , where 1 ≤ i ≤ n. Thus if Γ = Z,
in order for A to be strongly graded, we only need to have 1 ∈ A1 A−1 and
1 ∈ A−1 A1 . This will be used, for example, in Proposition 1.6.6 to show that
certain corner skew Laurent polynomial rings (§1.6.2) are strongly graded.
Example 1.1.16 Constructing strongly graded rings via tensor products
Let A and B be Γ-graded rings. Then by Example 1.1.6, A ⊗Z B is a Γ-graded
ring. If one of the rings is strongly graded (resp. crossed product) then A ⊗Z B
is so. Indeed, suppose A is strongly graded (resp. crossed product). Then the
claim follows from Proposition 1.1.15(4) and the graded homomorphism A →
A ⊗Z B, a → a ⊗ 1.
As a specific case, suppose A is a Z-graded ring. Then
A[x, x−1 ] = A ⊗ Z[x, x−1 ]
16 Graded rings and graded modules

is a strongly graded ring. Notice that with this grading, A[x, x−1 ]0  A.
Example 1.1.17 Strongly graded as a Γ/Ω-graded ring
Let A be a Γ-graded ring. Using Proposition 1.1.15, it is easy to see that if A
is a strongly Γ-graded ring, then it is also a strongly Γ/Ω-graded ring, where Ω
is a subgroup of Γ. However the strongly gradedness is not a “closed” property,
i.e, if A is a strongly Γ/Ω-graded ring and AΩ is a strongly Ω-graded ring, it
does not follow that A is strongly Γ-graded.

1.1.4 Crossed products

Natural examples of strongly graded rings are crossed product algebras (see
Proposition 1.1.15(3)). They cover, as special cases, the skew group rings and
twisted groups rings. We briefly describe the construction here.
Let A be a ring, Γ a group (as usual we use the additive notation), and let
φ : Γ → Aut(A) and ψ : Γ × Γ → A∗ be maps such that for any α, β, γ ∈ Γ and
a ∈ A,
(i) α (β a) = ψ(α, β) α+β a ψ(α, β)−1 ,
(ii) ψ(α, β)ψ(α + β, γ) = α ψ(β, γ) ψ(α, β + γ),
(iii) ψ(α, 0) = ψ(0, α) = 1
Here for α ∈ Γ and a ∈ A, φ(α)(a) is denoted by α a. The map ψ is called a
2-cocycle map. Denote by Aφψ [Γ] the free left A-module with the basis Γ, and
define the multiplication by
(aα)(bβ) = a α b ψ(α, β)(α + β). (1.3)
One can show that with this multiplication, Aφψ [Γ] is a Γ-graded ring with
homogeneous components Aγ, γ ∈ Γ. In fact γ ∈ Aγ is invertible, so Aφψ [Γ] is
a crossed product algebra [75, Proposition 1.4.1].
On the other hand, any crossed product algebra is of this form (see [75,
§1.4]): for any γ ∈ Γ choose uγ ∈ A∗ ∩ Aγ and define φ : Γ → Aut(A0 ) by
φ(γ)(a) = uγ au−1
γ for γ ∈ Γ and a ∈ A0 . Moreover, define the cocycle map

ψ : Γ × Γ −→ A∗0 ,
(ζ, η) −→ uζ uη u−1
ζ+η .

Then

A = A0 φψ [Γ] = γ∈Γ A0 γ,

with multiplication
(aζ)(bη) = aζ bψ(ζ, η)(ζ + η),
 1.1 Graded rings 17

where ζ b is defined as φ(ζ)(b).

Note that when Γ is cyclic, one can choose ui = ui1 for u1 ∈ A∗ ∩ A1 and thus
the cocycle map ψ is trivial, φ is a homomorphism and the crossed product is
a skew group ring. In fact, if Γ = Z, then the skew group ring becomes the
so-called skew Laurent polynomial ring, denoted by A0 [x, x−1 , φ]. Moreover, if
u1 is in the centre of A, then φ is the identity map and the crossed product ring
reduces to the group ring A0 [Γ]. A variant of this construction, namely corner
skew polynomial rings, is studied in §1.6.2.
Skew group rings If ψ : Γ × Γ → A∗ is a trivial map, i.e., ψ(α, β) = 1
for all α, β ∈ Γ, then Conditions (ii) and (iii) trivially hold, and Condition (i)
reduces to α (β a) = α+β a which means that φ : Γ → Aut(A) becomes a group
homomorphism. In this case Aφψ [Γ], denoted by A φ Γ, is a skew group ring
with multiplication
(aα)(bβ) = a α b (α + β). (1.4)

Twisted group ring If φ : Γ → Aut(A) is trivial, i.e., φ(α) = 1A for all

α ∈ Γ, then Condition (i) implies that ψ(α, β) ∈ C(A) ∩ A∗ for any α, β ∈ Γ.
Here C(A) stands for the centre of the ring A. In this case Aφψ [Γ], denoted by
Aψ [Γ], is a twisted group ring with multiplication

(aα)(bβ) = abψ(α, β)(α + β). (1.5)

A well-known theorem in the theory of central simple algebras states that

if D is a central simple F-algebra with a maximal subfield L such that L/F is
a Galois extension and [A : F] = [L : F]2 , then D is a crossed product, with
Γ = Gal(L/F) and A = L (see [35, §12, Theorem 1]).
Some of the graded rings we treat in this book are of the form K[x, x−1 ],
where K is a field. This is an example of a graded field.
A Γ-graded ring A = γ∈Γ Aγ is called a graded division ring if every
nonzero homogeneous element has a multiplicative inverse. If A is also a com-
mutative ring, then A is called a graded field.
Let A be a Γ-graded division ring. It follows from Proposition 1.1.1(4) that
ΓA is a group, so we can write A = γ∈ΓA Aγ . Then, as a ΓA -graded ring, A is
a crossed product and it follows from Proposition 1.1.15(3) that A is strongly
ΓA -graded. Note that if ΓA  Γ, then A is not strongly Γ-graded. Also note that
if A is a graded division ring, then A0 is a division ring.

Remark 1.1.18 Graded division rings and division rings which are graded
Note that a graded division ring and a division ring which is graded are
different. By definition, A is a graded division ring if and only if Ah \{0} is
a group. A simple example is the Laurent polynomial ring D[x, x−1 ], where
18 Graded rings and graded modules

D is a division ring (Example 1.1.19). Other examples show that a graded

division ring does not need to be a domain (Example 1.1.21). However, if the
grade group is totally ordered, then a domain which is also graded has to be
concentrated in degree zero. Thus a division
 ring which is graded by a totally
ordered grade group Γ is of the form A = γ∈Γ Aγ , where A0 is a division ring
and Aγ = 0 for γ  0. This will not be the case if Γ is not totally ordered (see
Example 1.1.20).

In the following we give some concrete examples of graded division rings.

Example 1.1.19 The Veronese subring


Let A = γ∈Γ Aγ be a Γ-graded ring, where Γ is a torsion-free group. For

n ∈ Z\{0}, the nth-Veronese subring of A is defined as A =(n)
γ∈Γ Anγ . This
is a Γ-graded ring with A(n) γ = Anγ . It is easy to see that the support of A
(n)
is
Γ if the support of A is Γ. Note also that if A is strongly graded, so is A(n) .
Clearly A(1) = A and A(−1) is the graded ring with the components “flipped”,
i.e., A(−1)
γ = A−γ . For the case of A(−1) we don’t need to require the grade group
to be torsion-free.
Let D be a division ring and let A = D[x, x−1 ] be the Laurent polynomial
 i∈Z ai x , where ai ∈ D.
i
ring. The elements of A consist of finite sums Then
A is a Z-graded division ring with A = A
i∈Z i , where A i = {ax i
| a ∈
n −n
D}. Consider the nth-Veronese subring A which is the ring D[x , x ]. The
(n)

elements of A(n) consist of finite sums  i∈Z ai x , where ai ∈ D. Then A

in (n)

is a Z-graded division ring, with A = (n)

i∈Z Ain . Here both A and A
(n)
are
strongly graded rings.
There is also another way to consider the Z-graded ring B = D[xn , x−n ]
such that it becomes a graded subring of A = D[x, x−1 ]. Namely, we define
B= i∈Z Bi , where Bi = Dx if i ∈ nZ and Bi = 0 otherwise. This way B is a
i

graded division ring and a graded subring of A. The support of B is clearly the
subgroup nZ of Z. With this definition, B is not strongly graded.

Example 1.1.20 Different gradings on a graded division ring

Let H = R⊕Ri⊕R j⊕Rk be the real quaternion algebra, with multiplication
defined by i2 = −1, j2 = −1 and i j = − ji = k. It is known that H is a
noncommutative division ring with centre R. We give H two different graded
division ring structures, with grade groups Z2 × Z2 and Z2 respectively as
follows.

Z2 × Z2 -grading Let H = R(0,0) ⊕ R(1,0) ⊕ R(0,1) ⊕ R(1,1) , where

R(0,0) = R, R(1,0) = Ri, R(0,1) = R j, R(1,1) = Rk.

 1.1 Graded rings 19

It is routine to check that H forms a strongly Z2 × Z2 -graded division

ring.
Z2 -grading Let H = C0 ⊕C1 , where C0 = R⊕Ri and C1 = C j = R j⊕Rk. One
can check that C0 C0 = C0 , C0 C1 = C1 C0 = C1 and C1 C1 = C0 . This
makes H a strongly Z2 -graded division ring. Note that this grading on
H can be obtained from the first part by considering the quotient grade
group Z2 × Z2 /0 × Z2 (§1.1.2). Quaternion algebras are examples of
Clifford algebras (see Example 1.1.24).
The following generalises the above example of quaternions as a Z2 × Z2 -
graded ring.
Example 1.1.21 Symbol algebras
Let F be a field, ξ be a primitive nth root of unity and let a, b ∈ F ∗ . Let
n−1 n−1
A= F xi y j
i=0 j=0

be the F-algebra generated by the elements x and y, which are subject to the
relations xn = a, yn = b and xy = ξyx. By [35, Theorem 11.1], A is an n2 -
dimensional central simple algebra over F. We will show that A forms a graded
division ring. Clearly A can be written as a direct sum
A= A(i, j) , where A(i, j) = F xi y j
(i, j)∈Zn ⊕Zn

and each A(i, j) is an additive subgroup of A. Using the fact that ξ−k j xk y j = y j xk
for each j, k, with 0 ≤ j, k ≤ n − 1, we can show that
A(i, j) A(k,l) ⊆ A([i+k],[ j+l]) ,
for i, j, k, l ∈ Zn . A nonzero homogeneous element f xi y j ∈ A(i, j) has an inverse
f −1 a−1 b−1 ξ−i j xn−i yn− j ,
proving A is a graded division ring. Clearly the support of A is Zn × Zn , so A
is strongly Zn × Zn -graded.
These examples can also be obtained from graded free rings (see Exam-
ple 1.6.3).
Example 1.1.22 A good counter-example
In the theory of graded rings, in many instances it has been established that if
the grade group Γ is finite (or in some cases, finitely generated), then a graded
property implies the corresponding nongraded property of the ring (i.e., the
20 Graded rings and graded modules

property is preserved under the forgetful functor). For example, one can prove
that if a Z-graded ring is graded Artinian (Noetherian), then the ring is Ar-
tinian (Noetherian). One good example which provides counter-examples to
such phenomena is the following graded field.
Let K be a field and A = K[x1±1 , x2±1 , x3±1 , . . . ] a Laurent polynomial ring
in
countably many variables. This ring is a graded field with its “canonical”
∞ Z-grading and thus it is graded Artinian and Noetherian. However, A is
not Noetherian.

1.1.5 Graded ideals

Let A be a Γ-graded ring. A two-sided ideal I of A is called a graded ideal (or
homogeneous ideal) if
I= (I ∩ Aγ ). (1.6)
γ∈Γ

Thus I is a graded ideal if and only if for any x ∈ I, x = xi , where xi ∈ Ah ,

implies that xi ∈ I.
The notions of a graded subring, a graded left and a graded right ideal are
defined similarly.
Let I be a graded ideal of A. Then the quotient ring A/I forms a graded ring,
with
A/I = (A/I)γ , where (A/I)γ = (Aγ + I)/I. (1.7)
γ∈Γ

With this grading (A/I)0  A0 /I0 , where I0 = A0 ∩ I. From (1.6) it follows

that an ideal I of A is a graded ideal if and only if I is generated as a two-
sided ideal of A by homogeneous elements. Also, for a two-sided ideal I of A,
if (1.7) induces a grading on A/I, then I has to be a graded ideal. By Propo-
sition 1.1.15(4), if A is strongly graded or a crossed product, so is the graded
quotient ring A/I.
Example 1.1.23 Symmetric and exterior algebras as Z-graded rings
Recall from Example 1.1.3 that for a commutative ring A and an A-module
M, the tensor algebra T (M) is a Z-graded ring with support N. The symmetric
algebra S (M) is defined as the quotient of T (M) by the ideal generated by
elements x ⊗ y − y ⊗ x, x, y ∈ M. Since these elements are homogeneous of
degree two, S (M) is a Z-graded commutative ring.

Similarly, the exterior algebra of M, denoted by M, is defined as the quo-
tient of T (M) by the ideal generated by homogeneous elements x ⊗ x, x ∈ M.

So M is a Z-graded ring.
 1.1 Graded rings 21

Let I be a two-sided ideal of a Γ-graded ring A generated by a subset {ai } of

not necessarily homogeneous elements of A. If Ω is a subgroup of Γ such that
{ai } are homogeneous elements in Γ/Ω-graded ring A (see §1.1.2), then clearly
I is a Γ/Ω-graded ideal and consequently A/I is a Γ/Ω-graded ring.

Example 1.1.24 Clifford algebras as Z2 -graded rings

Let V be a F-vector space and q : V → F be a quadratic form with its
associated nondegenerate symmetric bilinear form B : V × V → F.
The Clifford algebra associated with (V, q) is defined as

Cl(V, q) := T (V)/v ⊗ v + q(v).

Considering T (V) as a Z/2Z-graded ring (see §1.1.2), the elements v ⊗ v −

q(v) are homogeneous of degree zero. This induces a Z2 -graded structure on
Cl(V, q). Identifying V with its image in the Clifford algebra Cl(V, q), V lies in
the odd part of the Clifford algebra, i.e., V ⊂ Cl(V, q)1 .
If char(F)  2, as B is nondegenerate, there exist x, y ∈ V such that B(x, y) =
1/2, and thus

xy + yx = 2B(x, y) = 1 ∈ Cl(V, q)1 Cl(V, q)1 .

Similarly, if char(F) = 2, there exist x, y ∈ V such that B(x, y) = 1, so

xy + yx = B(x, y) = 1 ∈ Cl(V, q)1 Cl(V, q)1 .

It follows from Proposition 1.1.15 that Clifford algebras are strongly Z2 -

graded rings.

Recall that for Γ-graded rings A and B, a Γ-graded ring homomorphism f :

A → B is a ring homomorphism such that f (Aγ ) ⊆ Bγ for all γ ∈ Γ. It can easily
be shown that ker( f ) is a graded ideal of A and im( f ) is a graded subring of B.
It is also easy to see that f is injective (surjective/bijective) if and only if for
any γ ∈ Γ, the restriction of f on Aγ is injective (surjective/bijective).
Note that if Γ is an abelian group, then the centre of a graded ring A, C(A), is
a graded subring of A. More generally, the centraliser of a set of homogeneous
elements is a graded subring.

