Quantum Information Processing From Theory To Experiment Dimitris G Angelakis

lehanamitai 13 views 85 slides May 19, 2025
Slide 1
Slide 1 of 85
Slide 1
1
Slide 2
2
Slide 3
3
Slide 4
4
Slide 5
5
Slide 6
6
Slide 7
7
Slide 8
8
Slide 9
9
Slide 10
10
Slide 11
11
Slide 12
12
Slide 13
13
Slide 14
14
Slide 15
15
Slide 16
16
Slide 17
17
Slide 18
18
Slide 19
19
Slide 20
20
Slide 21
21
Slide 22
22
Slide 23
23
Slide 24
24
Slide 25
25
Slide 26
26
Slide 27
27
Slide 28
28
Slide 29
29
Slide 30
30
Slide 31
31
Slide 32
32
Slide 33
33
Slide 34
34
Slide 35
35
Slide 36
36
Slide 37
37
Slide 38
38
Slide 39
39
Slide 40
40
Slide 41
41
Slide 42
42
Slide 43
43
Slide 44
44
Slide 45
45
Slide 46
46
Slide 47
47
Slide 48
48
Slide 49
49
Slide 50
50
Slide 51
51
Slide 52
52
Slide 53
53
Slide 54
54
Slide 55
55
Slide 56
56
Slide 57
57
Slide 58
58
Slide 59
59
Slide 60
60
Slide 61
61
Slide 62
62
Slide 63
63
Slide 64
64
Slide 65
65
Slide 66
66
Slide 67
67
Slide 68
68
Slide 69
69
Slide 70
70
Slide 71
71
Slide 72
72
Slide 73
73
Slide 74
74
Slide 75
75
Slide 76
76
Slide 77
77
Slide 78
78
Slide 79
79
Slide 80
80
Slide 81
81
Slide 82
82
Slide 83
83
Slide 84
84
Slide 85
85

About This Presentation

Quantum Information Processing From Theory To Experiment Dimitris G Angelakis
Quantum Information Processing From Theory To Experiment Dimitris G Angelakis
Quantum Information Processing From Theory To Experiment Dimitris G Angelakis


Slide Content

Quantum Information Processing From Theory To
Experiment Dimitris G Angelakis download
https://ebookbell.com/product/quantum-information-processing-
from-theory-to-experiment-dimitris-g-angelakis-906768
Explore and download more ebooks at ebookbell.com

Here are some recommended products that we believe you will be
interested in. You can click the link to download.
Quantum Information Processing From Theory To Experiment Dimitris G
Angelakis Et Al
https://ebookbell.com/product/quantum-information-processing-from-
theory-to-experiment-dimitris-g-angelakis-et-al-4106688
Quantum Information Processing 1st Edition Gerd Leuchs
https://ebookbell.com/product/quantum-information-processing-1st-
edition-gerd-leuchs-55210308
Quantum Information Processing 2nd Revised And Enlarged Edition Thomas
Beth
https://ebookbell.com/product/quantum-information-processing-2nd-
revised-and-enlarged-edition-thomas-beth-2100728
Quantum Information Processing And Quantum Error Correction An
Engineering Approach Djordjevic
https://ebookbell.com/product/quantum-information-processing-and-
quantum-error-correction-an-engineering-approach-djordjevic-22214962

Quantum Information Processing Quantum Computing And Quantum Error
Correction An Engineering Approach 2nd Edition Ivan B Djordjevic
https://ebookbell.com/product/quantum-information-processing-quantum-
computing-and-quantum-error-correction-an-engineering-approach-2nd-
edition-ivan-b-djordjevic-23549368
Quantum Information Processing Theory And Implementation 2nd Edition
Jnos A Bergou
https://ebookbell.com/product/quantum-information-processing-theory-
and-implementation-2nd-edition-jnos-a-bergou-34801000
Quantum Information Processing With Diamond Principles And
Applications 1st Edition Steven Prawer
https://ebookbell.com/product/quantum-information-processing-with-
diamond-principles-and-applications-1st-edition-steven-prawer-4737244
Quantum Information Processing With Finite Resources Mathematical
Foundations 1st Edition Marco Tomamichel Auth
https://ebookbell.com/product/quantum-information-processing-with-
finite-resources-mathematical-foundations-1st-edition-marco-
tomamichel-auth-5353996
Quantum Information Processing Lecture Notes Dieter Van Melkebeek
https://ebookbell.com/product/quantum-information-processing-lecture-
notes-dieter-van-melkebeek-51196972

QUANTUM INFORMATION PROCESSING

NATO Science Series
A series presenting the results of scientific meetings supported under the NATO Science
Programme.
The series is published by IOS Press and Springer Science and Business Media in conjunction
with the NATO Public Diplomacy Division.
Sub-Series
I.Life and Behavioural Sciences IOS Press
II.Mathematics, Physics and Chemistry Springer Science and Business Media
III.Computer and Systems Sciences IOS Press
IV.Earth and Environmental Sciences Springer Science and Business Media
V.Science and Technology Policy IOS Press
The NATO Science Series continues the series of books published formerly as the NATO ASI
Series.
The NATO Science Programme offers support for collaboration in civil science between
scientists of countries of the Euro-Atlantic Partnership Council. The types of scientific meeting
generally supported are “Advanced Study Institutes” and “Advanced Research Workshops”,
although other types of meeting are supported from time to time. The NATO Science Series
collects together the results of these meetings. The meetings are co-organized by scientists from
NATO countries and scientists from NATO’s Partner countries – countries of the CIS and
Central and Eastern Europe.
Advanced Study Institutes are high-level tutorial courses offering in-depth study of latest
advances in a field.
Advanced Research Workshops are expert meetings aimed at critical assessment of a field, and
identification of directions for future action.
As a consequence of the restructuring of the NATO Science Programme in 1999, the NATO
Science Series has been re-organized and there are currently five sub-series as noted above.
Please consult the following web sites for information on previous volumes published in the
series, as well as details of earlier sub-series:
http://www.nato.int/science
http://www.springeronline.nl
http://www.iospress.nl
http://www.wtv-books.de/nato_pco.htm
Series III: Computer and Systems Sciences – Vol. 199 ISSN: 1387-6694

Quantum Information Processing
From Theory to Experiment
Edited by
Dimitris G. Angelakis
University of Cambridge, United Kingdom
Matthias Christandl
University of Cambridge, United Kingdom
Artur Ekert
University of Cambridge, United Kingdom and
National University of Singapore
Alastair Kay
University of Cambridge, United Kingdom
and
Sergei Kulik
Moscow M.V. Lomonosov State University, Russia
Amsterdam • Berlin • Oxford • Tokyo • Washington, DC
Published in cooperation with NATO Public Diplomacy Division

Proceedings of the NATO Advanced Study Institute on
Quantum Computation and Quantum Information
Chania, Crete, Greece
2–13 May 2005
© 2006 IOS Press.
All rights reserved. No part of this book may be reproduced, stored in a retrieval system,
or transmitted, in any form or by any means, without prior written permission from the publisher.
ISBN 1-58603-611-4
Library of Congress Control Number: 2006927263
Publisher
IOS Press
Nieuwe Hemweg 6B
1013 BG Amsterdam
Netherlands
fax: +31 20 687 0019
e-mail: [email protected]
Distributor in the UK and Ireland Distributor in the USA and Canada
Gazelle Books Services Ltd. IOS Press, Inc.
White Cross Mills 4502 Rachael Manor Drive
Hightown Fairfax, VA 22032
Lancaster LA1 4XS USA
United Kingdom fax: +1 703 323 3668
fax: +44 1524 63232 e-mail: [email protected]
e-mail: [email protected]
LEGAL NOTICE
The publisher is not responsible for the use which might be made of the following information.
PRINTED IN THE NETHERLANDS

Quantum Information Processing v
D.G. Angelakis et al. (Eds.
IOS Press, 2006
© 2006 IOS Press. All rights reserved.
Introduction
By Artur EKERT
Chania, a picturesque little town on the coast of
the western Crete, the birthplace of Dimitris
Angelakis, the main organizer of this Advanced
Study Institute in Quantum Information Sci-
ence, is only a few miles away from the tiny
island of Antikythera. In 1900 a party of
sponge-divers were driven by a storm to anchor
near the island and there, at a depth of some
40 meters, they found the wreck of an ancient
cargo ship. Among the pieces of pottery and
marble statutes sprawled on the seabed was a
coral-encrusted lump of corroded bronze gear
wheels. The Antikythera mechanism, as it is
now known, was probably the world’s first
“analog computer” — a sophisticated device
for calculating the motions of stars and planets. This remarkable assembly of more than
30 gears with a differential mechanism, made on Rhodes or Cos in the first century
B.C., revised the view of what the ancient Greeks were capable of creating at that time.
A comparable level of engineering didn’t become widespread until the industrial revo-
lution nearly two millennia later. Thus it is hardly surprising that Richard Feynman,
who saw the Antikythera mechanism on display in Athens, called it “nearly impossi-
ble”. In one of his letters, reprinted in “What Do You Care What Other People Think?:
Further Adventures of a Curious Character”, he wrote
“… Yesterday morning I went to the archaeological museum. … Also, it was
slightly boring because we have seen so much of that stuff before. Except for one thing:
among all those art objects there was one thing so entirely different and strange that it
is nearly impossible. It was recovered from the sea in 1900 and is some kind of ma-
chine with gear trains, very much like the inside of a modern wind-up alarm clock. The
teeth are very regular and many wheels are fitted closely together…”
It seems likely that the Antikythera tradition of complex mechanical technology
was transmitted via the Arab world to medieval Europe where it formed the basis of
clockmaking techniques. As such, the Antikythera mechanism is a venerable precursor
of mechanical computing devices based on the meshing of metal gears. Indeed, for
many years the basic raw material of the computer industry was brass. The 17th cen-
tury calculators of Wilhelm Schickard, Blaise Pascal and Gottfried Wilhelm Leibniz
testify to the importance of gears in the history of computing. When, in 1837, Charles
Babbage was tinkering with a design of the first programmable computer, known as the
Analytical Engine, he was thinking in terms of rods, gears and wheels.

