Textbook Geometry of Quantum States An Introduction To Quantum Entanglement Ingemar Bengtsson 2 Ebook All Chapter PDF

Geometry of Quantum States An
Introduction to Quantum Entanglement

Ingemar Bengtsson
Visit to download the full and correct content document:
https://textbookfull.com/product/geometry-of-quantum-states-an-introduction-to-quant
um-entanglement-ingemar-bengtsson-2/
More products digital (pdf, epub, mobi) instant
download maybe you interests ...
Geometry of Quantum States An Introduction to Quantum

Entanglement Ingemar Bengtsson
https://textbookfull.com/product/geometry-of-quantum-states-an-
introduction-to-quantum-entanglement-ingemar-bengtsson-2/
An Introduction to quantum optics and quantum

fluctuations First Edition. Edition Milonni
https://textbookfull.com/product/an-introduction-to-quantum-
optics-and-quantum-fluctuations-first-edition-edition-milonni/
An Introduction to Quantum Physics First Edition

Anthony P. French
physics-first-edition-anthony-p-french/
An Introduction To Quantum Field Theory First Edition

Michael E. Peskin
field-theory-first-edition-michael-e-peskin/
An Introduction to Quantum and Vassiliev Knot
Invariants David M. Jackson
https://textbookfull.com/product/an-introduction-to-quantum-and-
vassiliev-knot-invariants-david-m-jackson/
An Introduction to Quantum Transport in Semiconductors

1st Edition David K. Ferry
transport-in-semiconductors-1st-edition-david-k-ferry/
Introduction to Quantum Mechanics David J. Griffiths
https://textbookfull.com/product/introduction-to-quantum-
mechanics-david-j-griffiths/
An Introduction to Quantum Optics Photon and Biphoton

Physics 2nd Edition Yanhua Shih
optics-photon-and-biphoton-physics-2nd-edition-yanhua-shih/
Introduction to Quantum Field Theory with Applications

to Quantum Gravity 1st Edition Iosif L Buchbinder Ilya
Shapiro
https://textbookfull.com/product/introduction-to-quantum-field-
theory-with-applications-to-quantum-gravity-1st-edition-iosif-l-
buchbinder-ilya-shapiro/
G E O M E T RY O F QUA N T U M S TAT E S
An Introduction to Quantum Entanglement
Quantum information theory is a branch of science at the frontiers of physics, mathematics

and information science, and offers a variety of solutions that are impossible using classical
theory. This book provides a detailed introduction to the key concepts used in processing
quantum information and reveals that quantum mechanics is a generalisation of classical
probability theory.
The second edition contains new sections and entirely new chapters: the hot topic of
multipartite entanglement; in-depth discussion of the discrete structures in finite dimen-
sional Hilbert space, including unitary operator bases, mutually unbiased bases, symmetric
informationally complete generalised measurements, discrete Wigner functions and unitary
designs; the Gleason and Kochen–Specker theorems; the proof of the Lieb conjecture; the
measure concentration phenomenon; and the Hastings’ non-additivity theorem.
This richly-illustrated book will be useful to a broad audience of graduates and
researchers interested in quantum information theory. Exercises follow each chapter,
with hints and answers supplied.
I N G E M A R B E N G T S S O N is a professor of physics at Stockholm University. After

gaining a Ph.D. in theoretical physics from the University of Göteborg (1984), he held
post-doctoral positions at CERN, Geneva, and Imperial College, London. He returned to
Göteborg in 1988 as a research assistant at Chalmers University of Technology, before
taking up a position as Lecturer in Physics at Stockholm University in 1993. He was
appointed Professor of Physics in 2000. Professor Bengtsson is a member of the Swedish
Physical Society and a former board member of its Divisions for Particle Physics and for
Gravitation. His research interests are related to geometry, in the forms of classical general
relativity and quantum theory.
is a professor at the Institute of Physics, Jagiellonian Univer-

K A R O L Ż Y C Z K O W S K I
sity, Kraków, Poland and also at the Center for Theoretical Physics, Polish Academy of
Sciences, Warsaw. He gained his Ph.D. and habilitation in theoretical physics at Jagiel-
lonian University, and has followed this with a Humboldt Fellowship in Essen, a Fulbright
Fellowship at the University of Maryland, College Park, and a visiting research position
at the Perimeter Institute, Waterloo, Ontario. He has been a docent at the Academy of
Sciences since 1999 and a full professor at Jagiellonian University since 2004. Professor
Życzkowski is a member of the Polish Physical Society and Academia Europaea. He works
on quantum information, dynamical systems and chaos, quantum and statistical physics,
applied mathematics and the theory of voting.
G E O M E T RY O F Q UA N T U M S TAT E S
An Introduction to Quantum Entanglement
SECOND EDITION
INGEMAR BENGTSSON
Stockholm University, Sweden
K A RO L Ż Y C Z KOW S K I
Jagiellonian University, Poland
University Printing House, Cambridge CB2 8BS, United Kingdom
One Liberty Plaza, 20th Floor, New York, NY 10006, USA
477 Williamstown Road, Port Melbourne, VIC 3207, Australia
4843/24, 2nd Floor, Ansari Road, Daryaganj, Delhi – 110002, India
79 Anson Road, #06-04/06, Singapore 079906
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/9781107026254
DOI: 10.1017/9781139207010
© Ingemar Bengtsson and Karol Życzkowski 2017
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 2006
Second edition 2017
Printed in the United Kingdom by Clays, St Ives plc
A catalogue record for this publication is available from the British Library.
ISBN 978-1-107-02625-4 Hardback
ISBN 978-1-107-65614-7 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
Preface page xi
Acknowledgements xv
1 Convexity, colours and statistics 1
1.1 Convex sets 1
1.2 High dimensional geometry 9
1.3 Colour theory 15
1.4 What is ‘distance’? 18
1.5 Probability and statistics 25
Problems 28
2 Geometry of probability distributions 29
2.1 Majorisation and partial order 29
2.2 Shannon entropy 35
2.3 Relative entropy 41
2.4 Continuous distributions and measures 46
2.5 Statistical geometry and the Fisher–Rao metric 48
2.6 Classical ensembles 55
2.7 Generalised entropies 56
Problems 62
3 Much ado about spheres 63
3.1 Spheres 63
3.2 Parallel transport and statistical geometry 68
3.3 Complex, Hermitian, and Kähler manifolds 74
3.4 Symplectic manifolds 80
3.5 The Hopf fibration of the 3-sphere 82
3.6 Fibre bundles and their connections 88
3.7 The 3-sphere as a group 95
v
vi Contents
3.8 Cosets and all that 99

Problems 101
4 Complex projective spaces 103
4.1 From art to mathematics 103
4.2 Complex projective geometry 108
4.3 Complex curves, quadrics and the Segre embedding 110
4.4 Stars, spinors and complex curves 114
4.5 The Fubini–Study metric 117
4.6 CPn illustrated 122
4.7 Symplectic geometry and the Fubini–Study measure 129
4.8 Fibre bundle aspects 131
4.9 Grassmannians and flag manifolds 133
Problems 137
5 Outline of quantum mechanics 138
5.1 Quantum mechanics 138
5.2 Qubits and Bloch spheres 140
5.3 The statistical and the Fubini–Study distances 143
5.4 A real look at quantum dynamics 146
5.5 Time reversals 150
5.6 Classical and quantum states: a unified approach 154
5.7 Gleason and Kochen–Specker 157
Problems 164
6 Coherent states and group actions 165
6.1 Canonical coherent states 165
6.2 Quasi-probability distributions on the plane 170
6.3 Bloch coherent states 178
6.4 From complex curves to SU(K) coherent states 184
6.5 SU(3) coherent states 186
Problems 190
7 The stellar representation 191
7.1 The stellar representation in quantum mechanics 191
7.2 Orbits and coherent states 194
7.3 The Husimi function 197
7.4 Wehrl entropy and the Lieb conjecture 202
7.5 Generalised Wehrl entropies 205
7.6 Random pure states 207
7.7 From the transport problem to the Monge distance 214
Problems 217
Contents vii
8 The space of density matrices 219

8.1 Hilbert–Schmidt space and positive operators 220
8.2 The set of mixed states 223
8.3 Unitary transformations 226
8.4 The space of density matrices as a convex set 229
8.5 Stratification 234
8.6 Projections and cross-sections 239
8.7 An algebraic afterthought 244
8.8 Summary 247
Problems 248
9 Purification of mixed quantum states 249
9.1 Tensor products and state reduction 250
9.2 The Schmidt decomposition 252
9.3 State purification and the Hilbert–Schmidt bundle 255
9.4 A first look at the Bures metric 258
9.5 Bures geometry for N = 2 261
9.6 Further properties of the Bures metric 263
Problems 266
10 Quantum operations 267
10.1 Measurements and POVMs 267
10.2 Algebraic detour: matrix reshaping and reshuffling 274
10.3 Positive and completely positive maps 278
10.4 Environmental representations 283
10.5 Some spectral properties 286
10.6 Unital and bistochastic maps 288
10.7 One qubit maps 291
Problems 295
11 Duality: maps versus states 296
11.1 Positive and decomposable maps 296
11.2 Dual cones and super-positive maps 304
11.3 Jamiołkowski isomorphism 305
11.4 Quantum maps and quantum states 309
Problems 311
12 Discrete structures in Hilbert space 313
12.1 Unitary operator bases and the Heisenberg groups 313
12.2 Prime, composite and prime power dimensions 317
12.3 More unitary operator bases 322
viii Contents
12.4 Mutually unbiased bases 326

12.5 Finite geometries and discrete Wigner functions 333
12.6 Clifford groups and stabilizer states 338
12.7 Some designs 343
12.8 SICs 347
Problems 353
13 Density matrices and entropies 355
13.1 Ordering operators 356
13.2 Von Neumann entropy 359
13.3 Quantum relative entropy 365
13.4 Other entropies 369
13.5 Majorisation of density matrices 372
13.6 Proof of the Lieb conjecture 376
13.7 Entropy dynamics 380
Problems 384
14 Distinguishability measures 386
14.1 Classical distinguishability measures 386
14.2 Quantum distinguishability measures 391
14.3 Fidelity and statistical distance 396
Problems 401
15 Monotone metrics and measures 402
15.1 Monotone metrics 402
15.2 Product measures and flag manifolds 408
15.3 Hilbert–Schmidt measure 410
15.4 Bures measure 414
15.5 Induced measures 416
15.6 Random density matrices 418
15.7 Random operations 422
15.8 Concentration of measure 426
Problems 431
16 Quantum entanglement 433
16.1 Introducing entanglement 433
16.2 Two-qubit pure states: entanglement illustrated 436
16.3 Maximally entangled states 441
16.4 Pure states of a bipartite system 444
16.5 A first look at entangled mixed states 454
16.6 Separability criteria 460
Contents ix
16.7 Geometry of the set of separable states 466