Example 1.1.25 The centre of the graded ring

If a group Γ is not abelian, then the centre of a Γ-graded ring may not be a
graded subring. For example, let Γ = S 3 = {e, a, b, c, d, f } be the symmetric
group of order 3, where

a = (23), b = (13), c = (12), d = (123), f = (132).

22 Graded rings and graded modules

Let A be a ring, and consider the group ring R = A[Γ], which is a Γ-graded ring
by Example 1.1.2. Let x = 1d + 1 f ∈ R, where 1 = 1A , and we note that x is not
homogeneous in R. Then x ∈ Z(R), but the homogeneous components of x are
not in the centre of R. As xis expressed uniquely as the sum of homogeneous
components, we have x  γ∈Γ (Z(R) ∩ Rγ ).
This example can be generalised by taking a nonabelian finite group Γ with
a subgroup Ω which is normal and noncentral. Let A be a ring and consider the
group ring R = A[Γ] as above. Then x = ω∈Ω 1ω is in the centre of R, but the
homogeneous components of x are not all in the centre of R.

Remark 1.1.26 Let Γ and Λ be two groups. Let A be a Γ-graded ring and B be
a Λ-graded ring. Suppose f : A → B is a ring homomorphism and g : Γ → Λ a
group homomorphism such that for any γ ∈ Γ, f (Aγ ) ⊆ Bg(γ) . Then f is called
a Γ−Λ-graded homomorphism. In the case Γ = Λ and g = id, we recover the
usual definition of a Γ-graded homomorphism. For example, if Ω is a subgroup
of Γ, then the identity map 1A : A → A is a Γ−Γ/Ω-graded homomorphism,
where A is considered as Γ and Γ/Ω-graded rings, respectively (see §1.1.2).
Throughout this book, we fix a given group Γ and we work with the Γ-
graded category and all our considerations are within this category. (See Re-
mark 2.3.14 for references to literature where mixed grading is studied.)

1.1.6 Graded prime and maximal ideals

A graded ideal P of Γ-graded ring A is called a graded prime ideal of A if
P  A and for any two graded ideals I, J, I J ⊆ P, implies I ⊆ P or J ⊆ P. If
A is commutative, we obtain the familiar formulation that P is a graded prime
ideal if and only if for x, y ∈ Ah , xy ∈ P implies that x ∈ P or y ∈ P. Note that
a graded prime ideal is not necessarily a prime ideal.
A graded ideal P is called a graded semiprime ideal if for any graded ideal I
in A, I 2 ⊆ P, implies I ⊆ P. A graded ring A is called a graded prime (graded
semiprime) ring if the zero ideal is a graded prime (graded semiprime) ideal.
A graded maximal ideal of a Γ-graded ring A is defined to be a proper graded
ideal of A which is maximal among the set of proper graded ideals of A. Using
Zorn’s lemma, one can show that graded maximal ideals exist, and it is not
difficult to show that a graded maximal ideal is a graded prime. For a graded
commutative ring, a graded ideal is maximal if and only if its quotient ring is a
graded field. There are similar notions of graded maximal left and right ideals.
Parallel to the nongraded setting, for a Γ-graded ring A, the graded Jacobson
radical, J gr (A), is defined as the intersection of all graded left maximal ideals
of A. This coincides with the intersection of all graded right maximal ideals and
 1.1 Graded rings 23

so J gr (A) is a two-sided ideal (see [75, Proposition 2.9.1]). We denote by J(A)

the usual Jacobson radical. It is a theorem of G. Bergman that for a Z-graded
ring A, J(A) is a graded ideal and J(A) ⊆ J gr (A) (see [19]).

1.1.7 Graded simple rings

A nonzero graded ring A is said to be graded simple if the only graded two-
sided ideals of A are {0} and A. The structure of graded simple Artinian rings
are known (see Remark 1.4.8). Following [52] we prove that a graded ring A
is simple if and only if A is graded simple and C(A), the centre of A, is a field.
For a Γ-graded ring A, recall the support ΓA of A, from §1.1.1. For a ∈ A,
writing a = γ∈Γ aγ where aγ ∈ Ah , define the support of a to be
 
Γa = γ | aγ  0 .
We also need the notion of minimal support. A finite set X of Γ is called a
minimal support with respect to an ideal I if X = Γa for 0  a ∈ I and there is
no b ∈ I such that b  0 and Γb  Γa .
We start with a lemma.
Lemma 1.1.27 Let A be a Γ-graded simple ring and I an ideal of A. Let
0  a ∈ I with Γa = {γ1 , . . . , γn }. Then for any α ∈ ΓA , there is a 0  b ∈ I with
Γb ⊆ {γ1 − γn + α, . . . , γn − γn + α}.
Proof Let 0  x ∈ Aα , where α ∈ ΓA and 0  a ∈ I with Γa = {γ1 , . . . , γn }.
Write a = ni=1 aγi , where deg(aγi ) = γi . Since A is graded simple,

x= rl aγn sl , (1.8)
l

where rl , sl ∈ A . Thus there are rk , sk ∈ Ah such that rk aγn sk  0 which implies

h

that b := rk ask ∈ I is not zero. Comparing the degrees in Equation (1.8), it

follows that α = deg(rk ) + deg(sk ) + γn , or deg(rk ) + deg(sk ) = α − γn . So
Γb ⊆ Γa + deg(rk ) + deg(sk ) = { γ1 − γn + α, . . . , γn − γn + α }. 
Theorem 1.1.28 Let A be a Γ-graded ring. Then A is a simple ring if and only
if A is a graded simple ring and C(A) is a field.
Proof One direction is straightforward.
Suppose A is graded simple and C(A) is a field. We will show that A is a
simple ring. Suppose I is a nontrivial ideal of A and 0  a ∈ I with Γa a
minimal support with respect to I. For any x ∈ Ah , with deg(x) = α and γ ∈ Γa ,
we have
Γaxaγ −aγ xa  Γa + (γ + α). (1.9)
24 Graded rings and graded modules

Set b = axaγ − aγ xa ∈ I. Suppose b  0. By (1.9),

Γb  { γ1 + γ + α, . . . , γn + γ + α }.

Applying Lemma 1.1.27 with, say, γn + γ + α ∈ Γb and γn ∈ ΓA , we obtain a

0  c ∈ I such that

Γc ⊆ Γb + (γn − γn − γ − α)  Γa .

This is, however, a contradiction as Γa was a minimal support. Thus b = 0,

i.e., axaγ = aγ xa. It follows that for any γi ∈ Γa

aγi xaγ = aγ xaγi . (1.10)

Consider the R-bimodule map

φ : R = aγi  −→ aγ j  = R,
 
rl aγi sl −→ rl aγ j sl .
l l

To show that φ is well-defined, since φ(t + s) = φ(t) + φ(s), it is enough to show

that if t = 0 then φ(t) = 0, where t ∈ aγi . Suppose l rl aγi sl = 0. Then for
any x ∈ Ah , using (1.10) we have
     
0 = aγ j x rl aγi sl = aγ j xrl aγi sl = aγi xrl aγ j sl = aγi x rl aγ j sl .
l l l l

Since A is graded simple, aγi  = 1. It follows that l rl aγ j sl = 0. Thus φ is

well-defined, injective and also clearly surjective. Then aγ j = φ(aγi ) = aγi φ(1).
But φ(1) ∈ C(A). Thus a = j aγ j = aγi c where c ∈ C(A). But C(A) is a field,
so aγi = ac−1 ∈ I. Again, since R is graded simple, it follows that I = R. This
finishes the proof. 

Remark 1.1.29 If the grade group is not abelian, in order for Theorem 1.1.28
to be valid, the grade group should be hyper-central; A hyper-central group is
a group such that any nontrivial quotient has a nontrivial centre. If A is strongly
graded, and the grade group is torsion-free hyper-central, then A is simple if
and only if A is graded simple and C(A) ⊆ A0 (see [52]).

Remark 1.1.30 Graded simplicity implying simplicity

There are other cases that the graded simplicity of a ring implies that the
ring itself is simple. For example, if a ring is graded by an ordered group (such
as Z), and has a finite support, then graded simplicity implies the simplicity of
the ring [10, Theorem 3].
 1.1 Graded rings 25

1.1.8 Graded local rings

Recall that a ring is a local ring if the set of noninvertible elements form a
two-sided ideal. When A is a commutative ring, then A is local if and only if A
has a unique maximal ideal.
A Γ-graded ring A is called a graded local ring if the two-sided ideal M
generate by noninvertible homogeneous elements is a proper ideal. One can
easily observe that the graded ideal M is the unique graded maximal left, right,
and graded two-sided ideal of A. When A is a graded commutative ring, then
A is graded local if and only if A has a unique graded maximal ideal.
If A is a graded local ring, then the graded ring A/M is a graded division
ring. One can further show that A0 is a local ring with the unique maximal
ideal A0 ∩ M. In fact we have the following proposition.

Proposition 1.1.31 Let A be a Γ-graded ring. Then A is a graded local ring

if and only if A0 is a local ring.

Proof Suppose A is a graded local ring. Then by definition, the two-sided

ideal M generated by noninvertible homogeneous elements is a proper ideal.
Consider m = A0 ∩ M which is a proper ideal of A0 . Suppose x ∈ A0 \m. Then
x is a homogeneous element which is not in M. Thus x has to be invertible in
A and consequently in A0 . This shows that A0 is a local ring with the unique
maximal ideal m.
Conversely, suppose A0 is a local ring. We first show that any left or right
invertible homogeneous element is a two-sided invertible element. Let a be a
left invertible homogeneous element. Then there is a homogeneous element b
such that ba = 1. If ab is not right invertible, then ab ∈ m, where m is the
unique maximal ideal of the local ring A0 . Thus 1 − ab  m which implies that
1 − ab is invertible. But (1 − ab)a = a − aba = a − a = 0, and since 1 − ab
is invertible, we get a = 0 which is a contradiction to the fact that a has a left
inverse. Thus a has a right inverse and so is invertible. A similar argument can
be written for right invertible elements. Now let M be the ideal generated by
all noninvertible homogeneous elements of A. We will show that M is proper,
and thus A is a graded local ring. Suppose M is not proper. Thus 1 = i ri ai si ,
where ai are noninvertible homogeneous elements and ri , si are homogeneous
elements such that deg(ri ai si ) = 0. If ri ai si is invertible for some i, using the
fact that right and left invertibles are invertibles, it follows that ai is invertible,
which is a contradiction. Thus ri ai si , for all i, are homogeneous elements of
degree zero and not invertible. So they are all in m. This implies that 1 ∈ m,
which is a contradiction. Thus M is a proper ideal of A. 
26 Graded rings and graded modules

For more on graded local rings (graded by a cancellative monoid) see [64].
In §3.8 we determine the graded Grothendieck group of these rings.

1.1.9 Graded von Neumann regular rings

The von Neumann regular rings constitute an important class of rings. A unital
ring A is von Neumann regular if for any a ∈ A, we have a ∈ aAa. There are
several equivalent module theoretical definitions, such as A is von Neumann
regular if and only if any module over A is flat. This gives a comparison with
the class of division rings and semisimple rings. A ring is a division ring if and
only if any module is free. A semisimple ring is characterised by the property
that any module is projective. Goodearl’s book [40] is devoted to the class of
von Neumann regular rings. The definition extends to a nonunital ring in an
obvious manner.
If a ring has a graded structure, one defines the graded version of regularity
in a natural way: the graded ring A is called graded von Neumann regular if for
any homogeneous element a ∈ A we have a ∈ aAa. This means, for any homo-
geneous element a ∈ A, one can find a homogeneous element b ∈ A such that
a = aba. As an example, a direct sum of graded division rings is a graded von
Neumann regular ring. Many of the module theoretic properties established
for von Neumann regular rings can be extended to the graded setting; for ex-
ample, A is graded regular if and only if any graded module is (graded) flat.
We refer the reader to [74, C, I.5] for a treatment of such rings and [11, §2.2]
for a concise survey. Several of the graded rings we construct in this book are
graded von Neumann regular, such as Leavitt path algebras (Corollary 1.6.17)
and corner skew Laurent series (Proposition 1.6.8).
In this section, we briefly give some of the properties of graded von Neu-
mann regular rings. The following proposition is the graded version of [40,
Theorem 1.1] which has a similar proof.

Proposition 1.1.32 Let A be a Γ-graded ring. The following statements are

equivalent:

(1) A is a graded von Neumann regular ring;

(2) any finitely generated right (left) graded ideal of A is generated by a ho-
mogeneous idempotent.

Proof (1) ⇒ (2) First we show that any principal graded ideal is generated by
a homogeneous idempotent. So consider the principal ideal xA, where x ∈ Ah .
By the assumption, there is y ∈ Ah such that xyx = x. This immediately implies
xA = xyA. Now note that xy is homogeneous idempotent.
 1.1 Graded rings 27

Next we will prove the claim for graded ideals generated by two elements.
The general case follows by an easy induction. So let xA + yA be a graded ideal
generated by two homogeneous elements x, y. By the previous paragraph, xA =
eA for a homogeneous idempotent e. Note that y−ey ∈ Ah and y−ey ∈ xA+yA.
Thus
xA + yA = eA + (y − ey)A. (1.11)

Again, the previous paragraph gives us a homogeneous idempotent f such that

(y − ey)A = f A. Let g = f − f e ∈ A0 . Notice that e f = 0, which implies that
e and g are orthogonal idempotents. Moreover, f g = g and g f = f . It then
follows that gA = f A = (y − ey)A. Now from (1.11) we get

xA + yA = eA + gA = (e + g)A.

(2) ⇒ (1) Let x ∈ Ah . Then xA = eA for some homogeneous idempotent e.

Thus x = ea and e = xy for some a, y ∈ Ah . Then x = ea = eea = ex = xyx. 

Proposition 1.1.33 Let A be a Γ-graded von Neumann regular ring. Then

(1) any graded right (left) ideal of A is idempotent;

(2) any graded ideal is graded semiprime;
(3) any finitely generated right (left) graded ideal of A is a projective module.

Moreover, if A is a Z-graded regular ring then

(4) J(A) = J gr (A) = 0.

Proof The proofs of (1)–(3) are similar to the nongraded case [40, Corol-
lary 1.2]. We provide the easy proofs here.
(1) Let I be a graded right ideal. For any homogeneous element x ∈ I there
is y ∈ Ah such that x = xyx. Thus x = (xy)x ∈ I 2 . It follows that I 2 = I.
(2) This follows immediately from (1).
(3) By Proposition 1.1.32, any finitely generated right ideal is generated by
a homogeneous idempotent. However, this latter ideal is a direct summand of
the ring, and so is a projective module.
(4) By Bergman’s observation, for a Z-graded ring A, J(A) is a graded ideal
and J(A) ⊆ J gr (A) (see [19]). By Proposition 1.1.32, J gr (A) contains an idem-
potent, which then forces J gr (A) = 0. 