vi
From gears to relays to valves to transistors to integrated circuits and so on – in the
20th century brass gave way to silicon. Today’s advanced lithographic techniques can
etch logic gates and wires less than a micron across onto the surfaces of silicon chips.
Soon they will yield even smaller components, until we reach the point where logic
gates are so small that they consist of only a few atoms each. If computers are to con-
tinue to become faster (and therefore smaller), new, quantum technology must replace
or supplement what we have now, but it turns out that such technology can offer much
more than smaller and faster microprocessors. It can support entirely new modes of
computation, with new quantum algorithms that do not have classical analogues.
The very same person who was so fascinated by the ancient Antikythera laid down
the foundations of quantum computation. In 1981 Feynman observed that simulations
of some quantum experiments on any classical computer appear to involve an exponen-
tial slowdown in time as compared to the natural run of the experiment. Instead of
viewing this fact as an obstacle, Feynman regarded it as an opportunity. If it requires so
much computation to work out what will happen in a complicated quantum experiment
then, he argued, the very act of setting up an experiment and measuring the outcome is
tantamount to performing a complex computation. After all, any real computation is a
physical process, be it classical or quantum. Thus any computation can be viewed in
terms of physical experiments which produce outputs that depend on initial prepara-
tions called inputs. Since then, the hunt has been on for interesting things for quantum
computers to do, and at the same time, for the scientific and technological advances
that could allow us to build quantum computers.
The NATO Advanced Study Institute in Chania brought together a number of re-
searchers and students in both experimental and theoretical quantum information sci-
ence. During lectures and talks, and in numerous discussions over Raki, the participants
shared their views on just about everything; from quantum algorithms and intricacies of
computational complexity to the finer parts of Cretan cuisine, and from new technolo-
gies for realizing quantum computers to the spirit of traditional Greek dances. The
knowledge that nature can be coherently controlled and manipulated at the quantum
level was perceived as both a powerful stimulus and one of the greatest challenges fac-
ing experimental physics. Fortunately the exploration of quantum technology has many
staging posts along the way, each of which will yield scientifically and technologically
useful results and some of them are described in this volume.
We hope this collection of papers provides a good overview of the current state-of-
the-art of quantum information science. We do not know how a quantum Antikythera
will look like but all we know is that the best way to predict the future is to create it.
From the perspective of the future, it may well be that the real computer age has not yet
even begun.
We also wish to thank our sponsors NATO, the Cambridge-MIT Institute, the Un-
ion of Agricultural Cooperatives of Kidonia and Kissamos, the Cooperative Bank of
Chania, ANEK Lines, Olympic Airways, ABEA Olive Oil Products and Stigmes
Magazine. Finally we acknowledge the helpful collaboration from IOS Press in the
publication of this volume and also thank Kaija Hampson for being an excellent secre-
tary during the meeting.

vii
Contents
Introduction v
Artur Ekert
Chapter 1. Quantum Communication and Entanglement
Quantum Entanglement: Detection Methods and Usefulness as a Physical
Resource 3
Dagmar Bruß
On Quantum Cryptography with Bipartite Bound Entangled States 19
Paweł Horodecki and Remigiusz Augusiak
Unitary Local Permutations on Bell Diagonal States of Qudits and Quantum
Distillation Protocols 30
Hector Bombin and Miguel A. Martin-Delgado
Quantum Communication Channels in Infinite Dimensions 41
Alexander S. Holevo
Introduction to Relativistic Quantum Information 61
Daniel Terno
Generalized Bell Inequalities and the Entanglement of Pure States 83
Kwek Leong-Chuan, Chunfeng Wu, Jingling Chen, Dagomir Kaszlikowski
and C.H. Oh
Thermal Entanglement in Infinite Dimensional Systems 89
Aires Ferreira, Ariel Guerreiro and Vlatko Vedral
Improved Algorithm for Quantum Separability and Entanglement Detection 93
Lawrence M. Ioannou, Benjamin C. Travaglione, Donny Cheung
and Artur K. Ekert
Generalised Entanglement Swapping 99
Anthony J. Short, Sandu Popescu and Nicolas Gisin
Quantum Information Processing with Low-Dimensional Systems 103
Alexander Yu. Vlasov
Local Information and Nonorthogonal States 109
Jonathan Walgate
Optimal Alphabets for Noise-Resistant Quantum Cryptography 113
Denis Sych, Boris Grishanin and Victor Zadkov
Chapter 2. Quantum Algorithms and Error Correction
Quantum Algorithms and Complexity 121
Michele Mosca

viii
An Introduction to Measurement Based Quantum Computation 137
Richard Jozsa
Quantum Error Correction and Fault-Tolerance 159
Daniel Gottesman
Simulating Fourier Transforms for Open Quantum Systems in Higher
Encoding Bases 170
Ioannis N. Doxaras
Entanglement, Area Law and Group Theory 175
Radu Ionicioiu, Alioscia Hamma and Paolo Zanardi
A Quantum Algorithm for Closest Pattern Matching 180
Paulo Mateus and Yasser Omar
Classical and Quantum Fingerprinting in the One-Way Communication
Model 184
Anthony J. Scott, Jonathan Walgate and Barry C. Sanders
Chapter 3. Quantum Information Theory in Spin Systems
Introduction to Localizable Entanglement 191
Markus Popp, Frank Verstraete, Miguel A. Martin-Delgado
and Ignacio Cirac
State Transfer in Permanently Coupled Quantum Chains 218
Daniel Burgarth, Vittorio Giovannetti and Sougato Bose
Geometric Effects in Spin Chains 238
Marie Ericsson and Alastair Kay
Quantum Walks and Decoherence on a 1+1 Lattice 242
Isabelle Herbauts
Certain Aspects of Quantum Random Walk Asymptotics 246
Ioannis Smyrnakis
Chapter 4. Implementations of Quantum Information Processing
From Entanglement to Quantum Key Distribution 255
Hannes Hübel and Anton Zeilinger
Preparation and Measurement of Qutrits Based on Single-Mode Biphotons 281
Sergei Kulik
Cavity Quantum Electrodynamics: Quantum Entanglement and Information 294
Jean-Michel Raimond
Possibility of Quantum Computation by Utilizing Carbon Nanotubes
– Cooper Pair Splitting by Tomonaga-Luttinger Liquid – 312
Junji Haruyama, K. Murakami, J. Mizubayashi and N. Kobayashi
Implementing Quantum Processors in the Solid State 321
Crispin H.W. Barnes

ix
Quantum Field Trajectories Under Photon Number QND Measurement 326
Alexander A. Bukach and Sergei Ya. Kilin
Quantum Computation Beyond the “Standard Circuit Model” 330
Konstantinos Ch. Chatzisavvas, Costas Daskaloyannis
and Christos P. Panos
Entangled Light from Optical Time Boundaries 337
Ariel Guerreiro, Aires Ferreira, José T. Mendonça and Vlatko Vedral
Strong Light-Matter Coupling: Parametric Interactions in a Cavity and
Free-Space 341
Igor B. Mekhov, Valentin S. Egorov, Victor N. Lebedev,
Peter V. Moroshkin, Igor A. Chekhonin and Sergei N. Bagayev
Macroscopic Quantum Information Channel Via the Polarization-Sensitive
Interaction Between the Light and Spin Subsystems 346
Oksana S. Mishina, Dmitriy V. Kupriyanov and Eugene S. Polzik
From Network Complexity to Time Complexity Via Optimal Control 353
Thomas Schulte-Herbrüggen, Andreas Spörl, Navin Khaneja
and Steffen Glaser
Author Index 359

This page intentionally left blank

Chapter 1
Quantum Communication and Entanglement

This page intentionally left blank

Quantum Entanglement: Detection
Methods and Usefulness as a Physical
Resource
Dagmar Bruß
Institute for Theoretical Physics, Universit¨at D¨usseldorf, Germany
Abstract.Entanglement is one of the most fascinating features of quan-
tum information. In this lecture the focus is on two aspects of the broad
field of entanglement theory: first, on methods for the detection or ver-
ification of entanglement. Here, the method of Bell inequalities is com-
pared with the tool of witness operators and their local measurement.
An overview of other methods such as entropic and uncertainty rela-
tions, as well as physical approximations of unphysical maps, will be
given. Second, the zoo of multipartite entangled states will be visited,
mainly with respect to their usefulness as a resource for a given task.
As an example, distributed dense coding is presented.
As a connection between the two main topics, the detection of multi-
partite entangled states is discussed.
Keywords.Quantum entanglement, Bell inequalities, witness operators,
uncertainty relations, physical maps, multipartite entanglement
1. Detecting Entanglement
1.1. Bipartite Entanglement
As a reminder, let us recall the definition of bipartite entanglement for pure states.
We consider a pure state which is an element of a composite Hilbert space of two
subsystems, namely|ψffffiH
A⊗HB.Wecall|ψffseparable, iff
|ψff=|affΛ|bff, (1)
and entangled otherwise.
For mixed states, the definition of a separable state goes back to [1], and
reads:ffiis separable iff
ffi=

i
pi|aiffΨai|⊗|b iffΨbi|, (2)
with 0≤p
i≤1,

i
pi= 1, i.e. iff the density operator can be written as
a convex combination of projectors onto product states. Otherwiseffiis called
Quantum Information Processing
D.G. Angelakis et al. (Eds.)
IOS Press, 2006
© 2006 IOS Press. All rights reserved.
3

entangled. Note that eq. (2nota spectral decomposition, and that in general
Ψa
i|ajff⊗ =δ ij,Ψbi|bjff⊗ =δ ijmay hold.
There exist several operational separability criteria in the literature, for
overviews see [2]. For pure states a simple tool to find out whether a state is
entangled is the Schmidt decomposition. It states that any bipartite state can be
written as a bi-orthogonal superposition,

r
ff=
r

i
ai|eiff|fiff, (3)
where the Schmidt coefficientsa
ican be chosen positive,a i>0, and are nor-
malised,

r
i
a
2
i
= 1. The orthogonality condition readsΨe i|ejff=δ ij=Ψf i|fjff.
The Schmidt rankrhas to be smaller or equal to the minimum of the two di-
mensions of the two Hilbert spaces. The state|ψffis separable iffr=1.The
Schmidt rank can be easily determined by taking the partial trace over one of the
subsystems.
The most famous separability criterion for mixed states is the Peres-Horodecki
criterion [3,4], or criterion of positive partial transpose (PPT
pose is defined as the transpose of one of the two subsystems, e.g. the partial
transpose for Alice’s system reads
(ffi
TA
)mμ,nν=ffinμ,mν. (4)
Asher Peres realised that separability implies positivity of the partial transpose.
This can be seen directly from the definition of a separable state in eq. (2
reverse only holds in low dimensions, namely in dimensions 2×2and2×3, as
pointed out by the Horodecki family, who used properties of positive maps to
prove this statement.
1.2. Detecting Entanglement Via Bell Inequalities
One possibility to detect experimentally whether a given state is entangled relies
on a Bell inequality test. An example of a Bell inequality is the so-called CHSH-
inequality [5]. We assumea, b, c, dto be classical variables that can take the values
±1. Thus, (a+c)b+(−a+c)d=±2.If we consider several assignments for the
variables, and average over them, we arrive at
S=|Ψabff+Ψbcff+Ψcdff?Ψdaff| Ξ2, (5)
which is the CHSH inequality.
In quantum mechanics, we can consider observables that lead to assignments
to the variables. Explicitly, we can make a spin measurement in a specified direc-
tion. We thus arrive atΨabff=Tr[(Ψα·Ψσ
A)⊗(
Ψ
β·Ψσ B)ffi], whereΨσ
A(B)denotes the vec-
tor of Pauli matrices for Alice (Bob), andΨα(
Ψ
β) denotes a measurement direction
for Alice (Bob). By choosing appropriate measurement directions, as indicated in
figure 1, we find for the singlet|ψ

ff=
1

2
(|01ff?|10ff)thatS=2

2≥2, a
D. Bruß / Quantum Entanglement: Detection Methods and Usefulness as a Physical Resource4