16.8 Entanglement measures 471
16.9 Two-qubit mixed states 481
Problems 491
17 Multipartite entanglement 493
17.1 How much is three larger than two? 493
17.2 Botany of states 494
17.3 Permutation symmetric states 498
17.4 Invariant theory and quantum states 503
17.5 Monogamy relations and global multipartite entanglement 515
17.6 Local spectra and the momentum map 518
17.7 AME states and error-correcting codes 525
17.8 Entanglement in quantum spin systems 532
Problems 542
Epilogue 544
Appendix A Basic notions of differential geometry 545
A.1 Differential forms 545
A.2 Riemannian curvature 546
A.3 A key fact about mappings 547
Appendix B Basic notions of group theory 549
B.1 Lie groups and Lie algebras 549
B.2 SU(2) 550
B.3 SU(N) 551
B.4 Homomorphisms between low-dimensional groups 552
Appendix C Geometry: do it yourself 553
Appendix D Hints and answers to the exercises 557
References 572
Index 614
Preface
Preface to the first edition

The geometry of quantum states is a highly interesting subject in itself. It is also
relevant in view of possible applications in the rapidly developing fields of quantum
information and quantum computing.
But what is it? In physics words like ‘states’ and ‘system’ are often used. Skip-
ping lightly past the question of what these words mean – it will be made clear
by practice – it is natural to ask for the properties of the space of all possible
states of a given system. The simplest state space occurs in computer science: a
‘bit’ has a space of states that consists simply of two points, representing on and
off. In probability theory the state space of a bit is really a line segment, since the
bit may be ‘on’ with some probability between zero and one. In general the state
spaces used in probability theory are ‘convex hulls’ of a discrete or continuous set of
points. The geometry of these simple state spaces is surprisingly subtle – especially
since different ways of distinguishing probability distributions give rise to different
notions of distance, each with their own distinct operational meaning. There is an
old idea saying that a geometry can be understood once it is understood what linear
transformations are acting on it, and we will see that this is true here as well.
The state spaces of classical mechanics are – at least from the point of view that
we adopt – just instances of the state spaces of classical probability theory, with the
added requirement that the sample spaces (whose ‘convex hull’ we study) are large
enough, and structured enough, so that the transformations acting on them include
canonical transformations generated by Hamiltonian functions.
In quantum theory the distinction between probability theory and mechanics
goes away. The simplest quantum state space is these days known as a ‘qubit’. There
are many physical realisations of a qubit, from silver atoms of spin 1/2 (assuming
that we agree to measure only their spin) to the qubits that are literally designed
in today’s laboratories. As a state space a qubit is a three–dimensional ball; each
diameter of the ball is the state space of some classical bit, and there are so many
xi
xii Preface
bits that their sample spaces conspire to form a space – namely the surface of the
ball – large enough to carry the canonical transformations that are characteristic of
mechanics. Hence the word quantum mechanics.
It is not particularly difficult to understand a three–dimensional ball, or to see
how this description emerges from the usual description of a qubit in terms of a
complex two-dimensional Hilbert space. In this case we can take the word geometry
literally – there will exist a one-to-one correspondence between pure states of
the qubit and the points of the surface of the Earth. Moreover, at least as far
as the surface is concerned, its geometry has a statistical meaning when transcribed
to the qubit (although we will see some strange things happening in the interior).
As the dimension of the Hilbert space goes up, the geometry of the state spaces
becomes very intricate, and qualitatively new features arise – such as the subtle way
in which composite quantum systems are represented. Our purpose is to describe
this geometry. We believe it is worth doing. Quantum state spaces are more won-
derful than classical state spaces, and in the end composite systems of qubits may
turn out to have more practical applications than the bits themselves already have.
A few words about the contents of our book. As a glance at the Contents will
show, there are 17 chapters, culminating in a long chapter on ‘entanglement’. Along
the way, we take our time to explore many curious byways of geometry. We expect
that you – the reader – are familiar with the principles of quantum mechanics at
the advanced undergraduate level. We do not really expect more than that, and
should you be unfamiliar with quantum mechanics we hope that you will find some
sections of the book profitable anyway. You may start reading any chapter – if
you find it incomprehensible we hope that the cross-references and the index will
enable you to see what parts of the earlier chapters may be helpful to you. In the
unlikely event that you are not even interested in quantum mechanics, you may
perhaps enjoy our explanations of some of the geometrical ideas that we come
across.
Of course there are limits to how independent the chapters can be of each other.
Convex set theory (Chapter 1) pervades all statistical theories, and hence all our
chapters. The ideas behind the classical Shannon entropy and the Fisher–Rao
geometry (Chapter 2) must be brought in to explain quantum mechanical entropies
(Chapter 12) and quantum statistical geometry (Chapters 9 and 13). Sometimes
we have to assume a little extra knowledge on the part of the reader, but since no
chapter in our book assumes that all the previous chapters have been understood,
this should not pose any great difficulties.
We have made a special effort to illustrate the geometry of quantum mechan-
ics. This is not always easy, since the spaces that we encounter more often than
not have a dimension higher than three. We have simply done the best we could.
To facilitate self-study each chapter concludes with problems for the reader, while
some additional geometrical exercises are presented in the final appendix.
Preface xiii
Once and for all, let us say that we limit ourselves to finite-dimensional state
spaces. We do this for two reasons. One of them is that it simplifies the story very
much, and the other is that finite-dimensional systems are of great independent
interest in real experiments.
The entire book may be considered as an introduction to quantum entanglement.
This very non-classical feature provides a key resource for several modern applica-
tions of quantum mechanics including quantum cryptography, quantum computing
and quantum communication. We hope that our book may be useful for graduate
and postgraduate students of physics. It is written first of all for readers who do not
read the mathematical literature everyday, but we hope that students of mathematics
and of the information sciences will find it useful as well, since they also may wish
to learn about quantum entanglement.
We have been working on the book for about five years. Throughout this time
we enjoyed the support of Stockholm University, the Jagiellonian University in Cra-
cow, and the Center for Theoretical Physics of the Polish Academy of Sciences in
Warsaw. The book was completed in Waterloo during our stay at the Perimeter Insti-
tute for Theoretical Physics. The motto at its main entrance – AOYATO
EPI ŴEMETPIA MHEI EIIT – proved to be a lucky omen indeed,
and we are pleased to thank the Institute for creating optimal working conditions
for us, and to thank all the knowledgeable colleagues working there for their help,
advice, and support.
We are grateful to Erik Aurell for his commitment to Polish–Swedish colla-
boration; without him the book would never have been started. It is a pleasure
to thank our colleagues with whom we have worked on related projects: Johan
Brännlund, Åsa Ericsson, Sven Gnutzmann, Marek Kuś, Florian Mintert,
Magdalena Sinołȩcka, Hans-Jürgen Sommers and Wojciech Słomczyński. We
are grateful to them, and to many others who helped us to improve the manuscript.
If it never reached perfection, it was our fault, not theirs. Let us mention some
of the others: Robert Alicki, Anders Bengtsson, Iwo Białynicki-Birula, Rafał
Demkowicz-Dobrzański, Johan Grundberg, Sören Holst, Göran Lindblad, and
Marcin Musz. We have also enjoyed constructive interactions with Matthias
Christandl, Jens Eisert, Peter Harremoës, Michał, Paweł and Ryszard Horodeccy,
Vivien Kendon, Artur Łoziński, Christian Schaffner, Paul Slater, and William
Wootters.
Five other people provided indispensable support: Martha and Jonas in Stock-
holm, and Jolanta, Jaś, and Marysia in Cracow.
Waterloo, 12 March 2005 Ingemar Bengtsson
Karol Życzkowski
xiv Preface
Preface to the second edition