If the graded ring A is strongly graded then one can show that there is a
one-to-one correspondence between the right ideals of A0 and the graded right
ideals of A (similarly for the left ideals) (see Remark 1.5.6). This is always the
case for the graded regular rings as the following proposition shows.
28 Graded rings and graded modules

Proposition 1.1.34 Let A be a Γ-graded von Neumann regular ring. Then

there is a one-to-one correspondence between the right (left) ideals of A0 and
the graded right (left) ideals of A.

Proof Consider the following correspondences between the graded right ide-
als of A and the right ideals of A0 . For a graded right ideal I of A assign I0 in
A0 and for a right ideal J in A0 assign the graded right ideal JA in A. Note that
(JA)0 = J. We show that I0 A = I. It is enough to show that any homogeneous
element a of I belongs to I0 A. Since A is graded regular, axa = a for some
x ∈ Ah . But ax ∈ I0 and thus a = axa ∈ I0 A. A similar proof gives the left ideal
correspondence. 

In Theorem 1.2.20 we give yet another characterisation of graded von Neu-

mann regular rings based on the concept of divisible modules.
Later, in Corollary 1.5.10, we show that if A is a strongly graded ring, then
A is graded von Neumann regular if and only if A0 is a von Neumann regular
ring. The proof uses the equivalence of suitable categories over the rings A and
A0 . An element-wise proof of this fact can also be found in [96, Theorem 3].

1.2 Graded modules

1.2.1 Basic definitions
Let A be a Γ-graded ring. A graded right A-module 
M is defined to be a right
A-module M with a direct sum decomposition M = γ∈Γ Mγ , where each Mγ
is an additive subgroup of M such that Mλ Aγ ⊆ Mλ+γ for all γ, λ ∈ Γ.
For Γ-graded right A-modules M and N, a Γ-graded module homomorphism
f : M → N is a module homomorphism such that f (Mγ ) ⊆ Nγ for all γ ∈ Γ.
A graded homomorphism f is called a graded module isomorphism if f is
bijective and, when such a graded isomorphism exists, we write M gr N.
Notice that if f is a graded module homomorphism which is bijective, then its
inverse f −1 is also a graded module homomorphism.

1.2.2 Shift of modules

Let M be a graded right A-module. For δ ∈ Γ, we define the δ-suspended or
δ-shifted graded right A-module M(δ) as

M(δ) = M(δ)γ , where M(δ)γ = Mδ+γ .

γ∈Γ
 1.2 Graded modules 29

This shift plays a pivotal role in the theory of graded rings. For example, if M
is a Z-graded A-module, then the following table shows how the shift like “the
tick of the clock” moves the homogeneous components of M to the left.

degrees -3 -2 -1 0 1 2 3
M M−1 M0 M1 M2
M(1) M−1 M0 M1 M2
M(2) M−1 M0 M1 M2

Let M be a Γ-graded right A-module. A submodule N of M is called a graded

submodule if
N= (N ∩ Mγ ).
γ∈Γ

Example 1.2.1 aA as a graded ideal and a graded module

Let A be a Γ-graded ring and a ∈ A a homogeneous element of degree α.
Then aA is a graded right A-module with γ ∈ Γ homogeneous component
defined as
(aA)γ := aAγ−α ⊆ Aγ .

With this grading aA is a graded submodule (and graded right ideal) of A. Thus
for β ∈ Γ, a is a homogenous element of the graded A-module aA(β) of degree
α−β. This will be used throughout the book, for example in Proposition 1.2.19.
However, note that defining the grading on aA as

(aA)γ := aAγ ⊆ Aγ+α

makes aA a graded submodule of A(α), which is the image of the graded ho-
momorphism A → A(α), r → ar.

There are similar notions of graded left and graded bi-submodules (§1.2.5).
When N is a graded submodule of M, the factor module M/N forms a graded
A-module, with

M/N = (M/N)γ , where (M/N)γ = (Mγ + N)/N. (1.12)

γ∈Γ

Example 1.2.2 Let A be a Γ-graded ring. Define a grading on the matrix

ring Mn (A) as follows. For α ∈ Γ, Mn (A)α = Mn (Aα ) (for a general theory of
grading on matrix rings see §1.3). Let eii ∈ Mn (A), 1 ≤ i ≤ n, be a matrix unit,
30 Graded rings and graded modules

i.e., a matrix with 1 in the (i, i) position and zero everywhere else, and consider
eii Mn (A). By Example 1.2.1, eii Mn (A) is a graded right Mn (A)-module and
n
eii Mn (A) = Mn (A).
i=1

This shows that the graded module eii Mn (A) is a projective module. This is an
example of a graded projective module (see §1.2.9).
Example 1.2.3 Let A be a commutative ring. Consider the matrix ring Mn (A)
as a Z-graded ring concentrated in degree zero. Moreover, consider Mn (A) as a
graded Mn (A)-module with the grading defined as follows: Mn (A)i = eii Mn (A)
for 1 ≤ i ≤ n and zero otherwise. Note that all nonzero homogeneous elements
of this module are zero-divisors, and thus can’t constitute a linear independent
set. We will use this example to show that a free module which is graded is not
necessarily a graded free module (§1.2.4).
Example 1.2.4 Modules with no shift
It is easy to construct modules whose shifts don’t produce new (nonisomor-
phic) graded modules. Let M be a graded A-module and consider
N= M(γ).
γ∈Γ

We show that N gr N(α) for any α ∈ Γ. Define the map fα : N → N(α) on
homogeneous components as follows and extend it to N,
Nβ = Mγ+β −→ Mγ+α+β = N(α)β
γ∈Γ γ∈Γ

{mγ } −→ {mγ },
where mγ = mγ+α (i.e., shift the sequence α “steps”). It is routine to see that this
gives a graded A-module homomorphism with inverse homomorphism f−α . For
another example, see Corollary 1.3.18.

1.2.3 The Hom groups and the category of graded modules

For graded right A-modules M and N, a graded A-module homomorphism of
degree δ is an A-module homomorphism f : M → N, such that
f (Mγ ) ⊆ Nγ+δ
for any γ ∈ Γ. Let HomA (M, N)δ denote the subgroup of HomA (M, N) consist-
ing of all graded A-module homomorphisms of degree δ, i.e.,
HomA (M, N)δ = { f ∈ HomA (M, N) | f (Mγ ) ⊆ Nγ+δ , γ ∈ Γ}. (1.13)
 1.2 Graded modules 31

For graded A-modules, M, N and P, under the composition of functions, we

then have
HomA (N, P)γ × HomA (M, N)δ −→ Hom(M, P)γ+δ . (1.14)
Clearly a graded module homomorphism defined in §1.2.1 is a graded ho-
momorphism of degree 0.
By Gr-A (or GrΓ -A to emphasise the grade group of A), we denote a cate-
gory that consists of Γ-graded right A-modules as objects and graded homo-
morphisms as the morphisms. Similarly, A-Gr denotes the category of graded
left A-modules. Thus
HomGr-A (M, N) = HomA (M, N)0 .
Moreover, for α ∈ Γ, as a set of functions, one can write
   
HomGr-A M(−α), N = HomGr-A M, N(α) = HomA (M, N)α . (1.15)
A full subcategory of Gr-A consisted of all graded finitely generated A-
modules is denoted by gr-A.
For α ∈ Γ, the α-suspension functor or shift functor

Tα : Gr-A −→ Gr-A, (1.16)

M −→ M(α),
is an isomorphism with the property Tα Tβ = Tα+β , where α, β ∈ Γ.
Remark 1.2.5 Let A be a Γ-graded ring and Ω be a subgroup of Γ such that
ΓA ⊆ Ω ⊆ Γ. Then the ring A can be considered naturally as a Ω-graded
ring. Similarly, if A, B are Γ-graded rings and f : A → B is a Γ-graded homo-
morphism and ΓA , ΓB ⊆ Ω ⊆ Γ, then the homomorphism f can be naturally
considered as a Ω-graded homomorphism. In this case, to make a distinction,
we write GrΓ -A for the category of Γ-graded A-modules and GrΩ -A for the
category of Ω-graded A-modules.
Theorem 1.2.6 For graded right A-modules M and N, such that M is finitely
generated, the abelian group HomA (M, N) has a natural decomposition
HomA (M, N) = HomA (M, N)γ . (1.17)
γ∈Γ

Moreover, the endomorphism ring HomA (M, M) is Γ-graded.

Proof Let f ∈ HomA (M, N) and λ ∈ Γ. Define a map fλ : M → N as follows:
for m ∈ M,

fλ (m) = f (mγ−λ )γ , (1.18)
γ∈Γ
32 Graded rings and graded modules

where m = γ∈Γ mγ . One can check that fλ ∈ HomA (M, N).

Now let m ∈ Mα , α ∈ Γ. Then (1.18) reduces to
fλ (m) = f (m)α+λ ⊆ Mα+λ .
This shows that fλ ∈ HomA (M, N)λ . Moreover, fλ (m) is zero for all but a finite
number of λ ∈ Γ and
 
fλ (m) = f (m)α+λ = f (m).
λ λ

Now since M is finitely generated, there are a finite number of homogeneous

elements which generate any element m ∈ M. The above argument shows that
only a finite number of the fλ (m) are nonzero and f = λ fλ . This in turn
shows that

HomA (M, N) = HomA (M, N)γ .
γ∈Γ

Finally, it is easy to see that HomA (M, N)γ , γ ∈ Γ constitutes a direct sum.
For the second part, replacing N by M in (1.17), we get
HomA (M, M) = HomA (M, M)γ .
γ∈Γ

Moreover, by (1.14) if f ∈ HomA (M, M)γ and g ∈ HomA (M, M)λ then
f g ∈ HomA (M, M)γ+λ .
This shows that when M is finitely generated HomA (M, M) is a Γ-graded ring.

Let M be a graded finitely generated right A-module. Then the usual dual of
M, i.e., M ∗ = HomA (M, A), is a left A-module. Moreover, using Theorem 1.2.6,
one can check that M ∗ is a graded left A-module. Since
HomA (M, N)(α) = HomA (M(−α), N) = HomA (M, N(α)),
we have
M(α)∗ = M ∗ (−α). (1.19)
This should also make sense: the dual of “pushing forward” M by α, is the
same as “pulling back” the dual M ∗ by α.
Note that although HomA (M, N) is defined in the category Mod-A, the graded
structures of M and N are intrinsic in the grading defined on HomA (M, N).
Thus if M is isomorphic to N as a nongraded A-module, then EndA (M) is not
necessarily graded isomorphic to EndA (N). However if M gr N(α), α ∈ Γ,
then one can observe that EndA (M) gr EndA (N) as graded rings.
 1.2 Graded modules 33

When M is a free module, HomA (M, M) can be represented as a matrix ring

over A. Next we define graded free modules. In §1.3 we will see that if M is
a graded free module, the graded ring HomA (M, M) can be represented as a
matrix ring over A with a very concrete grading.

Example 1.2.7 The Veronese submodule

For a Γ-graded ring A, recall the construction of nth-Veronese subring

A(n) = Anγ
γ∈Γ

(Example 1.1.19). In a similar fashion, for a graded A-module M and n ∈ Z,

define the nth-Veronese module of M as

M (n) = Mnγ .
γ∈Γ

This is a Γ-graded A(n) -module. Clearly there is a natural “forgetful” functor

U : Gr-A −→ Gr-A(n) ,

which commutes with suspension functors as follows Tα U = UTnα , i.e.,

M (n) (α) = M(nα)(n) ,

for α ∈ Γ and n ∈ Z (see §1.2.7 for more on forgetful functors).

1.2.4 Graded free modules

A Γ-graded (right) A-module F is called a graded free A-module if F is a free
right A-module with a homogeneous base. Clearly a graded free module is a
free module but the converse is not correct, i.e., a free module which is graded
is not necessarily a graded free module. As an example, for A = R[x] consid-
ered as a Z-graded ring, A ⊕ A(1) is not a graded free A ⊕ A-module, whereas
A ⊕ A is a free A ⊕ A-module (see also Example 1.2.3). The definition of free
given here is consistent with the categorical definition of free objects over a set
of homogeneous elements in the category
 of graded modules ([50, I, §7]).
Consider a Γ-graded A-module i∈I A(δ i ), where I is an indexing set and
δi ∈ Γ. Note that for each i ∈ I, the element ei of the standard basis (i.e., 1 in
the ith component and  zero elsewhere) is homogeneous of degree −δi . The set
{ei }i∈I forms a base for i∈I A(δi ), which by definition makes this a graded free
A-module. On the other hand, a graded free A-module F with  a homogeneous
base {bi }i∈I , where deg(bi ) = −δi is graded isomorphic to i∈I A(δi ). Indeed
34 Graded rings and graded modules

one can easily observe that the map induced by

ϕ: A(δi ) −→ F (1.20)
i∈I
ei −→ bi
is a graded A-module isomorphism.
If the indexing set I is finite, say I = {1, . . . , n}, then
A(δi ) = A(δ1 ) ⊕ · · · ⊕ A(δn ),
i∈I

is also denoted by An (δ1 , . . . , δn ) or An (δ), where δ = (δ1 , . . . , δn ).

In §1.3.4, we consider the situation when the graded free right A-modules
An (δ) and Am (α), where δ = (δ1 , . . . , δn ) and α = (α1 , . . . , αm ), are isomorphic.
In §1.7, we will also consider the concept of graded rings with the graded
invariant basis numbers.

1.2.5 Graded bimodules

The notion of graded left A-modules is developed similarly. The category of
graded left A-modules with graded homomorphisms is denoted by A-Gr. In a
similar manner for Γ-graded rings A and B, we can consider the graded
 A− B-
bimodule M. That is, M is a A−B-bimodule and additionally M = γ∈Γ Mγ is
a graded left A-module and a graded right B-module, i.e.,
Aα Mγ Bβ ⊆ Mα+γ+β ,
where α, γ, β ∈ Γ. The category of graded A-bimodules is denoted by Gr-A-Gr.

Remark 1.2.8 Shift of nonabelian group graded modules

If the grade group Γ is not abelian, then in order that the shift of components
matches, for a graded left A-module M one needs to define
M(δ)γ = Mγδ ,
whereas for the graded right M-module A, shift is defined by
M(δ)γ = Mδγ .
With these definitions, for Tα , Tβ : Gr-A → Gr-A, we have Tα Tβ = Tβα ,
whereas for Tα , Tβ : A-Gr → A-Gr, we have Tα Tβ = Tαβ . For this reason,
in the nonabelian grade group setting, several books choose to work with the
graded left modules as opposed to the graded right modules we have adopted
in this book.
 1.2 Graded modules 35

1.2.6 Tensor product of graded modules

Let A be a Γ-graded ring and Mi , i ∈ I, be a direct system of Γ-graded A-
modules, i.e., I is a directed partially ordered set and for i ≤ j, there is a graded
A-homomorphism φi j : Mi → M j which is compatible with the ordering.
Then M := lim Mi is a Γ-graded A-module with homogeneous components
−−→
Mα = lim Miα (see Example 1.1.9 for the similar construction for rings).
−−→ 
In particular, let {Mi | i ∈ I} be Γ-graded
 right A-modules.
 Then i∈I Mi has
a natural graded A-module given by ( i∈I Mi )α = i∈I Miα , α ∈ Γ.
Let M be a graded right A-module and N be a graded left A-module. We
will observe that the tensor product M ⊗A N has a natural Γ-graded Z-module
structure. Since each of Mγ , γ ∈ Γ, is a right A0 -module and similarly Nγ ,
γ ∈ Γ, is a left A0 -module, then M ⊗A0 N can be decomposed as a direct sum

M ⊗ A0 N = (M ⊗ N)γ ,
γ∈Γ

where
 
(M ⊗ N)γ = mi ⊗ ni | mi ∈ M h , ni ∈ N h , deg(mi ) + deg(ni ) = γ .
i

Now note that M ⊗A N  (M ⊗A0 N)/J, where J is a subgroup of M ⊗A0 N

generated by the homogeneous elements

{ma ⊗ n − m ⊗ an | m ∈ M h , n ∈ N h , a ∈ Ah }.