4

4_π
4
αγ
β
δ
Figure 1.Optimal measurement directions on the Bloch sphere for testing the CHSH inequality.
clear indication of a violation of a CHSH inequality. This result is not compatible
with a local hidden variable model, and in particular it shows that the structure
of the singlet cannot be explained by classical correlations. However, as pointed
out by Werner [1], there exist mixed entangled states of two qubits that do not
violate any Bell inequality. Therefore, the tool of Bell inequalities is not the best
method to detect entanglement. This is not astonishing, as it was not designed
for entanglement detection, but for excluding local hidden variable models, i.e.
starting from purely classical arguments.
1.3. Detecting Entanglement Via Witness Operators
When designing a tool for the detection of quantum entanglement, it is advan-
tageous to start from the structure of quantum states. An important property
of quantum states is convexity: mixing two density operators leads to another
density operator. Separable states also form a convex set: from the definition in
eq. (2) it is clear that mixing two separable density operators (i.e. two sums of
projectors onto product states) leads to another separable density operator. The
concept of witness operators is based on this convex structure of quantum states,
which is illustrated in figure 2.
Entanglement witnesses were introduced in [4,6] and provide a necessary and
sufficient criterion for entanglement:ffiis entangled iff there exists a Hermitean
operatorWwith
Tr(Wffi)<0(6)
and Tr(Wffi
sep)≥0 for all separable statesffi sep.
The optimisation of witnesses has been studied (a witnessWis called optimal
if there is no witness that detects all the states that are detected byW,andsome
more in addition), and generic forms of witnesses have been constructed [7]. An
example for a witness for a stateffiwith non-positive partial transpose is given by
W=(|φffΨφ|)
TA
, (7)
withffi
TA
|φff=λ min|φff,whereλ mindenotes the minimal (negative) eigenvalue of
the partially transposed state. It is straightforward to see thatWis an entangle-
D. Bruß / Quantum Entanglement: Detection Methods and Usefulness as a Physical Resource5

Figure 2.Set of all states: convex subset of separable states (S
defined by witnessW, and by optimised witnessW
opt. The hatched area (in particular the state
ρ) is detected byW.
ment witness that detectsffi,asTr(Wffi)=Tr((|φffΨφ|)
TA
ffi)=Tr(|φffΨφ|ffi
TA
)=
Ψφ|ffi
TA
|φff=λ min<0andTr(Wffi sep)=Tr((|φffΨφ|)
TA
ffisep)=Tr(|φffΨφ|ffi
TA
sep)=
Ψφ|ffi
TA
sep|φff 0. Another example for an entanglement witness is a typical con-
struction that has been widely used in entanglement detection: given an entangled
pure state|ψff,
W
|ψffi=x·1l−|ψffΨψ| (8)
withx=max
|ϕsepffi|Ψψ|ϕ sepff|
2
is an optimal entanglement witness that de-
tects|ψffΨψ|. Again, this statement can be easily verified:Tr(W|ψffΨψ|)=
x?|Ψψ|ψff|
2
=x−1<0andTr(Wffi sep)=x?Ψψ|ffi sep|ψff 0. We ob-
serve thatWalso detects|ψffΨψ|plus some noise, as the noise only shifts the
expectation value towards zero, as long as it is small enough. The problem of
how to findxhas been solved in [8]. One writes|ψffas a certain bipartite su-
perposition, i.e.|ψff=

ij
Cij|iff|jff. A product state in the same bipartite
split reads|ϕ
sepff=|Aff|Bff,with|Aff=

i
Ai|iff,|Bff=

j
Bj|jff.Then
˜x=sup
A,B|

ij
A

i
B

j
Cij|
2
=maxλ(C), for the given bipartite split. In words,
˜xis the maximal (squared xis
then the maximum of ˜xoverallbipartite splits.
How can one use the mathematical concept of witness operators for the ex-
perimental detection of entanglement? In principle, one just has to determine the
expectation valueTr(Wffi). If this is negative, this is a clear signature of entan-
glement. However, how to measureW? In the spectral decomposition one will
find at least one entangled state. An important step for the implementation of
the witness is to realise thatanyobservable can be decomposed locally, and then
measured locally with present-day technology. A local decomposition of a wit-
ness is not unique. Therefore, in a further step one can find theoptimallocal
decomposition of a witness, where optimality here refers to the minimal number
of measurement settings. This concept has been developed in [9].
Let us note that a local decomposition of the witness
D. Bruß / Quantum Entanglement: Detection Methods and Usefulness as a Physical Resource6

W=

i
ci|aiffΨai|⊗|b iffΨbi|,ci∈1R,

i
ci=1, (9)
has to be a so-called pseudo mixture, where at least one of the coefficientsc
iis
negative. This is due to the fact that a witness operator is necessarily non-positive.
As a simple example, let us consider a Bell state of two qubits plus white
noise, namelyρ=p|ψ
+
ffΨψ
+
|+(1−p)1l/4, where|ψ
+
ff=
1

2
(|01ff+|10ff), and
pis a probability.
A suitable witness that is constructed according to the first method explained
above is of the formW=(|φffΨφ|)
TA
withρ
TA
|φff=λ min|φff, where here|φff=
α|00ff+β|11ffandα, βare parameters with|α|
2
+|β|
2
= 1, which can actually
be taken to be real.
It can be shown [9] that the optimal number of measurement settings for this
case is three. The optimal local decomposition reads
W=α
2
|z
+
z
+
ffΨz
+
z
+
|+β
2
|z

z

ffΨz

z

|
+αβ(|x
+
x
+
ffΨx
+
x
+
|+|x

x

ffΨx

x

|
−|y
+
y

ffΨy
+
y

|−|y

y
+
ffΨy

y
+
|) (10)
=
1
4
(1l⊗1l+σ
z⊗σz+(α
2
−β
2
)(σz⊗1l+1l⊗σ z)
+2αβ(σ
x⊗σx+σy⊗σy)) (11
where|z
±
ff=|0ff,|1ff;|x
±
ff=
1

2
(|0ff?|1ff)and|y
±
ff=
1

2
(|0ff?i|1ff).
Witnesses of this type were first measured in the experiment by F. DeMartini
and collaborators [10]. Note that for the two-qubit case considered here one could
also just perform state tomography, i.e. measure all elements of the density matrix.
By applying the partial transpose to the reconstructed density matrix one can
determine whether the produced state was entangled or not. In higher dimensions
or for more than two subsystems this is in general not possible, because there exist
bound entangled states which have a positive partial transpose. Therefore, the
tool of witness operators becomes more powerful for the detection of entanglement
than any previously known concept, when higher-dimensional and/or multipartite
systems are considered, see subsection 2.4.
Note that recently the two concepts of Bell inequalities (which can always be
reformulated as a non-optimal witness) and entanglement witnesses (which detect
more states as entangled than Bell inequalities, and are thus more powerful) has
been compared in [11].
1.4. Detecting Entanglement Via Uncertainty Relations
Recently, the method of using uncertainty relations for the detection of entan-
glement has been developed. There are two main ideas in this direction, namely
variance based uncertainty relations and entropic uncertainty relations.
D. Bruß / Quantum Entanglement: Detection Methods and Usefulness as a Physical Resource7

1.4.1. Variance Based Uncertainty Relations
Variance based uncertainty realtions forlocalobservables were introduced by Hof-
mann and Takeuchi [12] and generalised tonon-localobservables by G¨uhne [13].
The variance of an observableMis defined as
δ
2
(M)Ψ=Ψ(M?ΨMff Ψ)
2
ff=ΨM
2
ffΨ?ΨMff
2
Ψ
, (12)
whereΨMff
Ψ=Tr(ffiM). A well-known uncertainty relation for non-commuting
observablesPandXreadsδ(P)δ(X)≥ff/2. A general formulation of an uncer-
tainty relation for non-commuting observablesM
iis given as

i
δ
2
(Mi)≥C>0. (13)
In general, however, it is difficult to find the constantC.
The method for detection of entanglement consists in choosing{M
i}, finding
a lower boundCfor all product states, and then using concavity of the variance.
Let us first show that the variance is concave with respect to a decomposition
ffi=

k
pkffik:
δ
2
(Mi)Ψ=

k
pkΨ(Mi?ΨM iffΨ)
2
ffΨk
=

k
pk
Λ
ΨM
2
i
ffΨk
?ΨM iff
2
Ψ
k
+ΨM iff
2
Ψ
k
−2ΨM iffΨk
ΨMiffΨ+ΨM iff
2
Ψ
Ψ
=

k
pk
Λ
δ
2
(Mi)Ψk
+(ΨM iffΨk
?ΨM iffΨ)
2
Ψ


k
pkδ
2
(Mi)Ψk
, (14)
and therefore

i
δ
2
(Mi)Ψ≥

k
pk

i
δ
2
(Mi)Ψk
. (15)
Concavity of the variance means that one cannot decrease the uncertainty by
mixing states. Thus, once having established a lower boundCof the variance for
product states, a state with

i
δ
2
(Mi)<Cis entangled.
The local observables introduced in [12] read
M
i=Ai⊗1l+1l⊗B i. (16)
The local variances have lower bounds

i
δ
2
(Ai)ΨA
≥CAand

i
δ
2
(Bi)ΨB
≥CB.
It was shown that

i
δ
2
(Mi)Ψ≥CA+CB, (17)
D. Bruß / Quantum Entanglement: Detection Methods and Usefulness as a Physical Resource8

whenffiis separable. One problem of this approach is the generalisation to the mul-
tipartite scenario: no local uncertainty relation for the detection of multipartite
entanglement is known yet.
In the approach of [13] non-local observables were introduced. It was shown
that for any entangled pure bipartite state|ψff, there exists a set of observ-
ables{M
i}such that|ψffviolates an uncertainty relation. As an example for
the case that one subsystem is two-dimensional, i.e.|ψff=a|00ff+b|11ffin
the Schmidt decomposition, one can chooseM
i=|ψ iffΨψi|,i=1...4, with
{|ψ
1ff,|ψ 2ff,|ψ 3ff,|ψ 4ff}={|ψff,|ψff