More than a decade has passed since we completed the first edition of the book.
Much has happened during it. We have not tried to take all recent valuable contribu-
tions into account, but some of them can be found here. The Lieb conjecture for spin
coherent states has been proved by Lieb and Solovej, and an important additivity
conjecture for quantum channels has been disproved. Since these conjectures were
discussed at some length in the first edition we felt we had to say more about
them. However, our main concern with this second edition has been to improve
explanations and to remove mistakes (while trying not to introduce new ones).
There are two new chapters: one of them is centred around the finite Weyl–
Heisenberg group, and the discrete structures it gives rise to. The other tries to sur-
vey the vast field of multipartite entanglement, which was relegated to a footnote or
two in the first edition. There are also a few new sections: They concern Gleason’s
theorem and quantum contextuality, cross–sections and projections of the body of
mixed quantum states, and the concentration of measure in high dimensions (which
is needed to understand what is now Hastings’ non-additivity theorem).
We are grateful to several people who helped us with the work on this
edition. It is a pleasure to thank Felix Huber, Pedro Lamberti and Marcus Müller
for their comments on the first edition. We are indebted to Radosław Adam-
czak, Ole Andersson, Marcus Appleby, Adán Cabello, Runyao Duan, Shmuel
Friedland, Dardo Goyeneche, David Gross, Michał and Paweł Horodeccy, Ted
Jacobson, David Jennings, Marek Kuś, Ion Nechita, Zbigniew Puchała, Wojciech
Roga, Adam Sawicki, Stanisław Szarek, Stephan Weis, Andreas Winter, Iwona
Wintrowicz, and Huangjun Zhu, for reading some fragments of the text and
providing us with valuable remarks. Had they read all of it, we are sure that it
would have been perfect! We thank Kate Blanchfield, Piotr Gawron, Lia Pugliese,
Konrad Szymański, and Maria Życzkowska, for preparing for us two-dimensional
figures and three-dimensional printouts, models and photos for the book.
We thank Simon Capelin and his team at Cambridge University Press for the
long and fruitful work with us. Most of all we thank Martha and Jolanta for their
understanding and infinite patience.
Toruń, 11 June 2016 Ingemar Bengtsson
Karol Życzkowski
Acknowledgements
Figures 4.10–13 and 16.1–3 have appeared already in our reference [112], which
was published in the International Journal of Physics A17 Issue 31 (2002), coau-
thored with Johan Brännlund. They appear here with the kind permission of the
copyright holder, World Scientific.
Figure 5.2 and the Figure accompanying the Hint to Problem 5.2 are reprinted from
Dr. Kate Blanchfield’s PhD thesis, with the kind permission of its copyright holder
and author.
xv
1
Convexity, colours and statistics
1.1 Convex sets
What picture does one see, looking at a physical theory from a distance, so that the details
disappear? Since quantum mechanics is a statistical theory, the most universal picture which
remains after the details are forgotten is that of a convex set.
—Bogdan Mielnik1
Our object is to understand the geometry of the set of all possible states of a
quantum system that can occur in nature. This is a very general question; especially
since we are not trying to define ‘state’ or ‘system’ very precisely. Indeed we will
not even discuss whether the state is a property of a thing, or of the preparation of a
thing, or of a belief about a thing. Nevertheless we can ask what kind of restrictions
are needed on a set if it is going to serve as a space of states in the first place. There
is a restriction that arises naturally both in quantum mechanics and in classical
statistics: the set must be a convex set. The idea is that a convex set is a set such that
one can form ‘mixtures’ of any pair of points in the set. This is, as we will see, how
probability enters (although we are not trying to define ‘probability’ either).
From a geometrical point of view a mixture of two states can be defined as a
point on the segment of the straight line between the two points that represent the
states that we want to mix. We insist that given two points belonging to the set of
states, the straight line segment between them must belong to the set too. This is
certainly not true of any set. But before we can see how this idea restricts the set
of states we must have a definition of ‘straight lines’ available. One way to proceed
is to regard a convex set as a special kind of subset of a flat Euclidean space En .
Actually we can get by with somewhat less. It is enough to regard a convex set as a
subset of an affine space. An affine space is just like a vector space, except that no
1 Reproduced from [659].
1
2 Convexity, colours and statistics
special choice of origin is assumed. The straight line through the two points x1 and
x2 is defined as the set of points
x = μ1 x1 + μ2 x2 , μ1 + μ2 = 1. (1.1)
If we choose a particular point x0 to serve as the origin, we see that this is a one-
parameter family of vectors x − x0 in the plane spanned by the vectors x1 − x0 and
x2 − x0 . Taking three different points instead of two in Eq. (1.1) we define a plane,
provided the three points do not belong to a single line. A k-dimensional k–plane
is obtained by taking k + 1 generic points, where k < n. For k = n we describe
the entire space En . In this way we may introduce barycentric coordinates into an
n-dimensional affine space. We select n + 1 points xi , so that an arbitrary point x
can be written as
x = μ0 x0 + μ1 x1 + . . . + μn xn , μ0 + μ1 + . . . + μn = 1. (1.2)
The requirement that the barycentric coordinates μi add up to one ensures that they
are uniquely defined by the point x. (It also means that the barycentric coordinates
are not coordinates in the ordinary sense of the word, but if we solve for μ0 in terms
of the others then the remaining independent set is a set of n ordinary coordinates
for the n-dimensional space.) An affine map is a transformation that takes lines to
lines and preserves the relative length of line segments lying on parallel lines. In
equations an affine map is a combination of a linear transformation described by a
matrix A with a translation along a constant vector b, so x = Ax + b, where A is
an invertible matrix.
By definition a subset S of an affine space is a convex set if for any pair of points
x1 and x2 belonging to the set it is true that the mixture x also belongs to the set,
where
x = λ1 x1 + λ2 x2 , λ1 + λ2 = 1, λ1 , λ2 ≥ 0. (1.3)
Here λ1 and λ2 are barycentric coordinates on the line through the given pair
of points; the extra requirement that they be positive restricts x to belong to the
segment of the line lying between the pair of points.
It is natural to use an affine space as the ‘container’ for the convex sets since
convexity properties are preserved by general affine transformations. On the other
hand it does no harm to introduce a flat metric on the affine space, turning it into
an Euclidean space. There may be no special significance attached to this notion
of distance, but it helps in visualizing what is going on. See Figures 1.1 and 1.2.
From now on, we will assume that our convex sets sit in Euclidean space, whenever
it is convenient to do so.
Intuitively a convex set is a set such that one can always see the entire set from
whatever point in the set one happens to be sitting at. Still they can come in a variety
of interesting shapes. We will need a few definitions. First, given any subset of the
1.1 Convex sets 3
Figure 1.1 Three convex sets, two of which are affine transformations of each
other. The new moon is not convex. An observer in Singapore will find the new
moon tilted but still not convex, since convexity is preserved by rotations.
Figure 1.2 The convex sets we will consider are either convex bodies (like the
simplex on the left, or the more involved example in the centre) or convex cones
with compact bases (an example is shown on the right).
affine space we define the convex hull of this subset as the smallest convex set that
contains the set. The convex hull of a finite set of points is called a convex polytope.
If we start with p + 1 points that are not confined to any (p − 1)-dimensional
subspace then the convex polytope is called a p-simplex. The p-simplex consists of
all points of the form
x = λ0 x0 + λ1 x1 + . . . + λp xp , λ0 + λ1 + . . . + λp = 1, λi ≥ 0. (1.4)
(The barycentric coordinates are all non–negative.) The dimension of a convex set
is the largest number n such that the set contains an n-simplex. When discussing
a convex set of dimension n we usually assume that the underlying affine space
also has dimension n, to ensure that the convex set possesses interior points (in the
sense of point set topology). A closed and bounded convex set that has an interior
is known as a convex body.
The intersection of a convex set with some lower dimensional subspace of the
affine space is again a convex set. Given an n-dimensional convex set S there is also
a natural way to increase its dimension with one: choose a point y not belonging
to the n-dimensional affine subspace containing S. Form the union of all the rays
(in this chapter a ray means a half line), starting from y and passing through S.
Figure 1.3 Left: a convex cone and its dual, both regarded as belonging to
Euclidean 2-space. Right: a self dual cone, for which the dual cone coincides with
the original. For an application of this construction see Figure 11.6.
Figure 1.4 A convex body is homeomorphic to a sphere.
The result is called a convex cone and y is called its apex, while S is its base. A ray
is in fact a one dimensional convex cone. A more interesting example is obtained by
first choosing a p-simplex and then interpreting the points of the simplex as vectors
starting from an origin O not lying in the simplex. Then the p + 1 dimensional set
of points
x = λ0 x0 + λ1 x1 + . . . + λp xp , λi ≥ 0 (1.5)
is a convex cone. Convex cones have many nice properties, including an inbuilt
partial order among its points: x ≤ y if and only if y − x belongs to the cone. Linear
maps to R that take positive values on vectors belonging to a convex cone form
a dual convex cone in the dual vector space. Since we are in the Euclidean vector
space En , we can identify the dual vector space with En itself. If the two cones agree
the convex cone is said to be self dual. See Figure 1.3. One self dual convex cone
that will appear now and again is the positive orthant or hyperoctant of En , defined
as the set of all points whose Cartesian coordinates are non-negative. We use the
notation x ≥ 0 to denote the fact that x belongs to the positive orthant.
From a purely topological point of view all convex bodies are equivalent to an
n–dimensional ball. To see this choose any point x0 in the interior and then for every
point in the boundary draw a ray starting from x0 and passing through the boundary
point (as in Figure 1.4). It is clear that we can make a continuous transformation
1.1 Convex sets 5
of the convex body into a ball with radius one and its centre at x0 by moving the
points of the container space along the rays.
Convex bodies and convex cones with compact bases are the only convex sets
that we will consider. Convex bodies always contain some special points that cannot
be obtained as mixtures of other points – whereas a half space does not! These
points are called extreme points by mathematicians and pure points by physicists
(actually, originally by Weyl), while non-pure points are called mixed. In a convex
cone the rays from the apex through the pure points of the base are called extreme
rays; a point x lies on an extreme ray if and only if y ≤ x ⇒ y = λx with
λ between zero and one. A subset F of a convex set that is stable under mix-
ing and purification is called a face of the convex set. What the phrase means is
that if
x = λx1 + (1 − λ)x2 , 0<λ<1 (1.6)
then x lies in F if and only if x1 and x2 lie in F. The ‘only if’ part of the definition
forces x to lie on the boundary of the set. A face of dimension k is a k-face. A 0-face
is an extreme point, and an (n − 1)-face is also known as a facet. It is interesting to
observe that the set of all faces on a convex body form a partially ordered set; we
say that F1 ≤ F2 if the face F1 is contained in the face F2 . It is a partially ordered
set of the special kind known as a lattice, which means that a given pair of faces
always have a greatest lower bound (perhaps the empty set) and a lowest greater
bound (perhaps the convex body itself).
To stem the tide of definitions let us quote two theorems that have an ‘obvious’
ring to them when they are stated abstractly but which are surprisingly useful in
practice:
Minkowski’s theorem. Any convex body is the convex hull of its pure points.
Carathéodory’s theorem. Any point in an n-dimensional convex set X can be
expressed as a convex combination of at most n + 1 pure points in X .
Thus any point x of a convex body S may be expressed as a convex combination of
pure points:

p

x= λi x i , λi ≥ 0, p ≤ n + 1, λi = 1. (1.7)
i=1 i
This equation is quite different from Eq. (1.2) that defined the barycentric coordi-
nates of x in terms of a fixed set of points xi , because – with the restriction that all
the coefficients be non-negative – it may be impossible to find a finite set of xi so
that every x in the set can be written in this new form. An obvious example is a
circular disk. Given x one can always find a finite set of pure points xi so that the
Figure 1.5 In a simplex a point can be written as a mixture in one and only one
way. In general the rank of a point is the minimal number of pure points needed
in the mixture; the rank may change in the interior of the set as shown in the
rightmost example. The set on the right has two non-exposed points which form
faces of their own. Note also that its insphere is not unique.
equation holds, but that is a different thing. The exact number of pure points one
needs is related to the face structure of the body, as one can see from the proof of
Carathéodory’s theorem (which we give as Problem 1.1.)
It is evident that the pure points always lie in the boundary of the convex set, but
the boundary often contains mixed points as well. The simplex enjoys a very special
property, which is that any point in the simplex can be written as a mixture of pure
points in one and only one way (as in Figure 1.5). This is because for the simplex
the coefficients in Eq. (1.7) are barycentric coordinates and the result follows from
the uniqueness of the barycentric coordinates of a point. No other convex set has
this property. The rank of a point x is the minimal number p needed in the convex
combination (1.7). By definition the pure points have rank one. In a simplex the
edges have rank two, the faces have rank three, and so on, while all the points in
the interior have maximal rank. From Eq. (1.7) we see that the maximal rank of any
point in a convex body in Rn does not exceed n + 1. In a ball all interior points have
rank two and all points on the boundary are pure, regardless of the dimension of the
ball. It is not hard to find examples of convex sets where the rank changes as we
move around in the interior of the set (see Figure 1.5).
The simplex has another quite special property, namely that its lattice of faces is
self dual. We observe that the number of k-faces in an n dimensional simplex is