This shows that M ⊗A N is also a graded module. It is easy to check that, for
example, if N is a graded A-bimodule, then M ⊗A N is a graded right A-module.
It follows from the definition that

M ⊗ N(α) = M(α) ⊗ N = (M ⊗ N)(α). (1.21)

Observe that for a graded right A-module M, the map

M ⊗A A(α) −→ M(α), (1.22)

m ⊗ a −→ ma,

is a graded isomorphism. In particular, for any α, β ∈ Γ, there is a graded

A-bimodule isomorphism

A(α) ⊗A A(β) gr A(α + β). (1.23)

Example 1.2.9 Graded formal matrix rings

The construction of formal matrix rings (Example 1.1.4) can be carried over
36 Graded rings and graded modules

to the graded setting as follows. Let R and S be Γ-graded rings, M be a graded

R−S-bimodule and N be a graded S−R-bimodule. Suppose that there are graded
bimodule homomorphisms φ : M ⊗S N → R and ψ : N ⊗R M → S such that
(mn)m = n(nm ), where we denote φ(m, n) = mn and ψ(n, m) = nm. Consider
the ring
R M
T= ,
N S
and define, for any γ ∈ Γ,
Rγ Mγ
Tγ = .
Nγ Sγ
One checks that T is a Γ-graded ring, called a graded formal matrix ring. One
specific type of such rings is a Morita ring which appears in graded Morita
theory (§2.3).

1.2.7 Forgetting the grading

Most forgetful functors in algebra tend to have left adjoints, which have a
“free” construction. One such example is the forgetful functor from the cat-
egory of abelian groups to abelian monoids that we will study in Chapter 3 in
relation to Grothendieck groups. However, some of the forgetful functors in
the graded setting naturally have right adjoints, as we will see below.
Consider the forgetful functor

U : Gr-A −→ Mod-A, (1.24)

which simply assigns to any graded module M in Gr-A its underlying module
M in Mod-A, ignoring the grading. Similarly, the graded homomorphisms are
sent to the same homomorphisms, disregarding their graded compatibilities.
There is a functor F : Mod-A → Gr-A which is a right adjoint to U. The
construction
is as follows: let M be an A-module. Consider the abelian group
F(M) := γ∈Γ Mγ , where Mγ is a copy of M. Moreover, for a ∈ Aα and
m ∈ Mγ define m.a = ma ∈ Mα+γ . This defines a graded A-module structure
on F(M) and makes F an exact functor from Mod-A to Gr-A. One can prove
that for any M ∈ Gr-A and N ∈ Mod-A, we have a bijective map
  φ  
HomMod-A U(M), N −→ HomGr-A M, F(N) ,
f −→ φ f ,

where φ f (mα ) = f (mα ) ∈ Nα .

Another Random Scribd Document
with Unrelated Content
agreeable manners, he combined recklessness in false promises,
deceit and extortion even towards allies, and unscrupulous perjury
when it suited his purpose—this we find affirmed, and there is no
reason for disbelieving it.[1164] Such dissolving forces smoothed the
way for an efficient and admirable army, organized, and usually
commanded, by himself. Its organization adopted and enlarged the
best processes of scientific warfare employed by Epaminondas and
Iphikrates.[1165] Begun as well as completed by Philip, and
bequeathed as an engine ready-made for the conquests of
Alexander, it constitutes an epoch in military history. But the more
we extol the genius of Philip as a conqueror, formed for successful
encroachment and aggrandizement at the expense of all his
neighbors—the less can we find room for that mildness and
moderation which some authors discover in his character. If, on
some occasions of his life, such attributes may fairly be recognized,
we have to set against them the destruction of the thirty-two Greek
cities in Chalkidikê and the wholesale transportation of reluctant and
miserable families from one inhabitancy to another.
Besides his skill as a general and a politician, Philip was no mean
proficient in the Grecian accomplishments of rhetoric and letters.
The testimony of Æschines as to his effective powers of speaking,
though requiring some allowance, is not to be rejected. Isokrates
addresses him as a friend of letters and philosophy; a reputation
which his choice of Aristotle as instructor of his son Alexander, tends
to bear out. Yet in Philip, as in the two Dionysii of Syracuse and
other despots, these tastes were not found inconsistent either with
the crimes of ambition, or the licenses of inordinate appetite. The
contemporary historian Theopompus, a warm admirer of Philip’s
genius, stigmatizes not only the perfidy, of his public dealings, but
also the drunkenness, gambling, and excesses of all kinds in which
he indulged—encouraging the like in those around him. His
Macedonian and Grecian body-guard, eight hundred in number, was
a troop in which no decent man could live; distinguished indeed for
military bravery and aptitude, but sated with plunder, and stained
with such shameless treachery, sanguinary rapacity, and unbridled
lust, as befitted only Centaurs and Læstrygons.[1166] The number of
Philip’s mistresses and wives was almost on an Oriental scale;[1167]
and the dissensions thus introduced into his court through his
offspring by different mothers, were fraught with mischievous
consequences.
In appreciating the genius of Philip, we have to appreciate also
the parties to whom he stood opposed. His good fortune was
nowhere more conspicuous than in the fact, that he fell upon those
days of disunion and backwardness in Greece (indicated in the last
sentence of Xenophon’s Hellenica) when there was neither leading
city prepared to keep watch, nor leading general to take command,
nor citizen-soldiers willing and ready to endure the hardships of
steady service. Philip combated no opponents like Epaminondas, or
Agesilaus, or Iphikrates. How different might have been his career,
had Epaminondas survived the victory of Mantineia, gained only two
years before Philip’s accession! To oppose Philip, there needed a
man like himself, competent not only to advise and project, but to
command in person, to stimulate the zeal of citizen-soldiers, and to
set the example of braving danger and fatigue. Unfortunately for
Greece, no such leader stood forward. In counsel and speech
Demosthenes sufficed for the emergency. Twice before the battle of
Chæroneia—at Byzantium and at Thebes—did he signally frustrate
Philip’s combinations. But he was not formed to take the lead in
action, nor was there any one near him to supply the defect. In the
field, Philip encountered only that “public inefficiency,” at Athens and
elsewhere in Greece, of which even Æschines complains;[1168] and to
this decay of Grecian energy, not less than to his own distinguished
attributes, the unparalleled success of his reign was owing. We shall
find during the reign of his son Alexander (to be described in our
next volume) the like genius and vigor exhibited on a still larger
scale, and achieving still more wonderful results; while the once
stirring politics of Greece, after one feeble effort, sink yet lower, into
the nullity of a subject-province.
 FOOTNOTES
[1] Diodor. xiii. 86-114; xiv. 70; xv. 24. Another pestilence is alluded to by
Diodorus in 368 B. C. (Diodor. xv. 73).
Movers notices the intense and frequent sufferings of the ancient Phœnicians,
in their own country, from pestilence; and the fearful expiations to which these
sufferings gave rise (Die Phönizier, vol. ii. part ii. p. 9).

[2] Diodor. xiv. 78.

[3] Diodor. xiv. 78. Διονύσιος δ᾽ εἰς Μεσσήνην κατῴκισε χιλίους μὲν Λοκροὺς,
τετρακισχιλίους δὲ Μ ε δ ι μ ν α ί ο υ ς, ἑξακοσίους δὲ τῶν ἐκ Πελοποννήσου
Μεσσηνίων, ἔκ τε Ζακύνθου καὶ Ναυπάκτου φευγόντων.
The Medimnæans are completely unknown. Cluverius and Wesseling
conjecture Medmæans, from Medmæ or Medamæ, noticed by Strabo as a town in
the south of Italy. But this supposition cannot be adopted as certain; especially as
the total of persons named is so large. The conjecture of Palmerius—
Μηθυμναίους—has still less to recommend it. See the note of Wesseling.

[4] Diodor. xiv. 78.

[5] Diodor. xiv. 87.

[6] Diodor. xiv. 78. εἰς τὴν τῶν Σικελῶν χώραν πλεονάκις στρατεύσας, etc.
Wesseling shows in his note, that these words, and those which follow must refer
to Dionysius.

[7] Diodor. xiv. 87-103.

[8] Diodor. xiv. 8, 87, 106

[9] Diodor. xiv. 88.

[10] Diodor. xiv. 88. μετὰ δὲ τὴν ἀτυχίαν ταύτην, Ἀκραγαντῖνοι κ α ὶ

Μ ε σ σ ν ή ν ι ο ι τοὺς τὰ Διονυσίου φρονοῦντας μεταστησάμενοι, τῆς ἐλευθερίας
ἀντείχοντο, καὶ τῆς τοῦ τυράννου συμμαχίας ἀπέστησαν.
 It appears to me that the words καὶ Μεσσνήνιοι in this sentence cannot be
correct. The Messenians were a new population just established by Dionysius,
and relying upon him for protection against Rhegium: moreover they will appear,
during the events immediately succeeding, constantly in conjunction with him,
and objects of attack by his enemies.
I cannot but think that Diodorus has here inadvertently placed the word
Μεσσνήνιοι instead of a name belonging to some other community—what
community, we cannot tell.

[11] Diodor. xiv. 90-95.

[12] Diodor. xiii. 113.

[13] Diodor xiv. 90.

[14] Diodor. xiv. 95, 96.

[15] Diodor. xiv. 96.

[16] Livy, iv. 37-44; Strabo, v. p. 243-250. Diodorus (xii. 31-76) places the
commencement of the Campanian nation in 438 B. C., and their conquest of
Cumæ in 421 B. C. Skylax in his Periplus mentions both Cumæ and Neapolis as in
Campania (s. 10.) Thucydides speaks of Cumæ as being ἐν Ὀπικίᾳ (vi. 4).

[17] Strabo, v. p. 246.

[18] Thucydides (vii. 53-57) does not mention Campanians (he mentions
Tyrrhenians) as serving in the besieging Athenian armament before Syracuse
(414-413 B. C.) He does not introduce the name Campanians at all; though
alluding to Iberian mercenaries as men whom Athens calculated on engaging in
her service (vi. 90).
But Diodorus mentions, that eight hundred Campanians were engaged by the
Chalkidian cities in Sicily for service with the Athenians under Nikias, and that
they had escaped during the disasters of the Athenian army (xiii. 44).
The conquest of Cumæ in 416 B. C. opened to these Campanian Samnites an
outlet for hired military service beyond sea. Cumæ being in its Origin Chalkidic,
would naturally be in correspondence with the Chalkidic cities in Sicily. This forms
the link of connection, which explains to us how the Campanians came into
service in 413 B. C. under the Athenian general before Syracuse, and afterwards
so frequently under others in Sicily (Diodor. xiii. 62-80, etc.).

[19] Strabo, vi. p. 253, 254. See a valuable section on this subject in Niebuhr,
Römisch. Geschichte, vol. i. p. 94-98.
 It appears that the Syracusan historian Antiochus made no mention either of
Lucanians or of Bruttians, though he enumerated the inhabitants of the exact line
of territory afterwards occupied by these two nations. After repeating the
statement of Antiochus that this territory was occupied by Italians, Œnotrians,
and Chonians, Strabo proceeds to say—Οὗτος μὲν οὖν ἁπλουστέρως εἴρηκε καὶ
ἀρχαϊκῶς, οὐδὲν διορίσας περὶ τῶν Λευκανῶν καὶ τῶν Βρεττίων. The German
translator Grosskurd understands these words as meaning, that Antiochus “did
not distinguish the Lucanians from the Bruttians.” But if we read the paragraph
through, it will appear, I think, that Strabo means to say, that Antiochus had
stated nothing positive respecting either Lucanians or Bruttians. Niebuhr (p. 96 ut
suprà) affirms that Antiochus represented the Lucanians as having extended
themselves as far as Läus; which I cannot find.
The date of Antiochus seems not precisely ascertainable. His work on Sicilian
history was carried down from early times to 424 B. C. (Diodor. xii. 71). His silence
respecting the Lucanians goes to confirm the belief that the date of their
conquest of the territory called Lucania was considerably later than that year.
Polyænus (ii. 10. 2-4) mentions war as carried on by the inhabitants of Thurii,
under Kleandridas the father of Gylippus, against the Lucanians. From the age
and circumstances of Kleandridas, this can hardly be later than 420 B. C.

[20] Strabo, vi. p. 256. The Periplus of Skylax (s. 12, 13) recognizes Lucania
as extending down to Rhegium. The date to which this Periplus refers appears to
be about 370-360 B. C.: see an instructive article among Niebuhr’s Kleine
Schriften, p. 105-130. Skylax does not mention the Bruttians (Klausen, Hekatæus
and Skylax, p. 274. Berlin, 1831).

[21] Diodor. xiv. 91-101. Compare Polybius, ii. 39. When Nikias on his way to
Sicily, came near to Rhegium and invited the Rhegines to coöperate against
Syracuse, the Rhegines declined, replying, ὅ,τι ἂν καὶ τοῖς ἄλλοις Ἰταλιώταις
ξυνδοκῇ, τοῦτο ποιήσειν (Thucyd. vi. 44).

[22] Diodor. xiv. 101.

[23] Diodor. xiv. 100.

[24] Diodor. xiv 100.

[25] Herodot. vi. 21; Strabo, vi. p. 253.

[26] See the description of this mountainous region between the Tarentine
Gulf and the Tyrrhenian Sea, in an interesting work by a French General
employed in Calabria in 1809—Calabria during a military residence of Three
Years, Letters, 17, 18, 19 (translated and published by Effingham Wilson. London,
1832).
 [27] Diodor. xiv. 101. βουλόμενοι Λᾶον, πόλιν εὐδαίμονα, πολιορκῆσαι. This
appears the true reading: it is an acute conjecture proposed by Niebuhr
(Römisch. Geschicht. i. p. 96) in place of the words—βουλόμενοι λαὸν καὶ πόλιν
εὐδαίμονα, πολιορκῆσαι.

[28] Diodor. xiv. 102.

[29] Diodor. xiv. 103.

[30] Polybius (i. 6) gives us the true name of this river: Diodorus calls it the
river Helôris.

[31] Diodor. xiv. 105. παρέδωκαν αὑτοὺς περὶ ὀγδόην ὥραν, ἤδη τὰ σώματα
παρείμενοι.

[32] Diodor. xiv. 105. Καὶ πάντων αὐτοῦ ὑποπτευόντων τὸ θηριῶδες,

τοὐναντίον ἐφάνη πάντων ἐπιεικέστατος.

[33] Diodor. xiv. 105. καὶ σχεδὸν τοῦτ᾽ ἔδοξε πράττειν ἐν τῷ ζῇν κάλλιστον.
Strabo, vi. p. 261.

[34] Diodor. xiv. 106. καὶ παρακαλέσαι μηδὲν περὶ αὐτῶν ὑ π ὲ ρ ἄ ν θ ρ ω π ο ν

βουλεύεσθαι.