,|ψ3ff=a|01ff+b|10ff,|ψ 3ff

},where|ψff

denotes the state orthogonal to|ψff. It can be shown that for separable states

i
δ
2
(Mi)≥2a
2
b
2
. However, for the state|ψffone finds

i
δ
2
(Mi)=0,which
proves that|ψffis entangled, unlessa=0orb=0.
It is important to note that non-local measurements can be decomposed lo-
cally [9], and thus the non-locality of the observables is not a practical problem.
For the detection of bound entangled states, e.g. the UPB states, one can find
non-local measurement bases such that

i
δ
2
(Mi) = 0. Furthermore, multipartite
entanglement can be detected via a generalisation of the ideas described above.
1.4.2. Entropic Uncertainty Relations
Entropic uncertainty relations were discussed in [14,15,16]. Again, one measures
observablesM
i. This leads to a probability distribution, with an according entropy
S(M
i). An entropic uncertainty relation for non-commuting observablesM ireads

i
S(M i)≥C>0. (18)
Note that the difference of this entanglement criterion to other entropic entan-
glement criteria such as the reduction criterion lies in the fact that there the
eigenvalues offfi(i.e. outcomes of a measurement in the eigenbasis) are employed,
whereas here the outcomes of more general measurements are used.
The idea for the detection of entanglement via entropic uncertainty relations is
similar to the variance based uncertainty relations discusssed in 1.4.1: one chooses
observables{M
i}, finds a lower boundCfor all product states, and uses concavity
of the entropy, which leads to the fact that a state with

i
S(M i)<Cis detected
as entangled.
An important open question is which entropy is the most suitable for the
detection of entanglement. (In the literature the Tsallis entropy has been widely
used.) Let us remark that the concept of entropic uncertainty relations is ex-
tendible to the detection of multipartite entanglement.
1.5. Physical Approximations of Positive Maps
A completely different idea for the detection of entanglement are physical ap-
proximations of positive maps [17]. Let us remember the separability criterion
based on positive, but not completely positive (CP) maps: a bipartite stateffiis
separable iff [1l⊗Λ](ffi)≥0 for all positive maps Λ. For the detection of entangle-
ment only those maps are interesting which are not completely positive, i.e. an
D. Bruß / Quantum Entanglement: Detection Methods and Usefulness as a Physical Resource9

extension of which in some higher-dimensional space is not positive. Maps which
are not CP are considered to be unphysical.
The idea in [17] is to add a suitable proportion of the identity map to [1l⊗Λ],
such that the new map is positive, i.e. a physical map. This physical map for a
d×dsystem is given as
[ffi1l⊗Λ](ffi)=p
[1l⊗1l]
d
2
(ffi)+(1−p)[1l⊗Λ](ffi), (19)
wherep≥d
4
|λ|/(d
4
|λ|+1), andλ<0 is the most negative eigenvalue, obtained
after acting with the original map on a maximally entangled state.
We thus arrive at a modified separability criterion:ffiis separable iff
[ffi1l⊗Λ](ffi)≥
d
2
λ
d
4
λ+1
for all positive maps Λ. An important property of this ap-
proach is that in order to check this criterion, one only has to estimate a single
parameter, namely the smallest eigenvalue of the state after applying the physical
map. A problem of this entanglement criterion is that it requires the application
of non-local operations, which are difficult to implement. Therefore, this criterion
has not yet been realised in an experiment.
2. Usefulness of Multipartite Entanglement
2.1. Multipartite Entanglement
One can define the notion of multipartite entanglement as a generalisation of
the bipartite case: A pure multipartite state|ψffffiH
A⊗HB⊗...⊗H Nis fully
separable orn-separable, iff
|ψff=|affΛ|bffΛ...⊗|nff. (20)
If|ψffcan be written as a tensor product ofn−1 subsystems, it is calledn−1-
separable, and so forth. Note thatk-separability impliesk−1-separability. The
state|ψffis called biseparable iff|ψff=|s
1ffΛ|s 2ff.If|ψffis not biseparable , it
is genuinely multipartite entangled.
A mixed stateffi,actingonH
A⊗HB⊗...⊗H N, is fully separable orn-separable,
iff
ffi=

i
pi|aiffΨai|⊗|b iffΨbi|⊗...⊗|n iffΨni|, (21)
where 0≤p
i≤1,

i
pi= 1. Again, ifffican be written as a mixture of projectors
onto tensor products ofn−1 subsystems, it is calledn−1-separable, and so
forth. The stateffiis called biseparable iffffi=

i
pi|s1iffΨs1i|⊗|s 2iffΨs2i|. Here,
the terms contributing to the mixture can have different splittings. Ifffiis not
biseparable, it is genuinely multipartite entangled.
Unfortunately, we do not have many operational separability criteria in the
multipartite case. Even for pure states, the Schmidt decomposition does not al-
ways exist. For mixed states, the Peres-Horodecki criterion of a positive partial
transpose is still necessary for separability across a bipartite split.
D. Bruß / Quantum Entanglement: Detection Methods and Usefulness as a Physical Resource10

2.2. A Zoo of Multipartite Entangled States
Various multipartite entangled states have been studied in the literature, and
produced in experiments. By looking at their structure, we can understand the
main properties of some of the “animals” in the multipartite entanglement zoo.
A well-known, but very vulnerable state is the GHZ state
|GHZff=
1

2
(|000ff+|111ff), (22)
here for three qubits. Loosing (i.e. tracing out) one of the particles means that
no entanglement remains for the other subsystems. This state is as fragile as a
butterfly.
A state that is inequivalent to the GHZ state is the W state,
|Wff=
1

3
(|001ff+|010ff+|100ff), (23)
here for three qubits. Loosing one of the particles here means that some entan-
glement remains for the other subsystems. Like a rainworm, this state can be cut
into two pieces, and still remain entangled.
The singlet state for four qubits reads

(4)
ff=
1

3
(|0011ff+|1100ff?|ψ
+
ff|ψ
+
ff), (24)
where|ψ
+
ff=
1

2
(|01ff+|10ff) is a Bell state. When changing the basis in all
subsystems in the same way, this state remains invariant - like the chamaleon,
the colour may change, but the shape remains.
Finally, an important animal in the zoo of entangled states is a cluster state,
for example for four qubits:
|C
4ff=
1

4
(|+0+0ff+|+0−1ff+|−1−0ff+|−1+1ff), (25)
where|?ff=
1

2
(|0ff?|1ff). This state and its generalisations allow to implement
the one-way quantum computer, a very powerful concept in quantum information.
This is the lion in the zoo of multipartite entangled states.
We do not aim at a complete zoology of entangled states in this lecture.
Regarding their usefulness, with the GHZ states one can e.g. perform quantum
teleportation, multi-party quantum key distribution (QKD), secret sharing, dis-
tributed dense coding, quantum imaging, and improve frequency standards. The
W states are useful for multi-party QKD, secret sharing, and distributed dense
coding. Singlet states can be employed for error protection (decoherence free sub-
space), multi-party QKD, secret sharing, and telecloning. As mentioned above,
cluster states are useful for one-way quantum computing.
In order to provide a full classification of multipartite states, one should spec-
ify the quantum information processing task according to which the states are
classified. A state may be useful for one task, and not useful for another: the
classification scheme may thus depend on the task.
D. Bruß / Quantum Entanglement: Detection Methods and Usefulness as a Physical Resource11

U
1
U
2
...
oo o o
M
Figure 3.Original protocol for superdense coding with two parties.
2.3. Example: Distributed Dense Coding
As an example for the usefulness of multipartite entangled states for quantum
information processing we will consider distributed dense coding [18].
Let us remind the reader of the original protocol for superdense coding with
2 qubits, proposed in [19]. As sketched in figure 3, we have one sender (Alice)
and one receiver (Bob). Using entanglement, they manage to encode two classical
bits into one qubit, sent from Alice to Bob. The protocol consists of three steps,
as follows: i) Alice and Bob share the Bell state|ψ

ff=
1

2
(|01ff?|10ff). ii)
Alice applies one of the unitary rotationsU
i∈{1l,σ x,σy,σz}, occurring with
equal probabilitiesp
i=1/4, to her part of the state, and sends it to Bob. iii)
Bob performs a Bell measurement on thetotalstate, thus gaining two bits of
information.
The above protocol can be generalised to dense coding with higher dimen-
sional bipartite states, that are not necessarily pure. Alice and Bob now share a
stateρ
AB
. Alice performs a unitary operation from the set{U i}with probabilities
{p
i}on her part of the system, and sends it to Bob. The ensemble of possible total
states is thus{p
i,ρ
AB
i
}withρ
AB
i
=Ui⊗1ldB
ρ
AB
U

i
⊗1ldB
. The Holevo bound
[20] describes Bob’s accessible information (which is attainable asymptotically),
I
acc=S(
ρ)−

i
piS(ρ
AB
i
), (26)
whereS(ξ)=−tr(ξlogξ) is the von Neumann entropy ofξ, and the average state
is denoted as
ρ=

i
piρ
AB
i
.
We call the maximal accessible information (optimised over all unitary en-
codings) thecapacity of dense coding. For bipartite states it is given as
χ=max
{Ui},{p i}
Iacc=log
2dA+S(ρ
B
)−S(ρ
AB
). (27)
This can be seen by noting that the second term in eq. (26) is equal to−S(ρ
AB),
and by using the well-known fact thatS(
ρ
AB
)≤S(ρ
A
)+S(ρ
B
), together with
S(ρ
A
)≤log
2dA. The equal sign in the last two inequalities is reached by an
orthogonal complete set{W
j}with equal probabilitiesp jand the trace condition
1
dA

j
W

j
ΞWj=tr[Ξ]1l. It is an open question whether non-unitary encodings
may increase the dense coding capacity.
We can define the notion ofdense codeabilityas follows. A state is useful for
dense coding (dense codeable) ifχ>log
2dA, or equivalently if
S(ρ
B
)>S(ρ
AB
). (28)
D. Bruß / Quantum Entanglement: Detection Methods and Usefulness as a Physical Resource12

U
1
U
2
...
oo
oo
oo
U
1
U
2
...
U
1
U
2
...
M
o o...
...
Figure 4.Distributed dense coding with N senders and one receiver.
We can make some observations and draw conclusions from the condition in eq.
(28).
•For separable states, eq. (28) is never fulfilled, i.e. separable states are never
useful for dense coding.
•From eq. (28) follows thatρ
AB
violates the reduction criterion [21]. Re-
member that the reduction criterion reads 1l
dA
⊗ρ
B
≥ρ
AB
. If the reduc-
tion criterion is violated, then the stateρ
AB
is distillable [22]. Thus we can
conclude that bound entanglement is never useful for dense coding. (For
d×d-systems, see [23].)
•Anypurebipartite entangled state is useful for dense coding.
•There existmixed entangledstates in dimension 2×2, which arenotuseful
for dense coding, e.g. Werner states withF<0.748. Thus it is not only
the degree of entanglement which is important for dense codeability, but
also the degree of mixedness. As quantum teleportation is possible for all
entangled two-qubit states, dense coding and teleportation are inequivalent
tasks.
Let us generalise dense coding to the case of N senders, so-calleddistributed
dense coding[18].
The protocol for distributed dense coding is indicated in figure 4 and proceeds
as follows. TheNAlices and Bob share anN+ 1-partite stateρ
A1...ANB
. Alice
numberjappliesU
Aj
ij
with probabilityp
Aj
ij
. It is important to find out whether
a joint action of the senders can improve the capacity of dense coding. All Alices
send their respective states to Bob, who makes a global measurement on the total
state. It turns out that the Holevo bound can be achieved forproduct encodings,
namely a complete orthogonallocalset{W
i=⊗jW
Aj
ij
}with the trace property
as above. Thus, we arrive at the capacity of dense coding with a single receiver,
χ
A1...ANB
=log
2dA1
+...+log
2dAN
+S(ρ
B
)−S(ρ
A1...ANB
).(29)
Surprisingly, the senders donotneed to perform global unitaries to achieve the
capacity.
What can be said about distributed dense coding with two receivers?
D. Bruß / Quantum Entanglement: Detection Methods and Usefulness as a Physical Resource13