n+1 n+1
= . (1.8)
k+1 n−k
Hence the set of n−k−1 dimensional faces can be put in one-to-one correspondence
with the set of k-faces. In particular, the pure points (k = 0) can be put in one-to-one
correspondence with the set of facets (by definition, the n − 1 dimensional faces).
1.1 Convex sets 7
Figure 1.6 Support hyperplanes of a convex set.
For this, and other, reasons its lattice of subspaces will have some exceptional
properties, turning it into what is technically known as a Boolean lattice.2
There is a useful dual description of convex sets in terms of supporting hyper-
planes. A support hyperplane of S is a hyperplane that intersects the set and which is
such that the entire set lies in one of the closed half spaces formed by the hyperplane
(see Figure 1.6). Hence a support hyperplane just touches the boundary of S, and
one can prove that there is a support hyperplane passing through every point of
the boundary of a convex body. By definition a regular point is a point on the
boundary that lies on only one support hyperplane, a regular support hyperplane
meets the set in only one point, and the entire convex set is regular if all its boundary
points as well as all its support hyperplanes are regular. So a ball is regular, while
a convex polytope or a convex cone is not – indeed all the support hyperplanes of
a convex cone pass through its apex. A face is said to be exposed if it equals the
intersection of the convex set and some support hyperplane. Convex polytopes arise
as the intersection of a finite number of closed half-spaces in Rn , and any pure point
of a convex polytope saturates n of the inequalities that define the half-spaces; again
a statement with an ‘obvious’ ring that is useful in practice.
In a flat Euclidean space a linear function to the real numbers takes the form
x → a · x, where a is some constant vector. Geometrically, this defines a family of
parallel hyperplanes. We have the important
Hahn–Banach separation theorem. Given a convex body and a point x0 that
does not belong to it. Then one can find a linear function f and a constant k such
that f (x) > k for all points belonging to the convex body, while f (x0 ) < k.
This is again almost obvious if one thinks in terms of hyperplanes.
It is useful to know a bit more about dual convex sets. For definiteness let us
start out with a three dimensional vector space in which a point is represented by a
vector y. Then its dual plane is the set of vectors x such that
x · y = −1. (1.9)
2 Because it is related to what George Boole thought were the laws of thought; see Varadarajan’s book [916] on
quantum logic for these things.
Figure 1.7 A square is dual to another square; on the left we see how the points
on the edge at the top define a corner at the bottom of the dual square. To the right
we see a more complicated convex set with non-exposed faces (that are points). Its
dual has non-polyhedral corners. In both cases the unit circle is shown dashed.
The constant on the right hand side was set to −1 for convenience. The dual of a
line is the intersection of a one-parameter family of planes dual to the points on the
line. This is in itself a line. The dual of a plane is a point, while the dual of a curved
surface is another curved surface – the envelope of the planes that are dual to the
points on the original surface. To define the dual of a convex body with a given
boundary we change the definition slightly, and include all points on one side of the
dual planes in the dual. Thus the dual X ◦ of a convex body X is defined to be
X ◦ = {x | c + x · y ≥ 0 ∀y ∈ X}, (1.10)
where c is a number that was set equal to 1 above (and also when drawing
Figure 1.7). The dual of a convex body including the origin is the intersection
of the half-spaces defined by the pure points y of X. The dual of the dual of a
body that includes the origin is equal to the convex hull of the original body. If we
enlarge a convex body the conditions on the dual become more stringent, and hence
the dual shrinks. The dual of a sphere centred at the origin is again a sphere, so a
sphere (of suitable radius) is self dual. The dual of a cube is an octahedron. The
dual of a regular tetrahedron is another copy of the original tetrahedron, possibly
of a different size. The copy can be made to coincide with the original by means of
an affine transformation, hence the tetrahedron is a self dual body.3
We will find much use for the concept of convex functions. A real function f (x)
defined on a closed convex subset X of Rn is called convex, if for any x, y ∈ X and
λ ∈ [0, 1] it satisfies
f (λx + (1 − λ)y) ≤ λf (x) + (1 − λ)f (y). (1.11)
3 To readers who wish to learn more about convex sets – or who wish to see proofs of the various assertions that
we left unproved – we recommend the book by Eggleston [283].
(a) (b)
0.5 0.5
f(x) −f(x)
0 0
−0.5 −0.5
0 0.5 1 0 0.5 1
x x
Figure 1.8 (a): the convex function f (x) = x ln x (b): the concave function
g(x) = −x ln x. The names stem from the shaded epigraphs of the functions which
are convex and concave, respectively.
The name refers to the fact that the epigraph of a convex function, that is the region
lying above the curve f (x) in the graph, is convex. Applying the inequality k − 1
times we see that
⎛ ⎞
k k
f⎝ λj xj ⎠ ≤ λj f (xj ), (1.12)
j=1 j=1
k
where xj ∈ X and the nonnegative weights sum to unity, j=1 λj = 1. If a function
f from R to R is differentiable, it is convex if and only if
f (y) − f (x) ≥ (y − x) f (x). (1.13)
If f is twice differentiable it is convex if and only if its second derivative is non-

negative. For a function of several variables to be convex, the matrix of second
derivatives must be positive definite. In practice, this is a very useful criterion. A
function f is called concave if −f is convex.
One of the main advantages of convex functions is that it is (comparatively) easy
to study their minima and maxima. A minimum of a convex function is always
a global minimum, and it is attained on some convex subset of the domain of
definition X. If X is not only convex but also compact, then the global maximum
sits at an extreme point of X.
1.2 High dimensional geometry

In quantum mechanics the spaces we encounter are often of very high dimension;
even if the dimension of Hilbert space is small the dimension of the space of density
matrices will be high. Our intuition on the other hand is based on two and three
dimensional spaces, and frequently leads us astray. We can improve ourselves by
asking some simple questions about convex bodies in flat space. We choose to
look at balls, cubes and simplices for this purpose. A flat metric is assumed. Our
questions will concern the inspheres and outspheres of these bodies (defined as the
largest inscribed sphere and the smallest circumscribed sphere, respectively). For
any convex body the outsphere is uniquely defined, while the insphere is not – one
can show that the upper bound on the radius of inscribed spheres is always attained
by some sphere, but there may be several of those.
Let us begin with the surface of a ball, namely the n-dimensional sphere. In
equations a sphere of radius r is given by the set
X02 + X12 + · · · + Xn2 = r2 (1.14)
in an n + 1 dimensional flat space En+1 . A sphere of radius one is denoted Sn . The

sphere can be parametrized by the angles φ, θ1 , . . ., θn−1 according to
⎧
⎪ X0 = r cos φ sin θ1 sin θ2 . . . sin θn−1
⎪
⎪
⎪
⎨ X1 = r sin φ sin θ1 sin θ2 . . . sin θn−1
0 < θi < π
X2 = r cos θ1 sin θ2 . . . sin θn−1 . (1.15)
⎪
⎪ 0 ≤ φ < 2π
⎪
⎪ · · · ···
⎩
Xn = r cos θn−1
The volume element dA on the unit sphere then becomes
dA = dφdθ1 . . . dθn−1 sin θ1 sin2 θ2 . . . sinn−1 θn−1 . (1.16)
We want to compute the volume vol(Sn ) of the n-sphere, that is to say its
‘hyperarea’ – meaning that vol(S2 ) is measured in square metres, vol(S3 ) in cubic
metres, and so on. A clever trick simplifies the calculation: Consider the well known
Gaussian integral

√
I = e−X0 −X1 − ... −Xn dX0 dX1 . . . dXn = ( π)n+1 .
2 2 2
(1.17)
Using the spherical polar coordinates introduced above our integral splits into two,
one of which
∞ is related to the integral representation of the Euler Gamma function,
(x) = 0 e−t tx−1 dt, and the other is the one we want to do:
∞
−r2 n 1 n+1
I= dr dAe r = vol(Sn ). (1.18)
0 Sn 2 2
We do not have to do the integral over the angles. We simply compare these results
and obtain (recalling the properties of the Gamma function)
⎧ p
⎪ 2(2π)
⎪
⎪ if n = 2p
π
n+1
2 ⎨ (2p − 1)! !
vol(S ) = 2 n+1 =
n
, (1.19)
( 2 ) ⎪
⎪
⎪ (2π)p+1
⎩ if n = 2p + 1
(2p)! !
where double factorial is the product of every other number, 5! ! = 5 · 3 · 1 and
6! ! = 6 · 4 · 2. An alarming thing happens as the dimension grows. For large x we
can approximate the Gamma function using Stirling’s formula