[35] Diodor. xiv. 106.

[36] Diodor. xiv. 106, 107.

[37] Plato, Epistol. vii. p. 332 D. Διονύσιος δὲ εἰς μίαν πόλιν ἀθροίσας πᾶσαν
Σικελίαν ὑπὸ σοφίας, etc.

[38] Diodor. xiv. 107, 108. Polyænus relates this stratagem of Dionysius
about the provisions, as if it had been practised at the siege of Himera, and not of
Rhegium (Polyæn. v. 3, 10).

[39] Diodor. xiv. 112. Ὁ δὲ Φύτων, κατὰ τὴν πολιορκίαν στρατηγὸς ἀγαθὸς
γεγενημένος, καὶ κατὰ τὸν ἄλλον βίον ἐπαινούμενος, οὐκ ἀγεννῶς ὑπέμενε τὴν ἐπὶ
τῆς τελευτῆς τιμωρίαν· ἀλλ᾽ ἀκατάπληκτον τὴν ψυχὴν φυλάξας, καὶ βοῶν, ὅτι τὴν
πόλιν οὐ βουληθεὶς προδοῦναι Διονυσίῳ τυγχάνει τῆς τιμωρίας, ἣν αὐτῷ τὸ
δαιμόνιον ἐκείνῳ συντόμως ἐπιστήσει· ὥστε τὴν ἀρετὴν τἀνδρὸς καὶ παρὰ τοῖς
στρατιώταις τοῦ Διονυσίου κατελεεῖσθαι, καί τινας ἤδη θορυβεῖν. Ὁ δὲ Διονύσιος,
εὐλαβηθεὶς μή τινες τῶν στρατιωτῶν ἀποτολμήσωσιν ἐξαρπάζειν τὸν Φύτωνα,
παυσάμενος τῆς τιμωρίας, κατεπόντωσε τὸν ἀτυχῆ μετὰ τῆς συγγενείας. Οὗτος
μὲν οὖν ἀναξίως τῆς ἀρετῆς ἐκνόμοις περιέπεσε τιμωρίαις, καὶ πολλοὺς ἔσχε καὶ
τότε τῶν Ἑλλήνων τοὺς ἀλγήσαντας τὴν συμφορὰν, καὶ μετὰ ταῦτα ποιητὰς τοὺς
θρηνήσοντας τὸ τῆς περιπετείας ἐλεεινόν.

[40] Strabo, vi. p. 258 ἐπιφανῆ δ᾽ οὖν πόλιν οὖσαν ... κατασκάψαι Διονύσιον,
etc.

[41] Polybius, ii. 39, 67.

[42] Polybius, i. 6.

[43] Chap. LXXVI. Vol. X.

[44] Livy has preserved the mention of this important acquisition of Dionysius
(xxiv. 3).
“Sed arx Crotonis, unâ parte imminens mari, alterâ vergente in agrum, situ
tantum naturali quondam munita, postea et muro cincta est, quâ per aversas
rupes ab Dionysio Siciliæ tyranno per dolum fuerat capta.”
Justin also (xx. 5) mentions the attack of Dionysius upon Kroton.
We may, with tolerable certainty, refer the capture to the present part of the
career of Dionysius.
See also Ælian, V. H. xii. 61.

[45] Aristotel. Auscult. Mirab. s. 96; Athenæus, xii. p. 541; Diodor. xiv. 77.
Polemon specified this costly robe, in his work Περὶ τῶν ἐν Καρχηδόνι
Πέπλων....

[46] Strabo, vi. p. 261.

[47] Strabo, v. p. 241. It would seem that the two maritime towns, said to
have been founded on the coast of Apulia on the Adriatic by Dionysius the
younger during the first years of his reign—according to Diodorus (xvi. 5)—must
have been really founded by the elder Dionysius, near about the time to which we
have now reached.

[48] Diodor. xv. 13, 14.

[49] Diodor. xv. 14; Strabo, v. p. 226; Servius ad Virgil. Æneid. x. 184.

[50] Justin, xx. 5; Xenoph. Hellen. vii. 1, 20.

[51] See Pseudo-Aristotel. Œconomic. ii. 20-41; Cicero, De Natur. Deor. iii.
34, 82, 85: in which passages, however, there must be several incorrect
assertions as to the actual temples pillaged; for Dionysius could not have been in
Peloponnesus to rob the temple of Zeus at Olympia, or of Æsculapius at
Epidaurus.
 Athenæus (xv. p. 693) recounts an anecdote that Dionysius plundered the
temple of Æsculapius at Syracuse of a valuable golden table; which is far more
probable.

[52] Diodor. xv. 74. See Mr. Fynes Clinton, Fast. Hellen. ad ann. 367 B. C.

[53] See a different version of the story about Philoxenus in Plutarch, De

Fortun. Alexand. Magni, p. 334 C.

[54] Diodor. xiv. 109; xv. 6.

[55] See Vol. VII. of this History, Ch. LV. p. 57 seqq.

[56] See above, in this work, Vol. X. Ch. LXXVII. p. 76. I have already noticed
the peculiarity of this Olympic festival of 384 B. C., in reference to the position
and sentiment of the Greeks in Peloponnesus and Asia. I am now obliged to
notice it again, in reference to the Greeks of Sicily and Italy—especially to
Dionysius.

[57] Diodor. xv. 14. Παρὰ δ᾽ Ἠλείοις Ὀλυμπιὰς ἤχθη ἐννενηκοστὴ ἐννάτη
(B. C. 384), καθ᾽ ἣν ἐνίκα στάδιον Δίκων Συρακούσιος.
Pausanias, vi. 3, 5. Δίκων δὲ ὁ Καλλιμβρότου πέντε μὲν Πυθοῖ δρόμου νίκας,
τρεῖς δὲ ἀνείλετο Ἰσθμίων, τέσσαρας δὲ ἐν Νεμέᾳ, καὶ Ὀλυμπιακὰς μίαν μὲν ἐν
παισὶ, δύο δὲ ἄλλας ἀνδρῶν· καὶ οἱ καὶ ἀνδριάντες ἴσοι ταῖς νίκαις εἰσὶν ἐν
Ὀλυμπίᾳ· παιδὶ μὲν δὴ ὄντι αὐτῷ Κ α υ λ ω ν ι ά τ ῃ , κ α θ ά π ε ρ γ ε κ α ὶ ἦ ν ,
ὑ π ῆ ρ ξ ε ν ἀ ν α γ ο ρ ε υ θ ῆ ν α ι· τὸ δὲ ἀπὸ τούτου Σ υ ρ α κ ο ύ σ ι ο ν α ὑ τ ὸ ν
ἀ ν η γ ό ρ ε υ σ ε ν ἐ π ὶ χ ρ ή μ α σ ι.
Pausanias here states, that Dikon received a bribe to permit himself to be
proclaimed as a Syracusan, and not as a Kauloniate. Such corruption did
occasionally take place (compare another case of similar bribery, attempted by
Syracusan envoys, Pausan. vi. 2, 4), prompted by the vanity of the Grecian cities
to appropriate to themselves the celebrity of a distinguished victor at Olympia.
But in this instance, the blame imputed to Dikon is more than he deserves.
Kaulonia had been already depopulated and incorporated with Lokri; the
inhabitants being taken away to Syracuse and made Syracusan citizens (Diodor.
xiv. 106). Dikon therefore could not have been proclaimed a Kauloniate, even had
he desired it—when the city of Kaulonia no longer existed. The city was indeed
afterwards reëstablished; and this circumstance doubtless contributed to mislead
Pausanias, who does not seem to have been aware of its temporary subversion
by Dionysius.

[58] Dionys. Hal. Judic. de Lysâ, p. 452, Reisk.

 [59] Lysias, Fragm. Orat. 33. ap. Dionys. Hal. p. 521. ὁρῶν οὕτως αἰσχρῶς
διακειμένην τὴν Ἑλλάδα, καὶ πολλὰ μὲν αὐτῆς ὄντα ὑπὸ τῷ βαρβάρῳ, πολλὰς δὲ
πόλεις ὑπὸ τυράννων ἀναστάτους γεγενημένας.

[60] Lysias, Fr. Or. 33. l. c. Ἐπίστασθε δὲ, ὅτι ἡ μὲν ἀρχὴ τῶν κρατούντων
τῆς θαλάττης, τῶν δὲ χρημάτων βασιλεὺς ταμίας· τὰ δὲ τῶν Ἑλλήνων σώματα
τῶν δαπανᾶσθαι δυναμένων· ναῦς δὲ πολλὰς αὐτὸς κέκτηται, πολλὰς δ᾽ ὁ
τύραννος τῆς Σικελίας.

[61] Lysias, Orat. Frag. l. c. Θαυμάζω δὲ Λακεδαιμονίους πάντων μάλιστα, τίνι

ποτε γνώμῃ χρώμενοι, καιομένην τὴν Ἑλλάδα περιορῶσιν, ἡγεμόνες ὄντες τῶν
Ἑλλήνων, οὐκ ἀδίκως, etc.
Οὐ γὰρ ἀλλοτρίας δεῖ τὰς τῶν ἀπολωλότων συμφορὰς νομίζειν ἀλλ᾽ οἰκείας·
οὐδ᾽ ἀναμεῖναι, ἕως ἂν ἐπ᾽ αὐτοὺς ἡμᾶς αἱ δυνάμεις
ἀμφοτέρων ἔλθωσιν, ἀλλ᾽ ἕως ἔτι ἔξεστι, τὴν τούτων ὕβριν
κ ω λ ῦ σ α ι.
I give in the text the principal points of what remains out of this discourse of
Lysias, without confining myself to the words.

[62] Diodor. xv. 23. οἱ μέγιστοι τῶν τότε δυναστῶν, etc.

[63] Diodor. xv. 13.

[64] Isokrates holds similar language, both about the destructive conquests
of Dionysius, and the past sufferings and present danger of Hellas, in his Orat. IV.
(Panegyric.) composed about 380 B. C., and (probably enough) read at the
Olympic festival of that year (s. 197). ἴσως δ᾽ ἂν καὶ τῆς ἐμῆς εὐηθείας πολλοὶ
καταγελάσειαν, εἰ δυστυχίας ἀνδρῶν ὀδυροίμην ἐν τοιούτοις καιροῖς, ἐν οἷς Ἰταλία
μὲν ἀνάστατος γέγονε, Σικελία δὲ καταδεδούλωται (compare s. 145), τοσαῦται δὲ
πόλεις τοῖς βαρβάροις ἐκδέδονται, τὰ δὲ λοιπὰ μέρη τῶν Ἑλλήνων ἐν τοῖς
μεγίστοις κινδύνοις ἐστίν.
Isokrates had addressed a letter to the elder Dionysius. He alludes briefly to it
in his Orat. ad Philippum (Orat. v. s. 93), in terms which appear to indicate that it
was bold and plain spoken (θρασύτερον τῶν ἄλλων). The first letter, among the
ten ascribed to Isokrates, purports to be a letter to Dionysius; but it seems rather
(to judge by the last words) to be the preface of a letter about to follow. Nothing
distinct can be made out from it as it now stands.

[65] Strabo, v. p. 212.

[66] Dionys. Hal. p. 519. Jud. de Lysiâ. Ἐστὶ δή τις αὐτῷ πανηγυρικὸς λόγος,
ἐν ᾧ πείθει τοὺς Ἕλληνας ... ἐκβάλλειν Διονύσιον τὸν τύραννον τῆς ἀρχῆς, καὶ
Σικελίαν ἐλευθερῶσαι, ἄρξασθαί τε τῆς ἐχθρᾶς αὐτίκα μάλα, διαρπάσαντας τὴν
τοῦ τυράννου σκηνὴν χρυσῷ τε καὶ πορφύρᾳ καὶ ἄλλῳ πλούτῳ πολλῷ
κεκοσμημένην, etc.
Diodor. xiv. 109. Λυσίας ... προετρέπετο τὰ πλήθη μὴ προσδέχεσθαι τοῖς ἱεροῖς
ἀγῶσι τοὺς ἐξ ἀσεβεστάτης τυραννίδος ἀπεσταλμένους θεωρούς.
Compare Plutarch Vit. x. Orator, p. 836 D.

[67] Diodor. xiv. 109. ὥστε τινὰς τολμῆσαι διαρπάζειν τὰς σκηνάς.

[68] Diodor. xiv. 109.

[69] Diodor. xiv. 109.

[70] Diodor. xv. 7. Ὁ δὲ Διονύσιος, ἀκούσας τὴν τῶν ποιημάτων

καταφρόνησιν, ἐνέπεσεν εἰς ὑπερβολὴν λύπης. Ἀεὶ δὲ μᾶλλον τοῦ πάθους ἐπίτασιν
λαμβάνοντος, μανιωδὴς διάθεσις κάτεσχε τὴν ψυχὴν αὐτοῦ, καὶ φθονεῖν αὐτῷ
φάσκων ἅπαντας, τοὺς φίλους ὑπώπτευεν ὡς ἐπιβουλεύοντας· καὶ πέρας, ἐπὶ
τοσοῦτο προῆλθε λύπης καὶ παρακοπῆς, ὥστε τῶν φίλων πολλοὺς μὲν ἐπὶ
ψευδέσιν αἰτίαις ἀνελεῖν, οὐκ ὀλίγους δὲ καὶ ἐφυγάδευσεν· ἐν οἷς ἦν Φίλιστος καὶ
Λεπτίνης ὁ ἀδελφὸς, etc.