U
1
U
2
...
oo
oo
U
1
U
2
...
M
o o
oo
U
1
U
2
...
oo
U
1
U
2
...
M
o o
...
...
i) no commun.
ii) LOCC
iii) global meas.
...
...
Figure 5.Distributed dense coding with N senders and two receivers.
The protocol, as sketched in figure 5, is now described as follows. TheNAlices
and 2 Bobs share anN+2-partite stateρ
A1...ANB1B2
. Alice numberjappliesU
Aj
ij
with probabilityp
Aj
ij
to her part of the state. The AlicesA 1, ...Aksend their states
toB
1; the AlicesA k+1, ...ANsend their states toB 2. There are now three different
scenarios for the two Bobs, depending on which type of measurement they are
allowed to make. i) No communication: the dense coding capacities are additive.
ii) Local operations and classical commuication (LOCC): there is a Holevo-like
upper bound, as derived in [24]. iii) Global measurement: here we are back to the
situation with one receiver (with a system of higher dimension).
From these considerations we find the following bounds on the LOCC-capacity
of dense coding with 2 receivers:
χ
LOCC
≥χ
B1

B2

B1B2
, (30)
χ
LOCC
≤logd A1
+...+logd AN
+S(ρ
B1B2
)−S(ρ)=χ
glob
, (31)
χ
LOCC
≤logd A1
+...+logd AN
+S(ρ
B1
)+S(ρ
B2
)−max
x=1,2
S(ρ
x
) (32)
≡B
LOCC
.
In eq. (32) the bound derived in [24] has been employed.
We can now introduce a classification scheme for quantum states, according
to their dense codeability [18]. In figure 6 we illustrate this classification, for
two receivers: The convex set “S” denotes separable states. The next shell (not
including the separable states) are the bound entangled states with positive partial
transpose (PBE). The shell “NBE?” are the states that have non-positive partial
transpose and are conjectured to be undistillable. “D” stands for distillable states.
“G-DC” are the states that are globally dense codeable. The next shell, “LOCC-
DC”, denotes the states that are LOCC dense codeable, but not locally dense
codeable. Locally dense codeable states are denoted as “LO-DC” and form the
outermost shell. Note that for one receiver the three outermost shells collaps into
one shell.
D. Bruß / Quantum Entanglement: Detection Methods and Usefulness as a Physical Resource14

1) 2) 3)
G−DCDNBE?PBES LOCC−DCLO−DC
Figure 6.Classification of quantum states according to their dense codeability. For notation see
main text.
We will now give examples for the three shells connected to dense coding and
thus show that they are non-empty. They are also not of measure zero, because
one can add a finite proportion of the identity without loosing the property of
dense-codeability.
1)Example for G-DC, but not LOCC-DC:

V DDVff=
1
2
(|0000ff+|0101ff+|1000ff+|1110ff). (33)
2)Example for LOCC-DC, but not LO-DC:

GHZff=
1

2
(|0000ff+|1111ff). (34)
The upper bound given in eq. (32χ
LOCC
≤3. This can be shown to be
reachable by an explicit protocol, and thereforeχ
LOCC
=3.
3)Example for LO-DC:

2Bellff=
1

2
(|00ff+|11ff)⊗
1

2
(|00ff+|11ff). (35)
We have thus given a classification of multipartite quantum states according to
a specific quantum information processing task. Different tasks are expected to
lead to different classifications.
2.4. Detecting Multipartite Entanglement Via Witness Operators
In this subsection we connect the topics of sections 1 and 2, and address the
problem of detecting multipartite entanglement.
Generalised Bell inequalities for multipartite states have been designed for
example in [25,26,27] and tested experimentally. The Mermin-Klyshko inequali-
ties are a straightforward generalisation of the CHSH inequality. Each of then
parties has two measurements with two outcomes each. We denote the dichotomic
D. Bruß / Quantum Entanglement: Detection Methods and Usefulness as a Physical Resource15

measurements asA iandA
Ξ
i
,withi=1, ..., n, and the outcomes asa i,a
Ξ
i
=±1.
Then one can define a correlation operator in a recursive way,
M
n=
1
2
Mn−1(an+a
Ξ
n
)+
12
M
Ξ
n−1
(an−a
Ξ
n
), (36)
withM
1=a1,M
Ξ
1
=a
Ξ
1
. Using the same line of argument as for CHSH one finds
that for a local hidden variable model
|ΨM
nff| Ξ1 (37)
has to hold. This inequality is violated by many genuinely multipartite entan-
gled states. For example, the GHZ-state leads to a maximal violation, namely
ΨGHZ|M
n|GHZff=2
(n−1)/2
. However, there are multipartite states that do not
violate the Mermin-Klyshko inequality, and thus it does not allow for general
detection and/or classification of genuine multipartite entanglement.
The method of detecting multipartite entanglement via witness operators has
been developed in [9,28]. The idea is very similar to the one described for the
detection of bipartite entanglement in section 1.3: one first constructs a suitable
witness operatorW. This operator is constructed such that it is positive on all
biseparable states, and therefore detects genuine multipartite entanglement. For
the construction of the witness one has to take into account the convex structure
of the set of states. For three qubits this has been described in [29]. In order to
find out whether a given stateffiis multipartite entangled, one then measures the
expectation value of the witness. If
Tr(Wffi)<0, (38)
it follows thatffiis genuinely multipartite entangled. In order to be able to mea-
sure the witness with present-day technology, it is decomposed locally. Theopti-
maldecomposition is the one that needs the minimal number of local projection
measurements.
In a collaboration between the theory group of M. Lewenstein at the Uni-
versity of Hannover and the experimental group of H. Weinfurter at LMU Mu-
nich witnesses were implemented for the first time for the detection of multipar-
tite entanglement [8]. The states that were experimentally created with polarised
photons are the W state and the four-photon singlet state, which read
|Wff=(|001ff+|010ff+|100ff)/

3 (39)
and

(4)
ff=
1

3
Ξ
|0011ff+|1100ff?
1
2
(|0110ff+|1001ff+|0101ff+|1010ff)

.
(40)
The witnesses for their detection are given as
D. Bruß / Quantum Entanglement: Detection Methods and Usefulness as a Physical Resource16

W
(1)
W
=
2
3
1l−|WffΨW|, (41)
W
(2)
W
=
1
2
1l−|GHZffΨGHZ|, (42)
W
Ψ
(4)=
3
4
1l−|Ψ
(4)
ffΨΨ
(4)
|, (43)
where|
GHZffdenotes a GHZ state in a rotated basis. These witnesses have been
decomposed locally [28].
An example for the minimal decomposition of the witness in eq. (41) is given
as
W
W=
2
3
1l−|WffΨW|
=
1
24

17·1l
⊗3
+7·σ
⊗3
z
+3·

σ z1l1l+1lσ z1l+1l1lσ z

+5·

σ
zσz1l+σ z1lσz+1lσ zσz

−(1l+σ
z+σx)
⊗3
−(1l+σ z−σx)
⊗3
−(1l+σ z+σy)
⊗3
−(1l+σ z−σy)
⊗3

, (44)
which needs only 5 measurement settings. For the interpretation as projection
measurements, remember that
Tr[σ
z⊗σzρ]=Ψz
+
z
+
|ρ|z
+
z
+
ff?Ψz
+
z

|ρ|z
+
z


?Ψz

z
+
|ρ|z

z
+
ff+Ψz

z

|ρ|z

z

ff. (45)
Note that tripartite entanglement with noise can be detected with the same wit-
ness, as long as the noise is not too large. AsTr(W
W|WffΨW|)=−1/3, the state
ffi=p|WffΨW|+(1−p)1l/8 is detected to be genuinely multipartite entangled for
allp>13/21.
The experimental results when measuring the above witnesses are as follows:
Tr(W
(1)
W
ρW)exp=−0.197±0.018 ( theory: =−1/3), (46)
Tr(W
(2)
W
ρW)exp=−0.139±0.030 ( theory: =−1/4), (47)
Tr(W
Ψ
(4)ρ
Ψ
(4))exp=−0.151±0.01 ( theory: =−1/4). (48)
Thus, the existence of genuine multiparticle entanglement has been confirmed
beyond any doubt. Note that the systematic shift of the experimental values with
respect to the theoretical expectation is due to noise, which moves the negative
expectation value of a witness towards zero. These results also confirm in a very
convincing way the power of the tool of entanglement witnesses.
Acknowledgements
The research presented in this lecture resulted from collaborations with Mohamed
Bourennane, Mauro D’Ariano, Manfred Eibl, Artur Ekert, Sascha G¨artner, Ot-
D. Bruß / Quantum Entanglement: Detection Methods and Usefulness as a Physical Resource17