√ −x x− 21 1 1
(x) = 2π e x 1+ +o 2 . (1.20)
12x x
Hence for large n we obtain
n
√ 2πe 2
vol(S ) ∼ 2
n
. (1.21)
n
This is small if n is large! In fact the ‘biggest’ unit sphere – in the sense that it has
the largest hyperarea – is S6 , which has
16 3
vol(S6 ) = π ≈ 33.1. (1.22)
15
Incidentally Stirling’s formula gives 31.6, which is already rather good. We hasten
to add that vol(S2 ) is measured in square metres and vol(S6 ) in (metre)6 , so that the
direct comparison makes no sense.
There is another funny thing to be noticed. If we compute the volume of the
n-sphere without any clever tricks, simply by integrating the volume element dA
using angular coordinates, then we find that
π π π
vol(S ) = 2π
n
dθ sin θ 2
dθ sin θ . . . dθ sinn−1 θ = (1.23)
0
π 0 0
= vol(Sn−1 ) dθ sinn−1 θ.
0
As n grows the integrand of the final integral has an increasingly sharp peak close to
the equator θ = π/2. Hence we conclude that when n is high most of the hyperarea
of the sphere is confined to a ‘band’ close to the equator. What about the volume
of an n-dimensional unit ball Bn ? By definition it has unit radius and its boundary
1
is Sn−1 . Its volume, using the radial integral 0 rn−1 dr = 1/n and the fact that
(x + 1) = x(x), is
n n
vol(Sn−1 ) π2 1 2πe 2
vol(B ) =
n
= ∼√ . (1.24)
n ( 2 + 1)
n
2π n
Again, as the dimension grows the denominator grows faster than the numerator and
therefore the volume of a unit ball is small when the dimension is high. We can turn
this around if we like: a ball of unit volume has a large radius if the dimension is
high. Indeed since the volume is proportional to rn , where r is the radius, it follows
√
that the radius of a ball of unit volume grows like n when Stirling’s formula
applies.
The fraction of the volume of a unit ball that lies inside a radius r is rn . We
assume r < 1, so this is a quickly shrinking fraction as n grows. The curious
conclusion of this is that when the dimension is high almost all of the volume
of a ball lies very close to its surface. In fact this is a crucial observation in sta-
tistical mechanics. It is also the key property of n-dimensional geometry: when
n is large the ‘amount of space’ available grows very fast with distance from the
origin.
In some ways it is easier to see what is going on if we consider hypercubes n
rather than balls. Take a cube of unit volume. In n dimensions it has 2n corners,
and the longest straight line that we can√ draw inside the hypercube connects two
√
opposite corners. It has length L = 1 + . . . + 1 = n. Or expressed in
2 2
another way, a straight line of any length fits into a hypercube of unit volume
if the dimension is large enough. The reason why the longest line segment fit-
ting into the cube is large is clearly that we normalised the volume to one. If
√
we normalise L = 1 instead we find that the volume goes to zero like (1/ n)n .
Concerning the insphere – the largest inscribed sphere, with inradius rn – and the
outsphere – the smallest circumscribed sphere, with outradius Rn – we observe
that
√
n √
Rn = = nrn . (1.25)
2
√
The ratio between the two grows with the dimension, ζn ≡ Rn /rn = n. Inciden-
tally, the somewhat odd statement that the volume of a sphere goes to zero when
the dimension n goes to infinity can now be interpreted: since vol(n ) = 1 the real
statement is that vol(Sn )/vol(n ) goes to zero when n goes to infinity.
Now we turn to simplices, whose properties will be of some importance later
on. We concentrate on regular simplices n , for which the distance between any
pair of corners is one. For n = 1 this is the unit interval, for n = 2 a regular
triangle, for n = 3 a regular tetrahedron, and so on. Again we are interested
in the volume, the radius rn of the insphere, and the radius Rn of the outsphere.
We will also compute χn , the angle between the lines from the ‘centre of mass’
to a pair of corners. For a triangle it is arccos(−1/2) = 2π/3 = 120 degrees,
but it drops to arccos(−1/3) ≈ 110◦ for the tetrahedron. A practical way to go
about all this is to think of n as a pyramid (a part of a cone, namely the convex
Another random document with
no related content on Scribd:
Prâna cannot work alone without the help of âkâsa. All that we know
is that motion or vibration and every movement that we see is a
modification of this prâna, and everything that we know in the form of
matter, either as form or as resistance, is a modification of this
âkâsa. This prâna cannot exist alone, or act without a medium, but in
every state of it, whether as pure prâna, or when it changes into
other forces of nature, say gravitation or centrifugal attraction, it can
never be separate from âkâsa. You have never seen force without
matter or matter without force; what we call force and matter being
simply the gross manifestations of these same things, which, when
superfine, we call prâna and âkâsa. Prâna you can call in English the
life, or vital energy, but you must not restrict it to the life of man, nor
should you identify it with the spirit, Âtman. Creation is without
beginning and without end; it cannot have either, it is an eternal on-
going.
The next question that comes is rather a fine one. Some European
philosophers have asserted that this world exists because “I” exist,
and if “I” do not exist, the world will not exist. Sometimes it is
expressed in this way; they say, if all the people in the world were to
die, and there were no more human beings, and no animals with
powers of perception and intelligence, all manifestations would
disappear. It seems paradoxical, but gradually we shall see clearly
that this can be proved. But these European philosophers do not
know the psychology of it, although they know the principle; modern
philosophy has got only a glimpse of it.
First we will take another proposition of these old psychologists
which is rather startling, that the grossest elements are the bhutas,
but that all gross things are the results of fine ones. Everything that
is gross is composed of a combination of minute things, so the
bhutas must be composed of certain fine particles, called in Sanskrit
the tanmâtras. I smell a flower; to smell that, something must come
in contact with my nose; the flower is there and I do not see it move
towards me; but without something coming in contact with my nose I
cannot smell the flower. That which comes from the flower and into
contact with my nose are the tanmâtras, fine molecules of that
flower, so fine that no diminution can be perceived in the flower. So
with heat, light, sight, and everything. These tanmâtras can again be
subdivided into atoms. Different philosophers have different theories,
and we know these are only theories, so we leave them out of
discussion. Sufficient for us that everything gross is composed of
things that are very, very minute. We first get the gross elements,
which we feel externally, and composing them are the fine elements,
which our organs touch, which come in contact with the nerves of the
nose, eyes and ears. That ethereal wave which touches my eyes, I
cannot see, yet I know it must come in contact with my optic nerve
before I can see the light. So with hearing, we can never see the
particles that come in contact with our ears, but we know that they
must be there. What is the cause of these tanmâtras? A very
startling and curious answer is given by our psychologists,—self-
consciousness. That is the cause of these fine materials, and the
cause of the organs. What are these organs? Here is first the eye,
but the eye does not see. If the eyes did see, when a man is dead,
and his eyes are still perfect, they would still be able to see. There is
some change somewhere; something has gone out of the man, and
that something, which really sees, of which the eye is the instrument,
is called the organ. So this nose is an instrument, and there is an
organ corresponding to it. Modern physiology can tell you what that
is, a nerve centre in the brain. The eyes, ears, etc., are simply the
external instruments. It may be said that the organs, Indriyas, as
they are called in Sanskrit, are the real seats of perception.
What is the use of having one organ for the nose, and one for the
eyes, and so on? Why will not one serve the purpose? To make it
clear to you,—I am talking, and you are listening, and you do not see
what is going on around you because the mind has attached itself to
the organ of hearing, and has detached itself from the sight organ. If
there were only one organ the mind would hear and see at the same
time, it would see and hear and smell at the same time, and it would
be impossible for it not to do all three at the same time. Therefore it
is necessary that there should be separate organs for all these
centres. This has been borne out by modern physiology. It is
certainly possible for us to see and hear at the same time, but that is
because the mind attaches itself partially to both centres, which are
the organs. What are the instruments? We see that these are really
made of the gross materials. Here they are,—eyes, nose, and ears,
etc. What are the organs? They are also made of materials, because
they are centres. Just as this body is composed of gross material for
transforming prâna into different gross forces, so these finer organs
behind, are composed of the fine elements, for the manufacture of
prâna into the finer forces of perception and all kindred things. All
these organs or indriyas combined, plus the internal instrument or
antahkarana, are called the finer body of man,—the linga (or
sûkshma) sarira.
It has a real form, because everything material must have a form.
Behind the indriyas is what is called the manas, the chitta in vritti,
what might be called the vibratory state of the mind, the unsettled
state. If you throw a stone into a calm lake, first there will be
vibration, and then resistance. For a moment the water will vibrate
and then it will react on the stone. So, when any impression comes
on the chitta, or “mind stuff,” it vibrates a little. This state of the mind
is called the manas. Then comes the reaction, the will. There is
another thing behind this will which accompanies all the acts of the
mind, which is called egoism, the ahamkâra, the self-consciousness,
which says “I am,” and behind that is what is called Buddhi, the
intellect, the highest form of nature’s existence. Behind the intellect
is the true Self of man, the Purusha, the pure, the perfect, who is
alone the seer, and for whom is all this change. The Purusha is
looking on at all these changes; he himself is never impure; but by
implication, what the Vedantists call adhyâsam, by reflection, he
appears to be impure. It is like a red flower held before a piece of
crystal; the crystal will look red; or a blue flower and the crystal will
look blue; and yet the crystal itself is colorless. We will take for
granted that there are many selves; each self is pure and perfect, but
it is all these various divisions of gross matter and fine matter that
are imposing on the self, and making it variously colored. Why is
nature doing all this? Nature is undergoing all these changes for the
improvement of the soul; all this creation is for the benefit of the soul,
so that it may be free. This immense book which we call the universe
is stretched before man so that he may read, and come out, as an
omniscient and omnipotent being. I must here tell you that some of
our best psychologists do not believe in a personal God in the sense
in which you believe in Him. The real father of our psychologists,
Kapila, denies the existence of God as a Creator. His idea is that a
personal God is quite unnecessary; Nature itself is sufficient to work
out all that is good. What is called the “Design” theory he repudiated,
and said a more childish theory was never advanced. But he admits
a peculiar kind of God; he says we are all struggling to get free, and
when man becomes free he can, as it were, melt away into Nature
for the time being, only to come out at the beginning of the next cycle
and be its ruler; come out an omniscient and omnipotent being. In
that sense he can be called God; you and I and the humblest beings
will be gods in different cycles. Kapila says such a God will be
temporal, but an eternal God, eternally omnipotent and eternally
ruler of the universe, cannot be. If there were such a God, there
would be this difficulty: he must either be bound or free. A God who
is perfectly free would not create; there would be no necessity. If he
were bound, he would not create because he could not, he would be
weak himself. So, in either case, there cannot be an omnipotent or
omniscient eternal ruler. So wherever the word God is mentioned in
our Scriptures, Kapila says it means those perfected souls who have
become free. The Sânkhya system does not believe in the unity of all
souls. Vedânta believes that all individual souls are united in one
cosmic Being called Brahman, but Kapila, the founder of the
Sânkhya, was dualistic. His analysis of the universe so far as it goes
is really marvellous. He was the father of Hindu evolutionists, and all
the later philosophical systems are simply outcomes of his thought.
According to this system all souls will regain their freedom and their
natural rights, which are omnipotence and omniscience. Here the
question may be asked, whence is this bondage of the souls? The
Sânkhya says it is without beginning, but if it be without beginning it
must also be without end, and we shall never be free. Kapila
explains that this “without beginning” means not in a constant line.
Nature is without beginning and without end, but not in the same
sense as is the soul, because Nature has no individuality, just as a
river flowing by us is every moment getting a fresh body of water,
and the sum total of all these bodies of water is the river, so the river
is not a constant quantity. Similarly everything in Nature is constantly
changing, but the soul never changes. Therefore as Nature is always
changing, it is possible for the soul to come out of its bondage. One
theory of the Sânkhya is peculiar to this psychology. The whole of
the universe is built upon the same plan as one single man, or one
little being; so, just as I have a mind, there is also a cosmic mind.
When this macrocosm evolves there must be first intelligence, then
egoism, then the tanmâtras and the organs, and then the gross
elements. The whole universe according to Kapila is one body, all
that we see are the grosser bodies, and behind these are the finer
bodies, and behind them, a universal egoism, and behind that a
universal Intelligence, but all this is in Nature, all this is manifestation
of Nature, not outside of Nature. Each one of us is a part of that
cosmic consciousness. There is a sum-total of intelligence out of
which we draw what we require, so there is a sum-total of mental
force in the universe out of which we are drawing eternally, but the
seed for the body must come from the parents. The theory includes
heredity and reincarnation too. The material is given to the soul out
of which to manufacture a body, but that material is given by
hereditary transmission from the parents.
We come now to that proposition that in this process there is an
involution and an evolution. All is evolved out of that indiscreet
Nature; and then is involved again and becomes Avyaktam. It is
impossible, according to the Sânkhyas, for any material thing to
exist, which has not as its material some portion of consciousness.
Consciousness is the material out of which all manifestation is made.
The elucidation of this comes in our next lecture, but I will show how
it can be proved. I do not know this table as it is, but it makes an
impression; it comes to the eyes, then to the indriyas, and then to the
mind; the mind then reacts, and that reaction is what I call the table.
It is just the same as throwing a stone into a lake; the lake throws a
wave against the stone; this wave is what we know. The waves
coming out are all we know. In the same way the fashion of this wall
is in my mind; what is external nobody knows; when I want to know
it, it has to become that material which I furnish; I, with my own mind,
have furnished the material for my eyes, and the something which is
outside is only the occasion, the suggestion, and upon that
suggestion I project my mind, and it takes the form of what I see.
The question is, how do we all see the same things? Because we all
have a part of this cosmic mind. Those who have mind will see the