[71] For the banishment, and the return of Philistus and Leptinês, compare
Diodor. xv. 7, and Plutarch, Dion. c. 11. Probably it was on this occasion that
Polyxenus, the brother-in-law of Dionysius, took flight as the only means of
preserving his life (Plutarch, Dion. c. 21).
Plutarch mentions the incident which offended Dionysius and caused both
Philistus and Leptinês to be banished. Diodorus does not notice this incident; yet
it is not irreconcilable with his narrative. Plutarch does not mention the
banishment of Leptinês, but only that of Philistus.
On the other hand, he affirms (and Nepos also, Dion. c. 3) that Philistus did
not return until after the death of the elder Dionysius, while Diodorus states his
return conjointly with that of Leptinês—not indicating any difference of time. Here
I follow Plutarch’s statement as the more probable.
There is, however, one point which is perplexing. Plutarch (Timoleon, c. 15)
animadverts upon a passage in the history of Philistus, wherein that historian had
dwelt, with a pathos which Plutarch thinks childish and excessive, upon the
melancholy condition of the daughters of Leptinês, “who had fallen from the
splendor of a court into a poor and mean condition.” How is this reconcilable with
the fact stated by Diodorus, that Leptinês was recalled from exile by Dionysius
after a short time, taken into favor again, and invested with command at the
battle of Kronium, where he was slain? It seems difficult to believe that Philistus
could have insisted with so much sympathy upon the privations endured by the
daughters of Leptinês, if the exile of the father had lasted only a short time.
 [72] In a former chapter of this History (Vol. X. Ch. LXXVII. p. 75), I have
already shown grounds, derived from the circumstances of Central Greece and
Persia, for referring the discourse of Lysias, just noticed, to Olympiad 99 or 384
B. C. I here add certain additional reasons, derived from what is said about
Dionysius, towards the same conclusion.
In xiv. 109, Diodorus describes the events of 388 B. C., the year of Olympiad
98, during which Dionysius was still engaged in war in Italy, besieging Rhegium.
He says that Dionysius made unparalleled efforts to send a great display to this
festival; a splendid legation, with richly decorated tents, several fine chariots-and-
four, and poems to be recited by the best actors. He states that Lysias the orator
delivered a strong invective against him, exciting those who heard it to exclude
the Syracusan despot from sacrificing, and to plunder the rich tents. He then
details how the purposes of Dionysius failed miserably on every point; the fine
tents were assailed, the chariots all ran wrong or were broken, the poems were
hissed, the ships returning to Syracuse were wrecked, etc. Yet in spite of this
accumulation of misfortunes (he tells us) Dionysius was completely soothed by his
flatterers (who told him that such envy always followed upon greatness), and did
not desist from poetical efforts.
Again, in xv. 6, 7, Diodorus describes the events of 386 B. C. Here he again
tells us, that Dionysius, persevering in his poetical occupations, composed verses
which were very indifferent—that he was angry with and punished Philoxenus and
others who criticized them freely—that he sent some of these compositions to be
recited at the Olympic festival, with the best actors and reciters—that the poems,
in spite of these advantages, were despised and derided by the Olympic audience
—that Dionysius was distressed by this repulse, even to anguish and madness,
and to the various severities and cruelties against his friends which have been
already mentioned in my text.
Now upon this we must remark:—
1. The year 386 B. C. is not an Olympic year. Accordingly, the proceedings
described by Diodorus in xv. 6, 7, all done by Dionysius after his hands were free
from war, must be transferred to the next Olympic year, 384 B. C. The year in
which Dionysius was so deeply stung by the events of Olympia, must therefore
have been 384 B. C., or Olympiad 99 (relating to 388 B. C.).
2. Compare Diodor. xiv. 109 with xv. 7. In the first passage, Dionysius is
represented as making the most prodigious efforts to display himself at Olympia
in every way, by fine tents, chariots, poems, etc.—and also as having undergone
the signal insult from the orator Lysias, with the most disgraceful failure in every
way. Yet all this he is described to have borne with tolerable equanimity, being
soothed by his flatterers. But, in xv. 7 (relating to 386 B. C., or more probably to
384 B. C.) he is represented as having merely failed in respect to the effect of his
poems; nothing whatever being said about display of any other kind, nor about
an harangue from Lysias, nor insult to the envoys or the tents. Yet the simple
repulse of the poems is on this occasion affirmed to have thrown Dionysius into a
paroxysm of sorrow and madness.
Now if the great and insulting treatment, which Diodorus refers to 388 B. C.,
could be borne patiently by Dionysius—how are we to believe that he was driven
mad by the far less striking failure in 384 B. C.? Surely it stands to reason that the
violent invective of Lysias and the profound humiliation of Dionysius, are parts of
one and the same Olympic phænomenon; the former as cause, or an essential
part of the cause—the latter as effect. The facts will then read consistently and in
proper harmony. As they now appear in Diodorus, there is no rational explanation
of the terrible suffering of Dionysius described in xv. 7; it appears like a comic
exaggeration of reality.
3. Again, the prodigious efforts and outlay, which Diodorus affirms Dionysius
to have made in 388 B. C. for display at the Olympic games—come just at the
time when Dionysius, being in the middle of his Italian war, could hardly have had
either leisure or funds to devote so much to the other purpose; whereas at the
next Olympic festival, or 384 B. C., he was free from war, and had nothing to
divert him from preparing with great efforts all the means of Olympic success.
It appears to me that the facts which Diodorus has stated are nearly all
correct, but that he has misdated them, referring to 388 B. C., or Olymp. 98—
what properly belongs to 384 B. C., or Olymp. 99. Very possibly Dionysius may
have sent one or more chariots to run in the former of the two Olympiads; but his
signal efforts, with his insulting failure brought about partly by Lysias, belong to
the latter.
Dionysius of Halikarnassus, to whom we owe the citation from the oration of
Lysias, does not specify to which of the Olympiads it belongs.

[73] Diodor. xv. 7. διὸ καὶ ποιήματα γράφειν ὑπεστήσατο μετὰ πολλῆς
σπουδῆς, καὶ τοὺς ἐν τούτοις δόξαν ἔχοντας μετεπέμπετο, καὶ προτιμῶν αὐτοὺς
συνδιέτριβε, καὶ τῶν π ο ι η μ ά τ ω ν ἐ π ι σ τ ά τ α ς κ α ὶ δ ι ο ρ θ ω τ ὰ ς ε ἶ χ ε ν.
The Syracusan historian Athanis (or Athenis) had noticed some peculiar
phrases which appeared in the verses of Dionysius: see Athenæus, iii. p. 98.

[74] Thucyd. vi. 16. Οἱ γὰρ Ἕλληνες καὶ ὑπὲρ δύναμιν μείζω ἡμῶν τὴν πόλιν
ἐνόμισαν, τῷ ἐμῷ διαπρεπεῖ τῆς Ὀλυμπιάζε θεωρίας (speech of Alkibiadês).

[75] See a striking passage in the discourse called Archidamus (Or. vi. s. 111,
112) of Isokrates, in which the Spartans are made to feel keenly their altered
position after the defeat of Leuktra: especially the insupportable pain of
encountering, when they attended the Olympic festival, slights or disparagement
from the spectators, embittered by open taunts from the reëstablished
Messenians—instead of the honor and reverence which they had become
accustomed to expect.
 This may help us to form some estimate of the painful sentiment of Dionysius,
when his envoys returned from the Olympic festival of 384 B. C.

[76] There are different statements about the precise year in which Plato was
born: see Diogenes Laert. iii. 1-6. The accounts fluctuate between 429 and 428
B. C.; and Hermodorus (ap. Diog. L. iii. 6) appears to have put it in 427 B. C.: see
Corsini, Fast. Attic. iii. p. 230; Ast. Platon’s Leben, p. 14.
Plato (Epistol. vii. p. 324) states himself to have been about (σχεδὸν) forty
years of age when he visited Sicily for the first time. If we accept as the date of
his birth 428 B. C., he would be forty years of age in 388 B. C.
It seems improbable that the conversation of Plato with Dion at Syracuse
(which was continued sufficiently long to exercise a marked and permanent
influence on the character of the latter), and his interviews with Dionysius, should
have taken place while Dionysius was carrying on the Italian war or the siege of
Rhegium. I think that the date of the interview must be placed after the capture
of Rhegium in 387 B. C. And the expression of Plato (given in a letter written more
than thirty years afterwards) about his own age, is not to be taken as excluding
the supposition that he might have been forty-one or forty-two when he came to
Syracuse.
Athenæus (xi. p. 507) mentions the visit of Plato.

[77] Plutarch, Dion. c. 5.

[78] Plutarch, Dion, c. 5; Diodor. xv. 7; Diogen. Laert. iii. 17; Cornelius
Nepos, Dion, c. 2.

[79] Diodor. xiv. 6.3. It was in the construction of these extensive

fortifications, seemingly, that Dionysius demolished the chapel which had been
erected by the Syracusans in honor of Dioklês (Diodor. xiii. 635).
Serra di Falco (Antichità di Sicilia, vol. iv. p. 107) thinks that Dionysius
constructed only the northern wall up the cliff of Epipolæ, not the southern. This
latter (in his opinion) was not constructed until the time of Hiero II.
I dissent from him on this point. The passage here referred to in Diodorus
affords to my mind sufficient evidence that the elder Dionysius constructed both
the southern wall of Epipolæ and the fortification of Neapolis. The same
conclusion moreover appears to result from what we read of the proceedings of
Dion and Timoleon afterwards.

[80] Diodor. xv. 13.

[81] See Plato, Epist. vii. p. 333, 336—also some striking lines, addressed by
the poet Theokritus to Hiero II. despot at Syracuse in the succeeding century:
Theokrit. xvi. 75-85.
Dionysius—ἐζήτει λαβεῖν πρόφασιν εὔλογον τοῦ πολέμου, etc.
 [82] Diodor. xv. 15.

[83] Diodor. xv. 15.

[84] Diodor. xv. 16, 17.

[85] Diodor. xv. 17.

[86] Plato, Epistol. vii. p. 333 A. After reciting the advice which Dion and he
had given to Dionysius the younger, he proceeds to say—ἕτοιμον γὰρ εἶναι,
τούτων γενομένων, πολὺ μᾶλλον δουλώσασθαι Καρχηδονίους τῆς ἐπὶ Γέλωνος
αὐτοῖς γενομένης δουλείας, ἀ λ λ ᾽ ο ὐ χ , ὥ σ π ε ρ ν ῦ ν τ ο ὐ ν α ν τ ί ο ν , ὁ
π α τ ὴ ρ α ὐ τ ο ῦ φ ό ρ ο ν ἐ τ ά ξ α τ ο φ έ ρ ε ι ν τ ο ῖ ς β α ρ β ά ρ ο ι ς, etc.

[87] Diodor. xv. 24.

[88] Strabo, vi. p. 261; Pliny, H. N. iii. 10. The latter calls the isthmus twenty
miles broad, and says that Dionysius wished (intercisam) to cut it through: Strabo
says that he proposed to wall it across (διατειχίζειν), which is more probable.

[89] Xenoph. Hellen. vi. 2, 4, 33; vii i. 20-28. Diodor. xv. 70.

[90] Diodor. xxii. p. 304.

[91] Diodor. xv. 73; xvi. 5.

[92] Diodor. xv. 74.

[93] Plutarch, Timoleon, c. 15.

[94] Polyb. xv. 35. Διὸ καὶ Πόπλιον Σκιπίωνά φασι, τὸν πρῶτον
καταπολεμήσαντα Καρχηδονίους, ἐρωτηθέντα, τίνας ὑπολαμβάνει
πραγματικωτάτους ἄνδρας γεγονέναι καὶ σὺν νῷ τολμηροτάτους, εἰπεῖν, τοὺς περὶ
Ἀγαθοκλέα καὶ Διονύσιον τοὺς Σικελιώτας.

[95] Plutarch, Dion, c. 7.

[96] The example of Dionysius—his long career of success and quiet death—
is among those cited by Cotta in Cicero (De Nat. Deor. iii. 33, 81, 85) to refute the
doctrine of Balbus, as to the providence of the gods and their moral government
over human affairs.

[97] Isokratês, Or. v. (Philipp.) s. 73. Διονύσιος ... ἐπιθυμήσας μοναρχίας

ἀ λ ό γ ω ς κ α ὶ μ α ν ι κ ῶ ς, καὶ τολμήσας ἅπαντα πράττειν τὰ φέροντα πρὸς τὴν
δύναμιν ταύτην, etc.
 [98] Thucyd. vi. 55. ἀλλὰ καὶ διὰ τὸ πρότερον ξύνηθες, τοῖς μὲν πολίταις
φοβερὸν, τοῖς δὲ ἐπικούροις ἀκριβὲς, πολλῷ τῷ περιόντι τοῦ ἀσφαλοῦς ἐκράτησε
(Hippias).
On the liberality of the elder Dionysius to his mercenaries, see an allusion in
Plato, Epistol. vii. p. 348 A.
The extension and improvement of engines for warlike purposes, under
Dionysius, was noticed as a sort of epoch (Athenæus, De Machinis ap. Mathemat.
Veteres, ed. Paris, p. 3).

[99] Cornelius Nepos, De Regibus, c. 2. “Dionysius prior, et manu fortis, et

belli peritus fuit, et, id quod in tyranno non facile reperitur, minime libidinosus,
non luxuriosus, non avarus, nullius rei denique cupidus, nisi singularis
perpetuique imperii, ob eamque rem crudelis. Nam dum id studuit munire, nullius
pepercit vitæ, quem ejus insidiatorem putaret.” To the same purpose Cicero, Tusc.
Disp. v. 20.

[100] Aristotel. Politic. v. 9, 5.

[101] Pseudo-Aristotel. Œconomic. ii. c. 21, 42; Cicero, De Nat. Deorum, iii.
34, 83, 84; Valerius Maxim. i. 1.

[102] Plutarch, Dion, c. 28; Plutarch, De Curiositate, p. 523 A; Aristotel.

Politic. v. 9, 3. The titles of these spies—αἱ ποταγωγίδες καλούμεναι—as we read
in Aristotle; or οἱ ποταγωγεῖς—as we find in Plutarch—may perhaps both be
correct.

[103] Cicero in Verrem, v. 55, 143.

[104] Plutarch, De Fortunâ Alexandr. Magni, p. 338 B. What were the crimes
of Dionysius which Pausanias had read and describes by the general words
Διονυσίου τὰ ἀνοσιώτατα—and which he accuses Philistus of having intentionally
omitted in his history—we cannot now tell (Pausan. i. 13, 2: compare Plutarch,
Dion, c. 36). An author named Amyntianus, contemporary with Pausanias, and
among those perused by Photius (Codex 131), had composed parallel lives of
Dionysius and the Emperor Domitian.

[105] Plato, Epistol. vii. p. 332 A; Aristotel. Politic. v. 5, 6.

[106] Plato, Epistol. vii. p. 332 D. Διονύσιος δὲ εἰς μίαν πόλιν ἀθροίσας πᾶσαν
Σικελίαν ὑπὸ σοφίας, π ι σ τ ε ύ ω ν ο ὐ δ ε ν ὶ , μ ό γ ι ς ἐ σ ώ θ η, etc.
This brief, but significant expression of Plato, attests the excessive mistrust
which haunted Dionysius, as a general fact; which is illustrated by the anecdotes
of Cicero, Tuscul. Disput. v. 20, 23; and De Officiis, ii. 7; Plutarch, Dion, c. 9;
Diodor. xiv. 2.
 The well-known anecdote of Damoklês, and the sword which Dionysius caused
to be suspended over his head by a horsehair, in the midst of the enjoyments of
the banquet, as an illustration how little was the value of grandeur in the midst of
terror—is recounted by Cicero.

[107] Plutarch, Dion, c. 3; Plutarch, Timoleon, c. 6.

[108] This sentiment, pronounced by Plato, Isokratês, Cicero, Seneca,

Plutarch, etc., is nowhere so forcibly laid out as in the dialogue of Xenophon
called Hiero—of which indeed it forms the text and theme. Whoever reads this
picture of the position of a Grecian τύραννος, will see that it was scarcely possible
for a man so placed to be other than a cruel and oppressive ruler.

[109] See the citation from Plato, in a note immediately preceding.