fried G¨uhne, Philipp Hyllus, Christian Kurtsiefer, Maciej Lewenstein, Chiara
Macchiavello, Anna Sanpera, Aditi Sen(De
whom I wish to thank for many enlightening and exciting discussions. This work
was supported in part by the EU IST project QUPRODIS.
References
[1] R. Werner, Phys. Rev. A40, 4277 (1989
[2] M. Lewenstein et al, J. Mod. Opt.47, 2841 (2000 43,
4237 (2002 49, 1399 (2002
[3] A. Peres, Phys. Rev. Lett.77, 1413 (1996
[4] M. Horodecki, P. Horodecki, and R. Horodecki, Phys. Lett. A223, 1 (1996
[5] J. Clauser, M. Horne, A. Shimony, and R. Holt, Phys. Rev. Lett.23, 880 (1969
[6] B. Terhal, Phys. Lett. A271, 319 (2000
[7] M. Lewenstein, B. Kraus, I.J. Cirac, and P. Horodecki; Phys. Rev. A62, 052310
(2000).
[8] M. Bourennane, M. Eibl, C. Kurtsiefer, H. Weinfurter, S. G¨artner, O. G¨uhne, P.
Hyllus, D. Bruß, M. Lewenstein, and A. Sanpera, Phys. Rev. Lett.92, 087902 (2004
[9] O. G¨uhne, P. Hyllus, D. Bruß, A. Ekert, M. Lewenstein, C. Macchiavello and A.
Sanpera, Phys. Rev. A66, 062305 (2002
[10] M. Barbieri et al, Phys. Rev. Lett.91, 227901 (2003
[11] P. Hyllus, O. G¨uhne, D. Bruß, and M. Lewenstein, Phys. Rev. A72, 012321 (2005
[12] H. Hofmann and S. Takeuchi, Phys. Rev. A68, 032103 (2003
[13] O. G¨uhne, Phys. Rev. Lett.92, 117903 (2004
[14] R. Horodecki and P. Horodecki, Phys. Lett. A194, 147 (1994
[15] V. Giovanetti, Phys. Rev. A70, 012102 (2004
[16] O. G¨uhne and M. Lewenstein, Phys. Rev. A70, 022316 (2004
[17] P. Horodecki and A. Ekert, Phys. Rev. Lett.89, 127902 (2002
[18] D. Bruß, G. M. D’Ariano, M. Lewenstein, C. Macchiavello, A. Sen(De
Phys. Rev. Lett.93, 210501 (2004
[19] C.H. Bennett and S.J. Wiesner, Phys. Rev. Lett.69, 2881 (1992
[20] A.S. Holevo, IEEE Trans. Inf. Theory44, 269 (1998
[21] K. Vollbrecht and M. Wolf, quant-ph/0202058.
[22] M. Horodecki and P. Horodecki, Phys. Rev. A59, 4206 (1999
[23] M. Horodecki et al, Quant. Inf. Comp.1, 70 (2001).
[24] P. Badzi¸ag, M. Horodecki, A. Sen(De 91, 117901
(2003).
[25] N. Mermin, Phys. Rev. Lett.65, 1838 (1990
[26] A. Belinskii and D. Klyshko, Phys. Usp.36, 653 (1993
[27] D. Collins, N. Gisin, S. Popescu, D. Roberts, and V. Scarani, Phys. Rev. Lett.88,
170405 (2002
[28] O. G¨uhne et al, J. Mod. Opt.50, 1079 (2003
[29] A. Ac´
ın, D. Bruß, M. Lewenstein and A. Sanpera, Phys. Rev. Lett.87, 040401
(2001).
D. Bruß / Quantum Entanglement: Detection Methods and Usefulness as a Physical Resource18

On Quantum Cryptography With
Bipartite Bound Entangled States
Pawe∗l Horodecki
a
, Remigiusz Augusiak
a
,
a
Faculty of Applied Physics and Mathematics, Gda´nsk University of Technology,
Narutowicza 11/12, PL 80-952 Gda´nsk, Poland
Abstract.Recently the explicit applicability of bound entanglement in
quantum cryptography has been shown. In this paper some of recent
results respecting this topic are reviewed. In particular relevant no-
tions and definitions are reminded. A new construction of bound en-
tangled states containing secure correlations is presented. It provides
low-dimensional 6⊗6 bound entangled states with nonzero distillable
key.
Keywords.quantum cryptography, quantum security, secure key
distillation, bound entanglement
1. Introduction
The explicit application of quantum entanglement in quantum information theory
was the cryptographic protocol by Ekert [1]. The essential point of the protocol
(cf. further modification [2]) was the entanglement monogamy principle (see [3])
which says that if the two particles are maximally entangled with each other then
they are completely unentangled with any other (third
any correlation measurements on both particles must be completely safe from
the cryptographic point of view, as they are uncorrelated form results of any
other measurement performed on the rest of the world. This point was further
exploited in a nice application [4] of entanglement distillation [5] (cf. [6]). The
idea of Ref. [4] is called quantum privacy amplification (QPA). Given stationary
source of pure states

∗Ψ
ABE
γ
describing Alice, Bob (which are cooperating) and
Eve (eavesdropper) quantum correlations protocol QPA is focused on distilling
maximally entangled states


Ψ
(d)
+
γ
=
1

d
d−1
η
i=0


i
γ



i
γ
(1.1)
from entangled quantum states

AB=TrE


Ψ
ABE
γσ
Ψ
ABE


. (1.2)
Quantum Information Processing
D.G. Angelakis et al. (Eds.)
IOS Press, 2006
© 2006 IOS Press. All rights reserved.
19

The distillation protocol uses local operations and classical communication
(LOCC
communication (LOPC


Ψ
(d)
+
γ
by entanglement monogamy they share logdbits of classical secure bits.
This can be done by performing local measurements on the state (1.1
bases{|iη
A}
d−1
i=0
and{|iη B}
d−1
i=0
.
However, since 1998 it has been known that there is quantum entanglement
called bound entanglement that can not be distilled to a pure form [7]. For a long
time there was a common belief that distillation of secure key from quantum state
is possible only when QPA is. In other words, that bound entanglement is useless
for quantum cryptography. In fact the results of extensive analysis of two qubit
case [8] naturally suggested equivalence of entanglement distillation protocols and
secure key distillation.
Surprisingly it is not true, as it has been shown in papers [9,10]. Before we shall
recall main observations of the latter, let us point out the key ingredient of their
reasoning. Namely, why QPA might be not necessary for distilling secure key? In
fact, if Alice and Bob share maximally entangled state (1.1
much stronger security than they need. In fact they will get secure correlations
if they measure the state inanypair of bases of the form:{

∗e
(A)
i
γ
=U|iη
A}
d−1
i=0
and{

∗e
(B)
i
γ
=U

|iηB}
d−1
i=0
,whereUstands for arbitrary unitary operation. It is
crucial to understand that in quantum cryptography it would be enough to have
a single basis, measurement in which could give secure correlations. To be more
specific, the security requirement is to get (via local measurement of Alice and
Bob in one basis) the state that is product with Eve’s degrees of freedom.
This leads us to the notion of ccq state [12], i.e., tripartite state of Alice,
Bob and Eve that is, after local von Neumann measurements, classical on Alice
and Bob parts and quantum on Eve part. With this notion one can summarize
the idea of [9,10] as follows: Alice and Bob should proceed to distill such a state
γ
(2)
ABA

B
∗that, after considering its purification

∗Ψ ABA

B

E
γ
and performing a
local measurements in a standard product basis on itsABpart the resulting ccq
state∗
ABEis (i AB|E(ii
classical correlations between Alice and Bob.
In this work we shall briefly describe main mathematical elements of the
construction [9,10], recall the idea of one-way distillation provided in [11] and
provide a new construction of bound entangled states with secure quantum key.
2. Main Notions of General Secure Key Distillation Schemes
Here we shall remind and discuss main notions introduced in Refs. [9,10].
Assume that Alice and Bob wish to communicate but without participation
of an eavesdropper Eve. To this aim, as a source of quantum correlations, they use
a quantum state∗
ABA

B
∗. Here, subsystemAA
γ
(BB
γ
) belongs to Alice (Bob).
Moreover, following [9], subsystemAB(A
γ
B
γ
) shall be called thekey part (shield
part). To make the considerations more formal, each subsystem of∗
ABA

B
∗shall
be represented by respective Hilbert space, i.e., Alice’s subsystems byH
Aand
H
A
∗and Bob’s byH BandH B
∗. Hence, the key part of∗ ABA

B
∗is defined on
P. Horodecki and R. Augusiak / On Quantum Cryptography with Bipartite Bound Entangled States20

H=H A⊗H Band the shield part onH
γ
=H A
∗⊗H B
∗. Hereafter we shall be
assuming that dimH
A=dimH B=d,dimH A
∗=dA
∗, and dimH B
∗=dB
∗.
Firstly, let us remind the notion of ccq state. To this aim let us introduce a
product basis defined on Hilbert spaceH
A⊗H B∼C
d
⊗C
d
, i.e.,
B
AB=
ϕ


e
(A)
i
γ



e
(B)
j
γ
ρ
i,j=0,...,d−1
, (2.3)
where{


e
(A)
i
γ
}
i=0,...,d−1 and{


e
(B)
i
γ
}
i=0,...,d−1 are arbitrary bases spanning the
Hilbert spaces, respectively,H
AandH B. Of course, these bases may be chosen
to be standard and thereforeB
ABis the standard basis inC
d
⊗C
d
.Thenwe
have the following
Definition 1.(ccq state) We call the state˜∗
ABEaccq stateof initial state

ABA

B
∗with respect to the basisB ABif˜∗ABEis a result of measurement
of∗
ABE=TrA

B


∗Ψ
ABA

B

E
γσ
Ψ
ABA

B

E


in the product basisB
AB. Here

∗Ψ
ABA

B

E
γ
is a purification of∗
ABA

B
∗.
As an illustrative example let us consider a density matrixη
ABA

B
∗acting
on (C
2
)
⊗4
and given byη ABA

B
∗=p


0111
γσ
0111


+(1−p)


1000
γσ
1000


.As
one may easily verify its standard purification takes the form

∗Ψ
ABA

B

E
γ
=

p

∗01110
γ
+

1−p

∗10001
γ
. Therefore, the ccq state ofη ABA

B
∗with respect to
standard basisB
(st)
AB
≡{|iη A|jηB}
1
i,j=0
is ˜ηABE=p

∗010
γσ
010

∗+(1−p)

∗101
γσ
101

∗.
Now, one can ask when a given state is said to be secure. As an answer one
gives the following (see [10])
Definition 2.(security) We call the state∗
ABA

B
∗securewith respect to the basis
B
ABif its ccq state is of the form
˜∗
ABE=


d−1
η
i,j=0
pij

∗e
(A)
i
e
(B)
j
γσ
e
(A)
i
e
(B)
j



⎦⊗∗
E. (2.4)
The security of such a state follows from the fact that Eve is completely uncor-
related from distribution represented byABsystem after Alice and Bob mea-
surement. Note that if the distributionp
ijis homogenous, i.e.,p ij=1/d;(i, j=
0,...,d−1), then we say that∗
ABA

B
∗has aB AB-key.
A very important ingredient of construction discussed here is a special class
of controlled unitary operations (see [9,10]) that we recall by the following
Definition 3.(twistingU
(A

B

)
ij
be certain unitary operations acting on sub-
systemA
γ
B
γ
. For a given basisB ABwe call the operation
U=
d−1
η
i,j=0

∗e
(A)
i
e
(B)
j
γσ
e
(A)
i
e
(B)
j

∗⊗U
(A

B

)
ij
(2.5)
B
AB-twistingor shortly,twisting.
P. Horodecki and R. Augusiak / On Quantum Cryptography with Bipartite Bound Entangled States21

The importance of such a class of operations follows from the fact that applied
to a given state∗
ABA

B
∗,Udo notchange its ccq state. It means that if we take

ABA

B
∗andσ ABA

B
∗=U∗ ABA

B
∗U

, their ccq states are exactly the same, i.e.,
˜∗
ABE=˜σABE.
Now we define the central notion of the generalised approach provided in
[9,10]. This is the notion of private state that has logdbits of secure key encoded
in itsABpart ofd⊗dtype).
Definition 4.(pdit) Let∗
ABA

B
∗is a density operator acting on the Hilbert space
H⊗H
γ
and∗ A

B
∗is a density matrix acting onH
γ
.LetU
(A

B

)
i
(i=0,...,d−
1)be certain unitary operations acting onA
γ
B
γ
system. Then we call the state