thing, and those who have not will not see it. This goes to show that
since this universe has existed there has never been a want of mind,
of that one cosmic mind. Every human being, every animal, is also
furnished out of that cosmic mind, because it is always present and
furnishing material for their formation.
II
PRAKRITI AND PURUSHA
We will take up the categories we have been discussing and come to
the particulars. If we remember we started with Prakriti, or Nature.
This Nature is called by the Sânkhya philosophers indiscrete or
inseparate, which is defined as perfect balance of the materials in it;
and it naturally follows that in perfect balance there cannot be any
motion. All that we see, feel, and hear is simply a compound of
motion and matter. In the primal state, before this manifestation,
where there was no motion, perfect balance, this Prakriti was
indestructible, because decomposition comes only with limitation.
Again, according to the Sânkhya, atoms are not the primal state.
This universe does not come out of atoms, they may be the
secondary, or tertiary state. The original matter may compound into
atoms, which in turn compound into greater and greater things, and
as far as modern investigations go, they rather point towards that.
For instance, in the modern theory of ether, if you say ether is also
atomic, that will not solve the proposition at all. To make it clearer,
say that air is composed of atoms, and we know that ether is
everywhere, interpenetrating, omnipresent, and these atoms are
floating, as it were, in ether. If ether again be composed of atoms,
there will still be some space between two atoms of ether. What fills
up that? And again there will be another space between the atoms of
that which fills up this space. If you propose that there is another
ether still finer you must still have something to fill that space, and so
it will be regressus in infinitum, what the Sânkhya philosophers call
anavasthâ,—never reaching a final conclusion. So the atomic theory
cannot be final. According to the Sânkhyas this Nature is
omnipresent, one omnipresent mass of Nature in which are the
causes of everything that exists. What is meant by cause? Cause is
the more subtle state of the manifested state, the unmanifested state
of that which becomes manifested. What do you mean by
destruction? It is reverting to the cause,—the materials out of which
a body is composed go back into their original state. Beyond this
idea of destruction, any idea such as annihilation, is on the face of it
absurd. According to modern physical sciences, it can be
demonstrated that all destruction means that which Kapila called
ages ago “reverting to the causal state.” Going back to the finer form
is all that is meant by destruction. You know how it can be
demonstrated in a laboratory that matter is indestructible. Those of
you who have studied chemistry will know that if you burn a candle
and put a caustic pencil inside a glass tube beneath the candle,
when the candle has burned away, if you take the caustic pencil out
of the tube and weigh it, you will find that the pencil will weigh exactly
its previous weight, plus the weight of the candle,—the candle
became finer and finer, and went on to the caustic. So that in this
present stage of our knowledge, if any man claims that anything
becomes annihilated, he is only making himself absurd. It is only
uneducated people who would advance such a proposition, and it is
curious that modern knowledge coincides with what those old
philosophers taught. The ancients proceeded in their inquiry by
taking up mind as the basis; they analyzed the mental part of this
universe and came to certain conclusions, while modern science is
analyzing the physical part, and it also comes to the same
conclusions. Both analyses must lead to the same truth.
You must remember that the first manifestation of this Prakriti in the
cosmos is what the Sânkhyas called Mahat. We may call it universal
intelligence, the great principle; that is the literal meaning. The first
manifestation of Prakriti is this intelligence; I would not translate it by
self-consciousness, because that would be wrong. Consciousness is
only a part of this intelligence, which is universal. It covers all the
grounds of consciousness, sub-consciousness and super-
consciousness. In Nature, for instance, certain changes are going on
before your eyes which you see and understand, but there are other
changes so much finer that no human perception can catch them.
They are from the same cause, the same Mahat is making these
changes. There are other changes, beyond the reach of our mind or
reasoning, all this series of changes is in this Mahat. You will
understand it better when I come to the individual. Out of this Mahat
comes the universal egoism, and these are both material. There is
no difference between matter and mind save in degree. It is the
same substance in finer or grosser form; one changes into the other,
and this will exactly coincide with the modern physiological research,
and it will save you from a great deal of fighting and struggling to
believe that you have a mind separate from the brain, and all such
impossible things. This substance called Mahat changes into the
material egoism, the fine state of matter, and that egoism changes
into two varieties. In one variety it changes into the organs. Organs
are of two kinds—organs of sensation and organs of reaction. They
are not the eyes or nose, but something finer, what you call brain
centres, and nerve centres. This egoism becomes changed, and out
of this material are manufactured these centres and these nerves.
Out of the same substance, the egoism, is manufactured a yet finer
form, the tanmâtras, fine particles of matter, those for instance which
strike your nose and cause you to smell. You cannot perceive these
fine particles, you can only know that they are there. These
tanmâtras are manufactured out of that egoism, and out of these
tanmâtras, or subtle matter, is manufactured the gross matter, air,
water, earth, and all the things that we see and feel. I want to
impress this on your mind. It is very hard to grasp it, because, in
Western countries, the ideas are so queer about mind and matter. It
is hard to take these impressions out of our brains. I myself had a
tremendous difficulty, being educated in Western philosophy in my
childhood. These are all cosmic things. Think of this universal
extension of matter, unbroken, one substance, undifferentiated,
which is the first state of everything, and which begins to change just
as milk becomes curd, and it is changed into another substance
called Mahat, which in one state manifests as intelligence and in
another state as egoism. It is the same substance, and it changes
into the grosser matter called egoism; thus is the whole universe
itself built, as it were, layer after layer; first undifferentiated Nature
(Avyaktam), and that changes into universal intelligence (Mahat),
and that again is changed into universal egoism (Ahamkâra), and
that changes into universal sensible matter. That matter changes into
universal sense-organs, again changes into universal fine particles,
and these in turn combine and become this gross universe. This is
the cosmic plan, according to the Sânkhyas, and what is in the
cosmos or macrocosm, must be in the individual or microcosm.
Take an individual man. He has first a part of undifferentiated nature
in him, and that material nature in him becomes changed into mahat,
a small particle of the universal intelligence, and that small particle of
the universal intelligence in him becomes changed into egoism—a
particle of the universal egoism. This egoism in turn becomes
changed into the sense-organs, and out of these sense-organs come
the tanmâtras, and out of them he combines and manufactures his
world, as a body. I want this to be clear, because it is the first
stepping stone to Vedânta, and it is absolutely necessary for you to
know, because this is the philosophy of the whole world. There is no
philosophy in the world that is not indebted to Kapila, the founder of
this Sânkhya system. Pythagoras came to India and studied his
philosophy and carried some of these ideas to the Greeks. Later it
formed the Alexandrian school, and still later formed the basis of
Gnostic philosophy. It became divided into two parts; one went to
Europe and Alexandria, and the other remained in India, and
became the basis of all Hindu philosophy, for out of it the system of
Vyâsa was developed. This was the first rational system that the
world saw, this system of Kapila. Every metaphysician in the world
must pay homage to him. I want to impress on your mind that as the
great father of philosophy, we are bound to listen to him, and respect
what he said. This wonderful man, most ancient of philosophers, is
mentioned even in the Vedas. How wonderful his perceptions were!
If there is any proof required of the power of the Yogis to perceive
things beyond the range of the ordinary senses, such men are the
proofs. How could they perceive them? They had no microscopes, or
telescopes. How fine their perception was, how perfect their analysis
and how wonderful!
To revert again to the microcosm, man. As we have seen, he is built
on exactly the same plan. First the nature is “indiscrete” or perfectly
balanced, then it becomes disturbed, and action sets in and the first
change produced by that action is what is called mahat,—
intelligence. Now you see this intelligence in man is just a particle of
the cosmic intelligence,—the Mahat. Out of it comes self-
consciousness, and from this the sensory and the motor nerves, and
the finer particles out of which the gross body is manufactured. I will
here remark that there is one difference between Schopenhauer and
Vedânta. Schopenhauer says the desire, or will, is the cause of
everything. It is the will to exist that makes us manifest, but the
Advaitists deny this. They say it is the intelligence. There cannot be
a single particle of will which is not a reaction. So many things are
beyond will. It is only a manufactured something out of the ego, and
the ego is a product of something still higher, the intelligence, and
that is a modification of “indiscrete” Nature, or Prakriti.
It is very important to understand this mahat in man,—the
intelligence. This intelligence itself is modified into what we call
egoism, and this intelligence is the cause of all these motions in the
body. This covers all the grounds of sub-consciousness,
consciousness and super-consciousness. What are these three
states? The sub-conscious state we find in animals, what we call
instinct. This is nearly infallible, but very limited. Instinct almost never
fails. An animal instinctively knows a poisonous herb from an edible
one, but its instinct is limited to one or two things; it works like a
machine. Then comes the higher state of knowledge, which is
fallible, makes mistakes often, but has a larger scope, although it is
slow, and this you call reason. It is much larger than instinct, but
there are more dangers of mistake in reasoning than in instinct.
There is a still higher state of the mind, the super-conscious, which
belongs only to the Yogis, men who have cultivated it. This is as
infallible as instinct, and still more unlimited than reason. It is the
highest state. We must remember that in man this mahat is the real
cause of all that is here, that which is manifesting itself in various
ways, covers the whole ground of sub-conscious, conscious and
super-conscious states, the three states in which knowledge exists.
So in the Cosmos, this universal Intelligence, Mahat, exists as
instinct, as reason, and as super-reason.
Now comes a delicate question, which is always being asked. If a
perfect God created the universe, why is there imperfection in it?
What we call the universe is what we see, and that is only this little
plane of consciousness or reason, and beyond that we do not see at
all. Now the very question is an impossible one. If I take up only a bit
out of a mass and look at it, it seems to be imperfect. Naturally. The
universe seems imperfect because we make it so. How? What is
reason? What is knowledge? Knowledge is finding associations. You
go into the street and see a man, and know it is a man. You have
seen many men, and each one has made an impression on your
mind, and when you now see this man, you calmly refer to your store
of impressions, see many pictures of men there, and you put this
new one with the rest, pigeon-hole it and are satisfied. When a new
impression comes and it has associations in your mind, you are
satisfied, and this state of association is called knowledge.
Knowledge is, therefore, pigeon-holing one experience with the
already existing fund of experience, and this is one of the great
proofs that you cannot have any knowledge until you have already a
fund in existence. If you are without experience, or if, as some
European philosophers think, the mind is a tabula rasa, it cannot get
any knowledge, because the very fact of knowledge is the
recognition of the new by comparison with already existing
impressions. There must be a store ready to which to refer a new
impression. Suppose a child is born into this world without such a
fund, it would be impossible for him to get any knowledge. Therefore
the child must have been in a state in which he had a fund, and so
knowledge is eternally going on. Show me any way of getting out of
this. It is mathematical experience. This is very much like the
Spencerian and other philosophies. They have seen so far that there
cannot be any knowledge without a fund of past knowledge. They
have drawn out the idea that the child is born with knowledge. They
say that the cause has entered the effect. It comes in a subtle form
in order to be developed. These philosophers say that these
impressions with which the child comes, are not from the child’s own
past, but were in his forefathers’; that it is hereditary transmission.
Very soon they are going to find this theory untenable, and some of
them are now giving hard blows to these ideas of heredity. Heredity
is very good, but incomplete. It only explains the physical side. How
do you explain the influence of environment? Many causes produce
one effect. Environment is one of the modifying causes. On the other
hand we in turn make our own environment, because as our past
was, so we find our present. In other words, we are what we are
here and now, because of what we have been in the past.
You understand what is meant by knowledge. Knowledge is pigeon-
holing a new impression with old impressions—recognizing a new
impression. What is meant by recognition? Finding its association
with the similar impressions that we already have. Nothing further is
meant by knowledge. If that be the case, it must be that we have to
see the whole series of similars. Is it not? Suppose you take a
pebble; to find the association, you have to see the whole series of
pebbles similar to it. But with the universe we cannot do that,
because in our reasoning we can only go after one perception of our
universe, and neither see on this side nor on that side, and we
cannot refer it to its association. Therefore the universe seems
unintelligible, because knowledge and reason are always finding
associations. This bit of the universe cut off by our consciousness is
a startling new thing, and we have not been able to find its
associations. Therefore we are struggling with it, and thinking it is so
horrible, so wicked, and bad;—sometimes we think it is good, but
generally we think it is imperfect. The universe will be known only
when we find the associations. We shall recognize them when we go
beyond the universe and consciousness, and then the universe will
stand explained. Until we do that all our fruitless striving will never
explain the universe, because knowledge is the finding of similars,
and this conscious plane gives us only a partial view. So with our
idea of the universal Mahat, or what in our ordinary everyday
language we call God. All that we have of God is only one
perception, just as of the universe we see only one portion, and all
the rest is cut off and covered by our human limitation. “I, the
Universal, so great am I that even this universe is a part of me.” That
is why we see God as imperfect, and we can never understand Him,
because it is impossible. The only way to understand is to go beyond
reason, beyond consciousness. “When thou goest beyond the heard