[110] Plato, Epistol. iii. p. 315 E. (to the younger Dionysius). Φασὶ δ᾽ οὐκ
ὀλίγοι λέγειν σε πρός τινας τῶν παρά σε πρεσβευόντων, ὡς ἄρα σοῦ ποτὲ
λέγοντος ἀκούσας ἐγὼ μέλλοντος τ ά ς τ ε Ἑ λ λ η ν ί δ α ς π ό λ ε ι ς ἐ ν Σ ι κ ε λ ί ᾳ
ο ἰ κ ί ζ ε ι ν , κ α ὶ Σ υ ρ α κ ο υ σ ί ο υ ς ἐ π ι κ ο υ φ ί σ α ι, τὴν ἀρχὴν ἀντὶ τυραννίδος εἰς
βασιλείαν μεταστήσαντα, ταῦτ᾽ ἄρα σὲ μέν τοτε διεκώλυσα, σοῦ σφόδρα
προθυμουμένου, νῦν δὲ Δίωνα διδάσκοιμι δρᾷν αὐτὰ ταῦτα, καὶ τοῖς διανοήμασι
τοῖς σοῖς τὴν σὴν ἀρχὴν ἀφαιρούμεθά σε.
Ibid. p. 319 C. Μή με διάβαλλε λέγων, ὡς οὐκ εἴων σε πόλεις Ἑλληνίδας
ἐῤῥούσας ὑπὸ βαρβάρων οἰκίζειν, οὐδὲ Συρακουσίους ἐπικουφίσαι ... ὡ ς ἐ γ ὼ
μ ὲ ν ἐ κ έ λ ε υ ο ν , σ ὺ δ ᾽ ο ὐ κ ἤ θ ε λ ε ς π ρ ά τ τ ε ι ν α ὐ τ ά.
Again, see Epistol. vii. p. 331 F. 332 B. 334 D. 336 A.-D.—and the brief notice
given by Photius (Codex, 93) of the lost historical works of Arrian, respecting Dion
and Timoleon.
Epistol. viii. p. 357 A. (What Dion intended to do, had he not been prevented
by death)—Καὶ μετὰ ταῦτα Σικελίαν ἂν τὴν ἄλλην κατῴκισα, τ ο ὺ ς μ ὲ ν
βαρβάρους ἣν νῦν ἔχουσιν ἀφελόμενος, ὅσοι μὴ ὑπὲρ τῆς
κοινῆς ἐλευθερίας διεπολέμησαν πρὸς τὴν τυραννίδα, τοὺς δ᾽
ἔμπροσθεν οἰκητὰς τῶν Ἑλληνικῶν τόπων εἰς τὰς ἀρχαίας καὶ
π α τ ρ ῴ α ς ο ἰ κ ή σ ε ι ς κ α τ ο ι κ ί σ α ς. Compare Plutarch, Timoleon, c. 2. αἱ δὲ
πλεῖσται πόλεις ὑπὸ βαρβάρων μιγάδων καὶ στρατιωτῶν ἀμίσθων κατείχοντο.
The βάρβαροι to whom Plato alludes in this last passage, are not the
Carthaginians (none of whom could be expected to come in and fight for the
purpose of putting down the despotism at Syracuse), but the Campanian and
other mercenaries provided for by the elder Dionysius on the lands of the
extruded Greeks. These men would have the strongest interest in upholding the
despotism, if the maintenance of their own properties was connected with it. Dion
thought it prudent to conciliate this powerful force by promising confirmation of
their properties to such of them as would act upon the side of freedom.
 [111] Both Diodorus (xvi. 9) and Cornelius Nepos (Dion, c. 5) speak of one
hundred thousand foot and ten thousand horse. The former speaks of four
hundred ships of war; the latter of five hundred.
The numbers of foot and horse appear evidently exaggerated. Both authors
must have copied from the same original; possibly Ephorus.

[112] Plutarch, Dion, c. 6; Theopompus, Fr. 204, ed. Didot. ap. Athenæum,
x. p. 435; Diodor. xvi. 6; Cornel. Nepos (Dion, c. 1).
The Scholiast on Plato’s fourth Epistle gives information respecting the
personal relations and marriages of the elder Dionysius, not wholly agreeing with
what is stated in the sixth chapter of Plutarch’s Life of Dion.

[113] Plutarch, Dion, c. 3. The age of the younger Dionysius is nowhere

positively specified. But in the year 356 B. C.—or 355 B. C., at the latest—he had a
son, Apollokratês, old enough to be entrusted with the command of Ortygia,
when he himself evacuated it for the first time (Plutarch, Dion, c. 37). We cannot
suppose Apollokratês to have been less than sixteen years of age at the moment
when he was entrusted with such a function, having his mother and sisters under
his charge (c. 50). Apollokratês therefore must have been born at least as early
as 372 B. C.; perhaps even earlier. Suppose Dionysius the younger to have been
twenty years of age when Apollokratês was born; he would thus be in his twenty-
fifth year in the beginning of 367 B. C., when Dionysius the elder died. The
expressions of Plato, as to the youth of Dionysius the younger at that juncture,
are not unsuitable to such an age.

[114] Aristotel. Polit. v. 5, 6.

[115] Plato Epistol. vii. p. 347 A. Compare the offer of Dion to maintain fifty
triremes at his own expense (Plutarch, Dion, c. 6.)

[116] Dion was fifty-five years of age at the time of his death, in the fourth
year after his departure from Peloponnesus (Cornelius Nepos, Dion, c. 10).
His death took place seemingly about 354 B. C. He would thus be born about
408 B. C.

[117] Plato, Epistol. vii. p. 326 D. ἐλθόντα δέ με ὁ ταύτῃ λεγόμενος αὖ βίος

εὐδαίμων, Ἰταλιωτικῶν τε καὶ Συρακουσίων τραπεζῶν πλήρης, οὐδαμῆ οὐδαμῶς
ἤρεσκε, δίς τε τῆς ἡμέρας ἐμπιμπλάμενον ζῇν καὶ μηδέποτε κοιμώμενον μόνον
νύκτωρ, etc.

[118] Cicero, De Finibus, v. 20; De Republic. i. 10. Jamblichus (Vit.

Pythagoræ, c. 199) calls Dion a member of the Pythagorean brotherhood, which
may be doubted; but his assertion that Dion procured for Plato, though only by
means of a large price (one hundred minæ), the possession of a book composed
by the Pythagorean Philolaus, seems not improbable. The ancient Pythagoreans
wrote nothing. Philolaus (seemingly about contemporary with Sokrates) was the
first Pythagorean who left any written memorial. That this book could only be
obtained by the intervention of an influential Syracusan—and even by him only
for a large price—is easy to believe.
See the instructive Dissertation of Gruppe, Ueber die Fragmente des Archytas
und der älteren Pythagoreer, p. 24, 26, 48, etc.

[119] See a remarkable passage, Plato, Epist. vii p. 328 F.

[120] Plato, Epistol. vii. p. 335 F. Δίωνα γὰρ ἐγὼ σαφῶς οἶδα, ὡς οἷόν τε περὶ
ἀνθρώπων ἄνθρωπον διϊσχυρίζεσθαι, ὅτι τὴν ἀρχὴν εἰ κατέσχεν, ὡς οὐκ ἄν ποτε
ἐπ᾽ ἄλλο γε σχῆμα τῆς ἀρχῆς ἐτράπετο, ἢ ἐπὶ τὸ—Συρακούσας μὲν πρῶτον, τὴν
πατρίδα τὴν ἑαυτοῦ, ἐπεὶ τὴν δουλείαν αὐτῆς ἀπήλλαξε καὶ φαιδρύνας ἐλευθερίῳ
ἐν σχήματι κατέστησε, τὸ μετὰ τοῦτ᾽ ἂν πάσῃ μηχάνῃ ἐκόσμησε νόμοις τοῖς
προσήκουσί τε καὶ ἀρίστοις τοὺς πολίτας—τό τε ἐφεξῆς τούτοις προυθυμεῖτ᾽ ἂν
πρᾶξαι, πᾶσαν Σικελίαν κατοικίζειν καὶ ἐλευθέραν ἀπὸ τῶν βαρβάρων ποιεῖν, τοὺς
μὲν ἐκβάλλων, τοὺς δὲ χειρούμενος ῥᾷον Ἱέρωνος, etc.
Compare the beginning of the same epistle, p. 324 A.

[121] Plato, Epist. iv. p. 320 F. (addressed to Dion). ... ὡς οὖν ὑπὸ πάντων
ὁρώμενος παρασκευάζου τόν τε Λυκοῦργον ἐκεῖνον ἀρχαῖον ἀποδείξων, καὶ τὸν
Κῦρον καὶ εἴτις ἄλλος πώποτε ἔδοξεν ἤθει καὶ πολιτείᾳ διενεγκεῖν, etc.

[122] Plutarch, Kleomenes, c. 2-11.

[123] Plato, Epistol. vii. p. 327 A. Δίων μὲν γὰρ δὴ μάλ᾽ εὐμαθὴς ὢν πρός τε
τἄλλα, καὶ πρὸς τοὺς τότε ὑπ᾽ ἐμοῦ λεγομένους λόγους, οὕτως ὀξέως ὑπήκουσε
καὶ σφόδρα, ὡς οὐδεὶς πώποτε ὧν ἐγὼ προσέτυχον νέων, καὶ τὸν ἐπίλοιπον βίον
ζῇν ἠθέλησε διαφερόντως τῶν πολλῶν Ἰταλιωτῶν καὶ Σικελιωτῶν, ἀρετὴν περὶ
πλείονος ἡδονῆς τῆς τε ἄλλης τρυφῆς ποιούμενος· ὅθεν ἐπαχθέστερον τοῖς περὶ
τὰ τυραννικὰ νόμιμα ζῶσιν ἐβίω, μέχρι τοῦ θανάτου τοῦ περὶ Διονύσιον
γενομένου.
Plutarch, Dion, c. 4. ὡς πρῶτον ἐγεύσατο λόγου καὶ φιλοσοφίας ἡγεμονικῆς
πρὸς ἀρετήν, ἀνεφλέχθη τὴν ψυχὴν, etc.

[124] See the story in Jamblichus (Vit. Pythagoræ, c. 189) of a company of

Syracusan troops under Eurymenes the brother of Dion, sent to lay in ambuscade
for some Pythagoreans between Tarentum and Metapontum. The story has not
the air of truth; but the state of circumstances, which it supposes, illustrates the
relation between Dionysius and the cities in the Tarentine Gulf.

[125] Plutarch, Dion, c. 5, 6; Cornelius Nepos, Dion, c. 1, 2.

 [126] Plutarch, Dion, c. 17, 49. Respecting the rarity of the vote of Spartan
citizenship, see a remarkable passage of Herodotus, ix. 33-35.
Plutarch states that the Spartans voted their citizenship to Dion during his
exile, while he was in Peloponnesus after the year 367 B. C., at enmity with the
younger Dionysius then despot of Syracuse; whom (according to Plutarch) the
Spartans took the risk of offending, in order that they might testify their extreme
admiration for Dion.
I cannot but think that Plutarch is mistaken as to the time of this grant. In and
after 367 B. C. the Spartans were under great depression, playing the losing game
against Thebes. It is scarcely conceivable that they should be imprudent enough
to alienate a valuable ally for the sake of gratuitously honoring an exile whom he
hated and had banished. Whereas if we suppose the vote to have been passed
during the lifetime of the elder Dionysius, it would count as a compliment to him
as well as to Dion, and would thus be an act of political prudence as well as of
genuine respect. Plutarch speaks as if he supposed that Dion was never in
Peloponnesus until the time of his exile, which is, in my judgment, highly
improbable.

[127] Cornelius Nepos, Dion, c. 2; Plutarch, Dion, c. 6.

[128] Diodor. xv. 74.

[129] Plato, Epistol. vii. p. 338 E. Ὁ δὲ οὔτε ἄλλως ἐστὶν ἀφυὴς πρὸς τὴν τοῦ
μανθάνειν δύναμιν, φιλότιμος δὲ θαυμαστῶς, etc. Compare p. 330 A. p. 328 B.;
also Epist. iii. p. 316 C. p. 317 E.
Plutarch, Dion, c. 7-9.

[130] Plato, Epistol. vii. p. 332 E. ἐπειδὴ τὰ παρὰ τοῦ πατρὸς αὐτῷ
συνεβεβήκει οὕτως ἀνομιλήτῳ μὲν παιδείας, ἀνομιλήτῳ δὲ συνουσιῶν τῶν
προσηκουσῶν γεγονέναι, etc.

[131] Plutarch Dion, c. 6.

[132] Plutarch, Dion, c. 7. Ὁ μὲν οὖν Διονύσιος ὑπερφυῶς τὴν μεγαλοψυχίαν

ἐθαύμασε καὶ τὴν προθυμίαν ἠγάπησεν.

[133] Dionysius II. was engaged at war at the time when Plato first visited
him at Syracuse, within the year immediately after his accession (Plato, Epistol. iii.
p. 317 A). We may reasonably presume that this was the war with Carthage.
Compare Diodorus (xvi. 5), who mentions that the younger Dionysius also
carried on war for some little time, in a languid manner, against the Lucanians;
and that he founded two cities on the coast of Apulia in the Adriatic. I think it
probable that these two last-mentioned foundations were acts of Dionysius I., not
of Dionysius II. They were not likely to be undertaken by a young prince of
backward disposition, at his first accession.

[134] Tacitus, Histor. ii. 49. “Othoni sepulcrum exstructum est, modicum, et
mansurum.”
A person named Timæus was immortalized as the constructor of the funeral
pile: see Athenæus, v. p. 206. Both Göller (Timæi Fragm. 95) and M. Didot
(Timæi Fr. 126) have referred this passage to Timæus the historian, and have
supposed it to relate to the description given by Timæus of the funeral-pile. But
the passage in Athenæus seems to me to indicate Timæus as the builder, not the
describer, of this famous πυρά.
It is he who is meant, probably, in the passage of Cicero (De Naturâ Deor. iii.
35)—(Dionysius) “in suo lectulo mortuus in Tympanidis rogum illatus est, eamque
potestatem quam ipse per scelus erat nactus, quasi justam et legitimam
hereditatis loco filio tradidit.” This seems at least the best way of explaining a
passage which perplexes the editors: see the note of Davis.

[135] Plutarch (De Exilio. p. 637) and Cornelius Nepos (Dion, c. 3) represent
that Philistus was recalled at the persuasion of the enemies of Dion, as a
counterpoise and corrective to the ascendency of the latter over Dionysius the
younger. Though Philistus afterwards actually performed this part, I doubt
whether such was the motive which caused him to be recalled. He seems to have
come back before the obsequies of Dionysius the elder; that is, very early after
the commencement of the new reign. Philistus had described, in his history, these
obsequies in a manner so elaborate and copious, that this passage in his work
excited the special notice of the ancient critics (see Philisti Fragment. 42, ed.
Didot; Plutarch, Pelopidas, c. 34). I venture to think that this proves him to have
been present at the obsequies; which would of course be very impressive to him,
since they were among the first things which he saw after his long exile.

[136] Plutarch, Dion, c. 11. Ταῦτα πολλάκις τοῦ Δίωνος παραινοῦντος, καὶ
τῶν λόγων τοῦ Πλάτωνος ἔστιν οὕστινας ὑποσπείροντος, etc.

[137] Plutarch, Dion, c. 10, 11; Plato, Epist. vii. p. 327 C.

[138] Plato, Epist. vii. p. 328 A. p. 335 E.; Plato, Republic, vi. p. 499 C. D.

[139] Plato, Epist. vii. p. 327 E. ... Ὃ δὴ καὶ νῦν εἰ διαπράξαιτο ἐν Διονυσίῳ
ὡς ἐπεχείρησε, μεγάλας ἐλπίδας εἶχεν, ἄνευ σφαγῶν καὶ θανάτων καὶ τῶν νῦν
γεγονότων κακῶν, βίον ἂν εὐδαίμονα καὶ ἀληθινὸν ἐν πάσῃ τῇ χώρᾳ
κατασκευάσαι.