ABA

B
∗private stateorpditwith respect to the basisB ABif it is of the form

ABA

B
∗=
1
d
d−1
η
i,j=0


e
(A)
i
e
(B)
i
γσ
e
(A)
j
e
(B)
j


⊗U
(A

B

)
i
∗A

B
∗U
(A

B

)†
j
. (2.6)
Hereafter, as usual, we shall denote the private dits byγ
(d)
ABA

B
∗. In the case when
the dimension of the key part isd= 2 on each side, we have to do withprivate
bitorpbit.
Now it is important to pose the question:What is the essential feature that
allows private bit to be truly private?Namely, a detailed analysis shows [9,10]
that ccq state of private bitγ
(d)
ABA

B
∗is the same as the ccq state of the following
state (calledbasic pdit)
P
(d)
+
⊗σA

B
∗, (2.7)
whereσ
A

B
∗=TrABγ
(d)
ABA

B
∗andP
(d)
+
is a projector onto

∗Ψ
(d)
+
γ
.Thisisbecause
it is always possible for a given pbit to find such a twisting under whichγ
(2)
ABA

B

transforms it to (2.7
basisB
ABthe physical systemABEis in the same state irrespective of whether
before Alice and Bob state was (2.7) or just a pditγ
(d)
ABA

B
∗. Hence the security
with respect to the measurement inthat particular basisisthesameasifAlice
and Bob really shared maximal entangled stateP
(d)
+
! This is the key observation
for understanding the essence of private dit.
We conclude the preliminary section recalling the definition of distillable key
[9,10] and related theorem.
Definition 5.(distillable key) Letσ
ABbe a density matrix acting onC
dA
⊗C
dB
and letP nbe a sequence of LOCC operations such thatP n(σ
⊗n
AB
)=Σn,where
Σ
nis defined onC
dn
⊗C
dn
⊗H A
∗⊗H B
∗.ThesetP={P n}

n=1
is said to be a
pdit distillation protocolof a given stateσ
ABif the following relation
lim
n→∞





∗Σ
n−γ
(dn)
ABA

B







Tr
=0. (2.8)
The rate of this protocol is defined as
P. Horodecki and R. Augusiak / On Quantum Cryptography with Bipartite Bound Entangled States22

R(P) = lim sup
n,dn→∞
logd n
n
(2.9)
and distillable key ofσ
ABas
K
D(σAB)=sup
P
R(P). (2.10)
There is a problem however, since the above definition has a complicated
form. It is hard to see whether given state fulfills the above condition or not.
Fortunately, one can simplify the task providing necessary and sufficient condi-
tions for nonzero distillable key that are more operational then the definition
itself. Here we provide a summary of the conditions proven in [9,10], which are
enough to analyse cryptographic usefulness of many quantum states:
Theorem 1.The following three conditions are equivalent:
(i ∗
ABA

B
∗can be transformed with some
LOCC protocol into the stateσ
ABA

B
∗arbitrarily close in the trace norm to certain
pbitγ
(2)
ABA

B

(ii ∗ ABA

B
∗can be transformed with some
LOCC protocol into the stateσ
ABA

B
∗with trace norm of the elementA
(A

B

)
00,11
arbitrarily close to1/2, where the element is defined by the representation:

ABA

B
∗=
1
η
i,j,k,l=0

∗e
(A)
i
e
(B)
j
γσ
e
(A)
k
e
(B)
l

∗⊗A
(A

B

)
ij,kl
. (2.11)
(iii ∗has nonzero distillable key, i.e., one hasK
D(∗)>0.
Moreover any convergence of||A
(A

B

)
00,11
||Trto1/2from (ii
protocol is equivalent to a convergence of the state to a certain pbit during that
protocol.
Especially the condition (ii
judge an applicability of a given state to quantum cryptography. In next section
we shall illustrate its power.
3. New Class of Bound Entangled States With Secure Quantum Key
In this section we present the main results. We provide a construction of a state
that is useful for quantum cryptography simultaneously being bound entangled.
Note that the construction, however based on that presented in [11], is different
in details from known so far [9,10,11] and sheds some light on the still unexplored
domain of bound entanglement. Hereafter it is assumed thatd= 2, i.e., the key
part of∗
ABA

B
∗consists of qubits and that dimH A
∗=dimH B
∗=D,which
allows us to writeH
γ
=C
D
⊗C
D
. We may also assume for simplicity thatB AB
is standard basis inC
2
⊗C
2
.
P. Horodecki and R. Augusiak / On Quantum Cryptography with Bipartite Bound Entangled States23

3.1. The Construction
To obtain a better insight into the construction we begin our considerations from
an illustrative example with the shield part of dimensionD= 3 on each side. This
with the assumption thatd= 2 makes the considered state to be of dimension
6 in each side. Finally, we show that the construction is possible for arbitrary
D≥3.
3.1.1. The6⊗6Case
At the very beginning suppose that Alice and Bob possess the following state
(cf. [11])

ABA

B
∗=
11
40






|X
3|00 X 3
0


∗X
T
B

3



T
B

00
00


∗X
T
B

3



T
B

0
X
3 00 |X 3|






(3.12)
where the superscriptT
B
∗denotes the partial transposition with respect to the
systemB
γ
andX 3is a symmetric 9×9 matrix of the form
X
3=
1
11














−1000 1 000 1
0 100 0 000 0
0 010 0 000 0
0 001 0 000 0
1 000−1000 1
0 000 0 100 0
0 000 0 010 0
0 000 0 001 0
1 000 1 000−1














. (3.13)
which is defined on the Hilbert spaceH
A
∗⊗H B
∗=C
3
⊗C
3
.Inor-
der to show that the matrix∗
ABA

B
∗is Hermitian and nonnegative, i.e.,
represents a quantum state, let us observe thatXmay be decomposed as
X
3=(1/11)(P
(3)
+
−2P
(3)
+Q
(3)
). HereP
(3)
+
is a projector onto the maximally
entangled state

∗Ψ
(3)
+
γ
=(1/

3)
λ
2
i=0

∗i
γ∗
∗i
γ
belonging toC
3
⊗C
3
andP
(3)
,and
Q
(3)
are projectors given by relations
P
(3)
=
2
η
i=0

∗ii
γσ
ii

∗−P
(3)
+
,Q
(3)
=I9−
2
η
i=0

∗ii
γσ
ii

∗, (3.14)
whereI
9stands for an identity acting on the Hilbert spaceC
3
⊗C
3
. These above
projectors are orthogonal and therefore one obtains
P. Horodecki and R. Augusiak / On Quantum Cryptography with Bipartite Bound Entangled States24

|X3|=
1
11
α
P
(3)
+
+2P
(3)
+Q
(3)
β
=
1
11














5
3
000−
1
3
000−
1
3
0 100 0 000 0
0 010 0 000 0
0 001 0 000 0

1
3
000
5
3
000−
1
3
0 000 0 100 0
0 000 0 010 0
0 000 0 001 0

1
3
000−
1
3
000
5
3














. (3.15)
From the above equation one infers two facts, first that the trace norm
1
ofX 3is
||X
3||Tr= 1 and the second that|X 3|
T
B
Δ≥0. Moreover, the matrixX 3partially
transposed with respect to the subsystemB
γ
maybewrittenintheformX
T
B
Δ
3
=
(1/11)
ν
2S
(3)

ω
I 9−Q
(3)
⊆↑
,whereS
(3)
is a projector given by
S
(3)
=
1
2
δ
I
9+V
(3)
−2
2

i=0
Δ
Δ
ii
γσ
ii
Δ
Δ
ζ
(3.16)
withV
(3)
being the swap operator defined by relationV
(3)
Δ
Δϕ
1
γΔ
Δ
ϕ
2
γ
=
Δ
Δ
ϕ
2
γΔ
Δ
ϕ
1
γ
for
Δ
Δϕ
i
γ
∈C
3
(i=1,2). Note thatI 9−Q
(3)
andS
(3)
are orthogonal pro-
jectors and therefore
Δ
Δ
ΔX
T
B
Δ
3
Δ
Δ
Δ=(1/11)
ν
2S
(3)
+I9−Q
(3)

.Since
Δ
Δ
ΔX
T
B
Δ
3
Δ
Δ
Δ
T
B
Δ
=
(1/11)
α
3P
(3)
+
+Q
(3)
β
, one may easily conclude that
Δ
Δ
ΔX
T
B
Δ
3
Δ
Δ
Δ
T
B
Δ
≥0. This with the
aid of the fact that the matrixX
3is symmetric and real, ensures thatΔ ABA
Δ
B
Δ
represents a quantum state.
Now we are in position to prove thatΔ
ABA
Δ
B
Δsatisfies PPT (positive partial
transpose) criterion [13,14]. To this aim we show that transposition with respect
to subsystemBB
γ
preserves the positivity. Indeed, the stateΔ ABA
Δ
B
Δtransposed
with respect to theBB
γ
subsystem, remains a positive operator. To see this fact
explicitly, we write
Δ
T
BB
Δ
ABA
Δ
B
Δ=
11
40






|X
3|
T
B
Δ
00 0
0
Δ
Δ
ΔX
T
B
Δ
3
Δ
Δ
ΔX
T
B
Δ
3
0
0X
T
B
Δ
3
Δ
Δ
ΔX
T
B
Δ
3
Δ
Δ
Δ0
000 |X
3|
T
B
Δ






. (3.17)
Positivity of the above operator stems from two facts. As previously mentioned
|X
3|
T
B
Δ≥0 and on the other hand the off-diagonal elements are blocked by
Δ
Δ
ΔX
T
B
Δ
3
Δ
Δ
Δ.
3.1.2. Construction of General2D⊗2DCase
Trying to generalize the above investigations we start from the symmetric matrix
1
For an arbitrary matrix Ξ the trace norm is defined by relation||Ξ|| Tr=Tr

Ξ

Ξ.
P. Horodecki and R. Augusiak / On Quantum Cryptography with Bipartite Bound Entangled States25

XD=
1
D
2
+2D−4
α
(D−2)P
(D)
+
−2P
(D)
+Q
(D)
β
, (3.18)
where, as previously,P
(D)
+
represents a projector onto maximally entangled state


Ψ
(D)
+
γ
. Orthogonal projectorsP
(D)
andQ
(D)
are given by
P
(D)
=
D−1
η
i=0

∗ii
γσ
ii

∗−P
(D)
+
,Q
(D)
=I
D
2−
D−1
η
i=0


ii
γσ
ii


. (3.19)
From Eq. (3.18) we have
|X
D|=
1
D
2
+2D−4
α
(D−2)P
(D)
+
+2P
(D)
+Q
(D)
β
(3.20)
and therefore|X
D|
T
B
∗≥0. Subsequently, after elementary steps, we may obtain
X
T
B

D
=
ν
1/(D
2
+2D−4)

2S
(D)

ω
I
D
2−Q
(D)
⊆↑
withS
(D)
being defined as
S
(D)
=
1
2
δ
I
D
2+V
(D)
−2
D−1
η
i=0