and hearing, the thought and thinking, then alone wilt thou come to
truth.” (Bhagavad Gita II. 52.) “Go thou beyond the Scriptures,
because they teach only up to Nature, up to the three qualities.”
(Gita II. 45.) When we go beyond them we find the harmony, not
before.
So far it is clear that this macrocosm and microcosm are built on
exactly the same plan, and in this microcosm we know only one very
small part. We know neither the sub-conscious, nor the super-
conscious. We know only the conscious. If a man says “I am a
sinner,” he is foolish, because he does not know himself. He is the
most ignorant of men about himself; one part only he knows,
because the fact of knowledge covers only one part of the “mind-
ground” he is in. So with this universe; it is possible to know only one
part through reasoning, but Nature comprises the whole of it, the
sub-conscious, the conscious and the super-conscious, the
individual mahat and the universal Mahat with all their subsequent
modifications, and these lie beyond reason.
What makes nature change? We see that Up to this point everything,
all Prakriti, is jadâ (insentient). It is working under law; it is all
compound and insentient. Mind, intelligence, and will, all are
insentient. But they are all reflecting the sentiency, the Chit
(Intelligence) of some Being who is beyond all this, and whom the
Sânkhya philosophers call Purusha. This Purusha is the unwitting
cause of all these changes in Nature—in the universe. That is to say,
this Purusha, taking Him in the universal sense, is the God of the
universe. It is claimed that the will of the Lord created the universe.
This is very good as a common daily expression, but that is all. How
could it be will? Will is the third or fourth manifestation in Nature.
Many things exist before it, and what created them? Will is a
compound, and everything that is a compound is a production out of
Nature. Will itself cannot create Nature. It is not a simple. So to say
that the will of the Lord created the universe is illogical. Our will only
covers a little portion of self-consciousness, and moves our brain,
they say. If it did you could stop the action of the brain, but you
cannot. It is not the will. Who moves the heart? It is not the will,
because if it were you could stop it or not at your will. It is neither will
that is working your body, nor that is working the universe. But it is
something of which will itself is one of the manifestations. This body
is being moved by the power of which will is only a manifestation in
one part. So in the universe there is will, but that is only one part of
the universe. The whole of the universe is not guided by will, that is
why we do not find the explanation in will. Suppose I take it for
granted that the will is moving the body, and then I begin to fret and
fume. It is my fault, because I had no right to take it for granted that it
was will. In the same way, if I take the universe and think it is will that
moves it and then find things that do not coincide, it is my fault. This
Purusha is not will, neither can it be intelligence, because
intelligence itself is a compound. There cannot be any intelligence
without some sort of matter. In man, this matter takes the form which
we call brain. Wherever there is intelligence there must be matter in
some form or other. But that intelligence itself is a compound. What
then is this Purusha? It is neither intelligence nor buddhi (will), but
yet it is the cause of both these; it is His presence that sets them all
vibrating and combining. Purusha may be likened to some of these
substances which by their mere presence promote chemical
reaction, as in the case of cyanide of potassium which is added
when gold is being smelted. The cyanide of potassium remains
separate and unaffected, but its presence is absolutely necessary to
the success of the process. So with the Purusha. It does not mix with
Nature: it is not Intelligence, or Mahat, or any one of these, but the
Self, the Pure, the Perfect. “I am the Witness, and through My
witnessing, Nature is producing all that is sentient and all that is
insentient.” (Gita IX. 10.)
What is this sentiency in Nature? The basis of sentiency is in the
Purusha, is the nature of the Purusha. It is that which cannot be
spoken, but which is the material of all that we call knowledge. This
Purusha is not consciousness, because consciousness is a
compound, but whatever is light and goodness in this consciousness
belongs to It. Sentiency is in the Purusha, but the Purusha is not
intelligent, not knowing, it is knowledge itself. The Chit in the
Purusha, plus Prakriti, is what is known to us as intelligence and
consciousness. Whatever is pleasure and happiness and light in the
universe belongs to the Purusha, but it is a compound because it is
that Purusha plus Nature. “Wherever there is any happiness,
wherever there is any bliss, there is one spark of that immortality,
which is Purusha.” This Purusha is the great attraction of the
universe, untouched by, and unconnected with the universe, yet it
attracts the whole universe. You see a man going after gold,
because therein is a spark of the Purusha, even though he knows it
not. When a man desires children, or a woman a husband, what is
the attracting power? That spark of Purusha behind the child or wife,
behind everything. It is there, only overlaid with matter. Nothing else
can attract. “In this world of insentiency that Purusha alone is
sentient.” This is the Purusha of the Sânkhyas. As such it
necessarily follows that this Purusha must be omnipresent. That
which is not omnipresent must be limited. All limitations are caused;
that which is caused must have beginning and end. If the Purusha is
limited it will die, will not be final, will not be free, but will have been
caused. Therefore if not limited, it is omnipresent. According to
Kapila there are many Purushas, not one. An infinite number of
them, you are one, I am one, each is one; an infinite number of
circles, each one infinite, running through this universe. The Purusha
is neither born nor dies. It is neither mind nor matter, and the reflex
from it is all that we know. We are sure if it be omnipresent it knows
neither death nor birth. Nature is casting her shadow upon it, the
shadow of birth and death, but it is by its own nature eternal. So far
we have found the theory of Kapila wonderful.
Next we will have to take up the proofs against it. So far the analysis
is perfect, the psychology cannot be controverted. There is no
objection to it. We will ask of Kapila the question: Who created
Nature? and his answer will be that Nature (Prakriti) is uncreate. He
also says that the Purusha is omnipresent and that of these
Purushas there is an infinite number. We shall have to controvert this
last proposition, and find a better solution, and by so doing we shall
come to the ground taken by Vedânta. Our first doubt will be how
there can be these two infinites. Then our argument will be that it is
not a perfect generalization, and that therefore we have not found a
perfect solution. And then we shall see how the Vedantists find their
way out of all these difficulties and reach a perfect solution. Yet all
the glory really belongs to Kapila. It is very easy to give a finish to a
building that is nearly complete.
III
SÂNKHYA AND ADVAITA
I will give you first a resumé of the Sânkhya philosophy, through
which we have been going, because in this lecture we want to find
where its defects are, and where Vedânta comes in as
supplementary to these defects. You must remember that according
to the Sânkhya philosophy, Nature is causing all these
manifestations which we call thought and intellect, reason, love,
hatred, touch, taste; that everything is from Nature. This Nature
consists of three sorts of elements, one called Sattva, another Rajas,
and the third Tamas. These are not qualities, but the materials out of
which the whole universe is being evolved, and at the beginning of a
cycle they remain in equilibrium. When creation comes this
equilibrium is disturbed and these elements begin to combine and
recombine, and manifest as the universe. The first manifestation of
this is what the Sânkhya calls the Mahat (universal Intelligence), and
out of that comes consciousness. And out of consciousness is
evolved Manas (universal Mind). Out of this consciousness are also
evolved the organs of the senses, and the tanmâtras,—sound
particles, touch particles, taste particles, and so forth. All fine
particles are evolved from this consciousness, and out of these fine
particles come the gross particles which we call matter. After the
tanmâtras (those particles which cannot be seen, or measured)
come the gross particles which we can feel and sense. The chitta
(“mind-stuff”) in its three-fold functions of intellect, consciousness
and mind, is working and manufacturing the forces called prânas.
These prânas have nothing to do with breath, you must at once get
rid of that idea. Breath is one effect of the Prâna (universal Energy).
By these prânas are meant the nervous forces that are governing
and moving the whole body, which are manifesting themselves as
thought, and as the various functions of the body. The foremost and
the most obvious manifestation of these prânas is the breathing
motion. If it were caused by air, a dead man would breathe. The
prâna acts upon the air, and not air upon it. These prânas are the
vital forces which manipulate the whole body, and they in turn are
manipulated by the mind and the indriyas (the two kinds of organs).
So far so good. The psychology is very clear and most precise, and
just think of the age of it, the oldest rational thought in the world!
Wherever there is any philosophy or rational thought, it owes
something to Kapila. Wherever there is any attempt at psychology, or
philosophy, there is some indebtedness to the great father of this
thought, to this man Kapila.
So far we see that this psychology is wonderful, but we shall have to
differ with it on some points, as we go on. We find that the principal
idea on which Kapila works is evolution. He makes one thing evolve
out of another, because his very definition of causation is “the effect
is the cause reproduced in another form,” and because the whole
universe, so far as we see it, is progressive and evolving. This whole
universe must have evolved out of some material, out of Prakriti or
Nature. Therefore this Nature cannot be essentially different from its
cause, only when it takes form it becomes limited. The material itself
is without form. But according to Kapila, from undifferentiated nature
down to the last stage of differentiation, none of these is the same as
Purusha, the “Enjoyer,” or “Enlightener.” Just as a lump of clay, so is
a mass of mind, or the whole universe. By itself it has no light, but
we find reason and intelligence in it, therefore there must be some
Existence behind it, behind the whole of Nature, whose light is
percolating through it and appearing as Mahat and consciousness
and all these various things, and this Existence is what Kapila calls
the Purusha, the Âtman or Self of the Vedantist. According to Kapila,
the Purusha is a simple factor, not a compound. It is immaterial, the
only one that is immaterial, whereas all the various manifestations
are material. The Purusha alone knows. Suppose I see a
blackboard, first the external instruments will bring that sensation to
the organ (to the indriya according to Kapila), from the organ it will
go to the mind and make an impression; the mind will cover it up with
another factor,—consciousness, and will present it to the buddhi
(intelligence), but buddhi cannot act; it is the Purusha behind that
acts. These are all its servants, bringing the sensation to It, and It
gives the orders, and the buddhi reacts. The Purusha is the Enjoyer,
the Perceiver, the real One, the King on his throne, the Self of man,
and It is immaterial. Because It is immaterial, it necessarily follows
that It must be infinite, It cannot have any limitation whatever. So
each one of these purushas is omnipresent, each is all-pervading,
but can act only through fine and gross manifestations of matter. The
mind, the self-consciousness, the organs and the vital forces
compose what is called the fine body, or what in Christian philosophy
is called the “spiritual body” of man. It is this body that comes to
reward or punishment, that goes to the different heavens; that
incarnates and reincarnates; because we see from the very
beginning that the going and coming of the soul (Purusha) is
impossible. Motion means going and coming, and that which goes
from one place to another cannot be omnipresent. It is this linga-
sarira (subtle body) which comes and goes. Thus far we see from
Kapila’s psychology that the soul is infinite, and that the soul is the
only principle that is not an evolution of Nature. It is the only one that
is outside of Nature, but It has apparently got bound by Nature. This
Nature is around the Purusha and It has identified Itself with Nature.
It thinks “I am the linga-sarira,” It thinks “I am the gross matter, the
gross body,” and as such is enjoying pleasure and pain; but these do
not really belong to the soul, they belong to this linga-sarira, and to
the gross body. When certain nerves are hurt we feel pain. We
recognize that immediately. If the nerves in our fingers were dead we
could cut the fingers and not feel it. So pleasure and pain belong to
the nerve-centres. Suppose my organ of sight is destroyed, I do not
feel pleasure or pain from color, although my eyes are there. So it is
obvious that pleasure and pain do not belong to the soul. They
belong to the mind and the body.
The soul has neither pleasure nor pain; it is the Witness of
everything, the eternal Witness of things that are going on, but it
takes no fruits from any work. “As the sun is the cause of sight in
every eye, yet is not itself affected by the defects in any eye; as a
piece of crystal appears red when red flowers are placed before it,
so this Purusha appears to be affected by pleasure or pain from the
reflection cast upon It by Nature, but it remains ever unchanged.”
The nearest way to describe Its state is that it is meditation. This
meditative state is that in which you approach nearest to the
Purusha. Thus we see why the meditative state is always called the
highest state by the Yogi, neither a passive nor an active state, but
the meditative state. This is the Sânkhya philosophy.
Next, the Sânkhyas say that this manifestation of Nature is for the
soul, all the combinations are for something outside of Nature. So
these combinations which we call Nature, these constant changes
are going on for the enjoyment of the soul, for its liberation, that it
may gain all this experience from the lowest to the highest, and
when it has gained it, the soul finds that it never was in Nature. It
was entirely separate, and it finds that it is indestructible, that it
neither goes nor comes, that going to heaven and being born again
were in Nature and not in the soul. So the soul becomes free. All of
Nature is working for the enjoyment and experience of the soul. It is
getting this experience in order to reach the goal, and that goal is
freedom. These souls are many, according to the Sânkhya
philosophy. There is an infinite number of souls. And the other
conclusion is that there is no God, as the Creator of the universe.
Nature herself is sufficient to produce all these forms. God is not
necessary, say the Sânkhyas.
Now we shall have to contest these three positions of the Sânkhyas.
First that intelligence or anything of that sort does not belong to the
soul, but that it belongs entirely to Nature; the soul being simply
qualitiless, colorless. The second point is that there is no God, but
Vedânta will show that without a God there cannot be any
explanation whatever. Thirdly, we shall have to contend that there
cannot be many souls, that there cannot be an infinite number, that
there is only One Soul in the universe, and that One is appearing as
many.
We will take the first proposition, that intelligence and reason belong
entirely to Nature, and not to the soul. The Vedânta says that the
soul is in its essence Existence-Knowledge-Bliss; but we agree with
the Sânkhyas that all that they call intelligence is a compound. For
instance, let us look at our perceptions. We remember that the chitta
(or the “mind-stuff”) is what is combining all these things, and upon