[140] Plato, Epistol. vii. p. 333 B. Ταὐτὸν πρὸς Δίωνα Συρακόσιοι τότε
ἔπαθον, ὅπερ καὶ Διονύσιος, ὅτε αὐτὸν ἐπεχείρει παιδεύσας καὶ θρέψας βασιλέα
τῆς ἀρχῆς ἄξιον, οὕτω κοινωνεῖν αὐτῷ τοῦ βίου παντός.

[141] Plato, Epist. vii. p. 327 E.; Plutarch. Dion, c. 11. ἔσχεν ἔρως τὸν
Διονύσιον ὀξὺς καὶ περιμανὴς τῶν τε λόγων καὶ τῆς συνουσίας τοῦ Πλάτωνος.
Εὐθὺς οὖν Ἀθήναζε πολλὰ μὲν ἐφοίτα γράμματα παρὰ τοῦ Διονυσίου, πολλαὶ δ᾽
ἐπισκήψεις τοῦ Δίωνος, ἄλλαι δ᾽ ἐξ Ἰταλίας παρὰ τῶν Πυθαγορικῶν, etc.

[142] Plato, Epist. vii. p. 328.

[143] Plato, Epistol. vii. p. 328. Ταύτῃ μὲν τῇ διανοίᾳ καὶ τόλμῃ ἀπῇρα
οἴκοθεν, οὐχ ᾗ τινὲς ἐδόξαζον, ἀλλ᾽ αἰσχυνόμενος μὲν
ἐ μ α υ τ ὸ ν τ ὸ μ έ γ ι σ τ ο ν, μὴ δόξαιμί ποτε ἐμαυτῷ παντάπασι λόγος μόνον
ἀτεχνῶς εἶναί τις, ἔργου δὲ οὐδενὸς ἄν ποτε ἑκὼν ἀνθάψασθαι, κινδυνεύσειν δὲ
προδοῦναι πρῶτον μὲν τὴν Δίωνος ξενίαν ἐν κινδύνοις ὄντως γεγονότος οὐ
σμικροῖς· εἴτ᾽ οὖν πάθοι τι, εἴτ᾽ ἐκπεσὼν ὑπὸ Διονυσίου καὶ τῶν ἄλλων ἐχθρῶν
ἔλθοι παρ᾽ ἡμᾶς φεύγων, καὶ ἀνέροιτο, εἰπών, etc.

[144] This is contained in the words ο ὐ χ ᾗ τ ι ν ὲ ς ἐ δ ό ξ α ζ ο ν—before

cited.

[145] Plato, Epistol. vii. p. 350 E. ταῦτα εἶπον μεμισηκὼς τὴν περὶ Σικελίαν
πλάνην καὶ ἀτυχίαν, etc.
Xenokrates seems to have accompanied Plato to Sicily (Diogen. Laert. iv. 2, 1).

[146] Plutarch, De Adulator, et Amici Discrimine, p. 52 C.

[147] Plutarch, Dion, c. 13. Οὐ παύσῃ καταρώμενος ἡμῖν;

[148] Plutarch, Dion, c. 14. Ἔνιοι δὲ προσεποιοῦντο δυσχεραίνειν, εἰ

πρότερον μὲν Ἀθηναῖοι ναυτικαῖς καὶ πεζικαῖς δυνάμεσι δεῦρο πλεύσαντες
ἀπώλοντο καὶ διεφθάρησαν πρότερον ἢ λαβεῖν Συρακούσας, νυνὶ δὲ δι᾽ ἑ ν ὸ ς
σ ο φ ι σ τ ο ῦ καταλύουσι τὴν Διονυσίου τυραννίδα, etc.
Plato is here described as a Sophist, in the language of those who did not like
him. Plato, the great authority who is always quoted in disparagement of the
persons called Sophists, is as much entitled to the name as they, and is called so
equally by unfriendly commentators. I drew particular attention to this fact in my
sixty-eighth chapter (Vol. VIII.), where I endeavored to show that there was no
school, sect, or body of persons distinguished by uniformity of doctrine or
practice, properly called Sophists, and that the name was common to all literary
men or teachers, when spoken of in an unfriendly spirit.

[149] Plato, Epistol. vii. p. 330 B. Ἐγὼ δὲ πάντα ὑπέμενον, τὴν πρώτην
διάνοιαν φυλάττων ᾗπερ ἀφικόμην, εἴπως εἰς ἐπιθυμίαν ἔλθοι τ ῆ ς φ ι λ ο σ ό φ ο υ
ζ ω ῆ ς (Dionysius)—ὁ δ᾽ ἐνίκησεν ἀντιτείνων.
 [150] Plato, Epistol. vii. p. 332 E. Ἃ δὴ καὶ Διονυσίῳ συνεβουλεύομεν ἐγὼ καὶ
Δίων, ἐπειδὴ τὰ παρὰ τοῦ πατρὸς αὐτῷ συνεβεβήκει, οὕτως ἀνομιλήτῳ μὲν
παιδείας, ἀνομιλήτῳ δὲ συνουσιῶν τῶν προσηκουσῶν γεγονέναι, π ρ ῶ τ ο ν ἐπὶ
ταῦτα ὁρμήσαντα φίλους ἄλλους αὑτῷ τῶν οἰκείων ἅμα καὶ ἡλικιωτῶν καὶ
συμφώνους πρὸς ἀρετὴν κτήσασθαι, μ ά λ ι σ τ α δ ὲ α ὐ τ ὸ ν α ὑ τ ῷ , τ ο ύ τ ο υ
γὰρ αὐτὸν θαυμαστῶς ἐνδεᾶ γεγονέναι· λέγοντες οὐκ ἐναργῶς
ο ὕ τ ω ς — ο ὐ γ ὰ ρ ἦ ν ἀ σ φ α λ ὲ ς—ὡς οὕτω μὲν πᾶς ἀνὴρ αὑτόν τε καὶ ἐκείνους
ὧν ἂν ἡγεμὼν γένηται σώσει, μὴ ταύτῃ δὲ τραπόμενος τἀναντία πάντα ἀποτελεῖ·
πορευθεὶς δὲ ὡς λέγομεν, κ α ὶ ἑ α υ τ ὸ ν ἔ μ φ ρ ο ν α κ α ὶ σ ώ φ ρ ο ν α
π ο ι η σ ά μ ε ν ο ς, εἰ τὰς ἐξηρημωμένας Σικελίας πόλεις κατοικίσειε νόμοις τε
ξυνδήσειε καὶ πολιτείαις, etc.
Compare also p. 331 F.

[151] Horat. Satir. ii. 1, 17.

“Haud mihi deero

Cum res ipsa feret. Nisi dextro tempore, Flacci
Verba per attentam non ibunt Cæsaris aurem.
Cui male si palpere, recalcitrat undique tutus.”

[152] Plato, Epist. iii. 315 E. Φάσι δὲ οὐκ ὀλίγοι λέγειν σε πρός τινας τῶν
παρά σε πρεσβευόντων, ὡς ἄρα σοῦ ποτὲ λέγοντος ἀκούσας ἐγὼ μέλλοντος τάς
τε Ἑλληνίδας πόλεις ἐν Σικελίᾳ οἰκίζειν, καὶ Συρακουσίους ἐπικουφίσαι, τὴν ἀρχὴν
ἀντὶ τυραννίδος εἰς βασιλείαν μεταστήσαντα, τ α ῦ τ ᾽ ἄ ρ α σ ὲ μ ὲ ν τ ό τ ε , ὡ ς
σὺ φῇς, διεκώλυσα—νῦν δὲ Δίωνα διδάσκοιμι δρᾷν αὐτὰ, καὶ
τ ο ῖ ς δ ι α ν ο ή μ α σ ι τ ο ῖ ς σ ο ῖ ς τ ὴ ν σ ὴ ν ἀ ρ χ ὴ ν ἀφαιρούμεθά σε....
Ibid. p. 319 B. εἶπες δὲ καὶ μάλ᾽ ἀπλάστως γελῶν, εἰ μέμνημαι, ὡς
Παιδευθέντα με ἐκέλευες ποιεῖν πάντα ταῦτα, ἢ μὴ ποιεῖν.
Ἔ φ η ν ἐ γ ὼ Κ ά λ λ ι σ τ α μ ν η μ ο ν ε ῦ σ α ί σ ε.
Cornelius Nepos (Dion, c. 3) gives to Plato the credit, which belongs
altogether to Dion, of having inspired Dionysius with these ideas.

[153] Plutarch, De Adulator, et Amici Discrimine, p. 52 E. We may set against

this, however, a passage in one of the other treatises of Plutarch (Philosophand.
cum Principibus, p. 779 ad finem), in which he observes, that Plato, coming to
Sicily with the hope of converting his political doctrines into laws through the
agency of Dionysius, found the latter already corrupted by power, unsusceptible
of cure, and deaf to admonition.

[154] Plato, Phædon, c. 88. p. 89 D. Οὐκοῦν αἰσχρόν; καὶ δῆλον, ὅτι ἄνευ
τέχνης τῆς περὶ τἀνθρώπεια ὁ τοιοῦτος χρῆσθαι ἐπιχειρεῖ τοῖς ἀνθρώποις;
He is expounding the causes and growth of misanthropic dispositions; one of
the most striking passages in his dialogues.
 [155] Plutarch, Dion, c. 14, Plato, Epistol. vii. p. 333 C. Ὁ δὲ (Dionysius) τοῖς
διαβάλλουσι (ἐπίστευε) καὶ λέγουσιν ὡς ἐπιβουλεύων τῇ τυραννίδι Δίων πράττοι
πάντα ὅσα ἔπραττεν ἐν τῷ τότε χρόνῳ, ἵνα ὁ μὲν (Dionysius) παιδείᾳ δὴ τὸν νοῦν
κηληθεὶς ἀμελοῖ τῆς ἀρχῆς ἐπιτρέψας ἐκείνῳ, ὁ δὲ (Dion) σφετερίσαιτο, καὶ
Διονύσιον ἐκβάλοι ἐκ τῆς ἀρχῆς δόλῳ.

[156] Plato, Epistol. vii. p. 329 C. ἐλθὼν δὲ, οὐ γὰρ δεῖ μηκύνειν, εὗρον
στάσεως τὰ περὶ Διονύσιον μεστὰ ξύμπαντα καὶ διαβολῶν πρὸς τὴν τυραννίδα
Δίωνος πέρι· ἤμυνον μὲν οὖν καθ᾽ ὅσον ἠδυνάμην, σμικρὰ δ᾽ οἷός τε ἦ, etc.

[157] The story is found in Plutarch (Dion, c. 14), who refers to Timæus as
his authority. It is confirmed in the main by Plato, Epistol. vii. p. 329 D. μηνὶ δὴ
σχεδὸν ἴσως τετάρτῳ Δίωνα Διονύσιος, αἰτιώμενος ἐπιβουλεύειν τῇ τυραννίδι,
σμικρὸν εἰς πλοῖον ἐμβιβάσας, ἐξέβαλεν ἀτίμως.
Diodorus (xvi. 6) states that Dionysius sought to put Dion to death, and that
he only escaped by flight. But the version of Plato and Plutarch is to be preferred.
Justin (xxi. 1, 2) gives an account, different from all, of the reign and
proceedings of the younger Dionysius. I cannot imagine what authority he
followed. He does not even name Dion.

[158] Plato, Epistol. iii. p. 315 F.; Epist. vii. p. 329 D.; p. 340 A. Plutarch,
Dion, c. 15.

[159] Plato, Epist. vii. p. 329, 330.

[160] Plato, Epist. vii. p. 338 C.

[161] Plato, Epistol. iii. p. 317 B. C.

[162] Plato, Epist. vii. p. 338-346; Plutarch, Dion, c. 19. Æschines, the
companion of Sokrates along with Plato, is said to have passed a long time at
Syracuse with Dionysius, until the expulsion of that despot (Diogen. Laert. ii. 63).

[163] Plutarch, De Fortunâ Alex. Magn. p. 338 B. Δωρίδος ἐκ μητρὸς Φοίβου

κοινώμασι βλαστών.

[164] See a passage in Plato, Epistol. ii. p. 314 E.

[165] Plato, Epistol. iii. p. 318 A.; vii. p. 346, 347. Plutarch, Dion, c. 15, 16.

[166] Plutarch, Timoleon, c. 15—on the authority of Aristoxenus.

[167] Plato, Epistol. vii. p. 350 A. B.

 [168] Plato, Epistol. vii. p. 350 C. The return of Plato and his first meeting
with Dion is said to have excited considerable sensation among the spectators at
the festival (Diogenes Laert. iii. 25).
The Olympic festival here alluded to, must be (I conceive) that of 366 B. C.:
the same also in Epistol. ii. p. 310 D.

[169] Plutarch, Dion, c. 21; Cornel. Nepos, Dion, c. 4.

[170] Plutarch, Dion, c. 17; Athenæus, xi. p. 508. Plato appears also to have
received, when at Athens, pecuniary assistance remitted by Dionysius from
Syracuse, towards expenses of a similar kind, as well as towards furnishing a
dowry for certain poor nieces. Dion and Dionysius had both aided him (Plato,
Epistol. xiii. p. 361).
An author named Onêtor affirmed that Dionysius had given to Plato the
prodigious sum of eighty talents; a story obviously exaggerated (Diogenes Laert.
iii. 9).

[171] Plato, Epistol. vii. p. 350 F.

[172] Plato, Epistol. vii. p. 350. This is the account which Plato gives after
the death of Dion, when affairs had taken a disastrous turn, about the extent of
his own interference in the enterprise. But Dionysius supposed him to have been
more decided in his countenance of the expedition; and Plato’s letter addressed
to Dion himself, after the victory of the latter at Syracuse, seems to bear out that
supposition.
Compare Epistol. iii. p. 315 E.; iv. p. 320 A.

[173] Plutarch, Dion, c. 22. Eudemus was afterwards slain in one of the
combats at Syracuse (Aristotle apud Ciceron. Tusc. Disp. i. 25, 53).

[174] Plutarch, Dion, c. 23-25.

[175] Aristotel. Politic. v. 8, 17.

[176] See Orat. adv. Leptinem, s. 179. p. 506: an oration delivered about two
years afterwards; not long after the victory of Dion.
Compare Diodor. xvi. 9; Plutarch, Timoleon, c. 2.

[177] Plutarch, Dion, c. 22. Speusippus, from Athens, corresponded both

with Dion and with Dionysius at Syracuse; at least there was a correspondence
between them, read as genuine by Diogenes Laertius (iv. 1, 2, 5).

[178] Plato, Epistol. iii. p. 318 C.

Welcome to Our Bookstore - The Ultimate Destination for Book Lovers
Are you passionate about books and eager to explore new worlds of
knowledge? At our website, we offer a vast collection of books that
cater to every interest and age group. From classic literature to
specialized publications, self-help books, and children’s stories, we
have it all! Each book is a gateway to new adventures, helping you
expand your knowledge and nourish your soul
Experience Convenient and Enjoyable Book Shopping Our website is more
than just an online bookstore—it’s a bridge connecting readers to the
timeless values of culture and wisdom. With a sleek and user-friendly
interface and a smart search system, you can find your favorite books
quickly and easily. Enjoy special promotions, fast home delivery, and
a seamless shopping experience that saves you time and enhances your
love for reading.
Let us accompany you on the journey of exploring knowledge and
personal growth!

ebookgate.com