Textbook Geometry of Quantum States An Introduction To Quantum Entanglement Ingemar Bengtsson 2 Ebook All Chapter PDF

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Textbook Geometry of Quantum States An Introduction To Quantum Entanglement Ingemar Bengtsson 2 Ebook All Chapter PDF

Uploaded by

Copyright:

Available Formats

Geometry of Quantum States An

Introduction to Quantum Entanglement

Geometry of Quantum States An Introduction to Quantum

An Introduction to quantum optics and quantum

An Introduction to Quantum Physics First Edition

An Introduction To Quantum Field Theory First Edition

An Introduction to Quantum Transport in Semiconductors

Introduction to Quantum Mechanics David J. Griffiths

An Introduction to Quantum Optics Photon and Biphoton

Introduction to Quantum Field Theory with Applications

Quantum information theory is a branch of science at the frontiers of physics, mathematics

I N G E M A R B E N G T S S O N is a professor of physics at Stockholm University. After

is a professor at the Institute of Physics, Jagiellonian Univer-

Cambridge University Press is part of the University of Cambridge.

3.8 Cosets and all that 99

8 The space of density matrices 219

12.4 Mutually unbiased bases 326

16.7 Geometry of the set of separable states 466

Preface to the first edition

Preface to the second edition

1.1 Convex sets

1 Reproduced from [659].

Figure 1.4 A convex body is homeomorphic to a sphere.

Figure 1.6 Support hyperplanes of a convex set.

f (y) − f (x) ≥ (y − x) f (x). (1.13)

If f is twice differentiable it is convex if and only if its second derivative is non-

1.2 High dimensional geometry

X02 + X12 + · · · + Xn2 = r2 (1.14)

in an n + 1 dimensional flat space En+1 . A sphere of radius one is denoted Sn . The

The volume element dA on the unit sphere then becomes

dA = dφdθ1 . . . dθn−1 sin θ1 sin2 θ2 . . . sinn−1 θn−1 . (1.16)

You might also like