∗ii
γσ
ii


ζ
, (3.21)
whereV
(D)
is the swap operator acting onC
D
⊗C
D
. Again, the projectorsI
D
2−
Q
(D)
andS
(D)
are orthogonal and therefore


∗X
T
B

D


∗=
1
D
2
+2D−4
α
2S
(D)
+I
D
2−Q
(D)
β
. (3.22)
Finally, performing partial transposition with respect to subsystemB
γ
,wehave


∗X
T
B

D



T
B

=
ν
1/(D
2
+2D−4)

α
DP
(D)
+
+Q
(D)
β
and therefore


∗X
T
B

D



T
B

≥0.
Now we can introduce a class of mixed states

(D)
ABA

B
∗=
1
4
D
2
+2D−4
D
2
+D−2






|X
D|00 X D
0


∗X
T
B

D



T
B

00
00


∗X
T
B

D



T
B

0
X
D 00 |X D|






. (3.23)
Again, by virtue of the fact that|X
D|
T
B
∗≥0 one may infer that partial trans-
position with respect to subsystemBB
γ
preserves the positivity of∗
(D)
ABA

B
∗.In
other words, the following matrix

(D)T
BB

ABA

B
∗=
1
4
D
2
+2D−4
D
2
+D−2






|X
D|
T
B
∗ 00 0
0


∗X
T
B

D


∗X
T
B

D
0
0 X
T
B

D


∗X
T
B

D


∗0
000 |X
D|
T
B







. (3.24)
possess the nonnegative eigenvalues.
P. Horodecki and R. Augusiak / On Quantum Cryptography with Bipartite Bound Entangled States26

3.2. Proof of Nonzero Distillable KeyK D
Since the state∗
(D)
ABA

B
∗given by Eq. (3.23
separable - it satisfies necessary (PPT
it is not. This is in agreement with the fact [14] that PPT condition is sufficient
for separability only for the casesM⊗N,MN≤6 while here we haveMN=
(2D)
2
≥36. Like in [9] we shall prove nonseparability of∗
(D)
ABA

B
∗in a very
nonstandard way. We simply show that the state has nonzeroK
D. Such a state
must be entangled since, due to seminal result of Ref. [15], any separable state
hasK
D= 0. Because state is PPT and entangled it must be bound entangled [7].
To prove the cryptographic use of∗
ABA

B
∗, below we show that there exist
a LOCC protocol that allows Alice and Bob to approach arbitrarily closely to
some pbitγ
(2)
ABA

B
∗. Note that, obviously, since LOCC operations preserves PPT
property (see [7]), the resulting state is still bound entangled. Givenkcopies of
the state∗
ABA

B
∗in thei-th step of the protocol Alice and Bob perform the
following operations:
•They take the state∗
(i−1)
ABA

B
∗(i=1,...,k−1) and one of remainingk−i+1
copies of∗
ABA

B
∗(here∗
(0)
ABA

B
∗=∗ABA

B
∗).
•They perform C-NOT operation treating qubitsAandBof∗
ABA

B
∗as
source qubits and that of∗
(i−1)
ABA

B
∗as target qubits.
•They perform a measurement of target qubits in computational basis and
then compare their results. If both of them have the same results (00 or
11) then they keep the source state. Otherwise they get rid of it.
After performing allksteps, with some probability they arrive at the following
state

(D,k)
ABA

B
∗=
1
ND,k









|X
D|
⊗k
00 X
⊗k
D
0
θ


∗X
T
B

D



T
B

τ
⊗k
00
00
θ


∗X
T
B

D



T
B

τ
⊗k
0
X
⊗k
D
00 |X D|
⊗k









,(3.25)
where (forD≥3)
N
D,k=2Tr|X D|
⊗k
+2Tr
θ


∗X
T
B

D



T
B

τ
⊗k
=2
δ
1+
θ
D
2
D
2
+2D−4
τ
k
ζ
k→∞
−−−→2.
(3.26)
If we define, according to (2.11) the matrix,A
(A

B

)
00,11
(D, k)=(1/N D,k)X
⊗k
D
we
can see that||A
(A

B

)
00,11
(D, k)|| Tr→1/2 wheneverk→∞,forD≥3. This means
that repeating the whole procedure described above, one may find such akthat
the trace norm of the upper right block of∗
(D,k)
ABA

B
∗is close to 1/2 with arbi-
trary precision. According to the the Theorem 1 (see Section 2) this convergence
guarantees that the original state∗
(D)
ABA

B
∗defined by the formula (3.23) satisfies
P. Horodecki and R. Augusiak / On Quantum Cryptography with Bipartite Bound Entangled States27

KD>0. The Theorem 2 guarantees in particular, that (like it was in [9]) the
above sequence of bound entangled statesΔ
(D,k)
ABA
Δ
B
Δapproaches private bit.
4. Summary and Discussion
We have summarized main elements of general scheme of distillation of secure
key [9,10]. The central notion of the scheme is the idea of private bit (with its
natural generalization - private dit) which is the state that, in general, consists
of two parts: the key partABand the shield partA
γ
B
γ
. The first contains a bit
of secure key. The role of the second part is - in a sense - to protect the key
form Eve. A surprising fact, found in [9] is that PPT bound entangled states can
approach private bit in trace norm. Since this convergence is a necessary and
sufficient condition to distill secure key form the original state, this means that
bound entangled states can serve as a source of distillable key [9,10].
First bound entangled states with nonzero distillable keyK
Dwere provided
in [9]. They required however very high dimensions. The small (4⊗4) bound
entangled states with nonzero distillable key were provided later in paper [11].
Here we have provided new class of small (of, among others, 6⊗6type)
bound entangled states with that property. We have proven this fact applying
an easy criterion form [9] showing that a given sequence of quantum states
approaches the sequence of private bits. The LOCC protocol applied to produce
such a sequence was of two-way type. As noticed in [9] at some point the elements
of the sequence got one-way distillable key which can be distilled with help of
Devetak-Winter protocol. The original problem was that such states were of
very high dimensions. Quite surprisingly the low-dimensional bound entangled
states provided in [11] represent one-way distillable key. It was proven with help
of the observation that any biased mixture of two private bits with second of
them having the key part rotated locally withσ
xPauli matrix contains nonzero
distillable key. Namely one has the following theorem [11]:
Theorem 2.For two private bitsγ
(2)
1

(2)
2
one-way distillable key of the mixture
of the form:
ρ=p

(2)
1
+p2σ
(A)
x
γ
(2)
2
σ
(A)
x
(4.27)
with,σ
(A)
x
=[σ x]A⊗IA
Δ
BB
ΔsatisfiesK

D
(ρ)≥1−h(p 1)with binary entropy
h(p
1)
2
.
HereK

D
(ρ) stands for cryptographic key distillable fromρwith help of forward
classical communication. The natural question is whether one can modify the
6⊗6 bound entangled states provided in the present paper to get mixture of two
private bits of the above form. Our first analysis has shown that most probably
it is impossible to turn our example into a state of the form (4.27
ing bound entanglement property. In Ref. [11] the construction leading tobound
entangledstate of the form (4.27
2
Binary entropy of the distribution{p 1,p2}is defined ash(p 1)=−p 1log2p1−p2log2p2
P. Horodecki and R. Augusiak / On Quantum Cryptography with Bipartite Bound Entangled States28

were used in locking entanglement measures effects. May be that was the reason
why the construction was successful there. Still there is a natural question about
other methods to construct low-dimensional bound entangled states with one-way
distillable key.
Of course the most important open problem is whether any entangled bipar-
tite state contains nonzero distillable key or not. In multipartite case it is not
true - there are states that are entangled but no secure key between any of the
parties can be distilled [16]. In bipartite case lack of such states is guaranteed (via
entanglement distillation approach [4]) only for 2⊗2and2⊗3 cases, since, as
shown in [17,18], all the entangled states can be distilled to singlet form in those
cases. Ford⊗dwithd≥4 it is known that at least some bound entangled (ie.
nondistillable to singlets) states have nonzero distillable key [11]. No example of
3⊗3or2⊗4 states with that property is still known.
Acknowledgements
The authors thank Maciej Demianowicz and Micha√l Horodecki for fruitful discus-
sions. The work is partially supported by Polish Ministry of Scientific Research
and Information Technology grant under the (solicited
008/P03/2003 and by EC grant: RESQ, contract no. IST-2001-37559.
References
[1] A. Ekert, Phys. Rev. Lett.67, 661 (1991
[2] C. H. Bennett, G. Brassard, and N. D. Mermin, Phys. Rev. Lett.68, 557 (1992
[3]B.M.Terhal,IBMJ.Res.Dev. 48, 71 (2004).
[4] D.Deutsch,A.Ekert,R.Jozsa,C.Macchiavello,S.Popescu,andA.Sanpera,Phys.
Rev. Lett.77, 2818 (1996
[5] C. H. Bennett, G. Brassard, S. Popescu, B. Schumacher, J. Smolin, and W. K.
Wootters, Phys. Rev. Lett.76, 722 (1996
[6] C. H. Bennett, D. P. Di Vincenzo, J. Smolin, and W. K. Wootters, Phys. Rev. A
54, 3814 (1997
[7] M. Horodecki, P. Horodecki, and R. Horodecki, Phys. Rev. Lett.80, 5239 (1998
[8] N. Gisin and M. Wolf, Phys. Rev. Lett.83, 4200 (1999
N. Gisin, quant-ph/0303053.
[9] K. Horodecki, M. Horodecki, P. Horodecki, and J. Oppenheim, Phys. Rev. Lett.94,
160502 (2005
[10] K. Horodecki, M. Horodecki, P. Horodecki, and J. Oppenheim, quant-ph/0506189.
[11] K. Horodecki, √L. Pankowski, M. Horodecki, and P. Horodecki, quant-ph/0506203.
[12] I. Devetak and A. Winter, Proc. R. Soc. A461, 207 (2005
[13] A. Peres, Phys. Rev. Lett.77, 1413 (1996
[14] M. Horodecki, P. Horodecki, and R. Horodecki, Phys. Lett. A223, 1 (1996
[15] M. Curty, M. Lewenstein, and N. L¨utkenhaus, Phys. Rev. Lett.92, 217903 (2004
[16] R. Augusiak and P. Horodecki, quant-ph/0405187; R. Augusiak and P. Horodecki,
Phys. Rev. A73, 012318 (2005
[17] M. Horodecki, P. Horodecki, and R. Horodecki, Phys. Rev. Lett.78, 574 (1997
[18] M. Horodecki and P. Horodecki, Phys. Rev. A59, 4026 (1999
P. Horodecki and R. Augusiak / On Quantum Cryptography with Bipartite Bound Entangled States29

Unitary Local Permutations on Bell
Diagonal States of Qudits and Quantum
Distillation Protocols
Hector Bombin, Miguel A. Martin-Delgado
Departamento de F´ısica Te´orica I, Universidad Complutense, 28040. Madrid,
Spain
Abstract.Entanglement distillation protocols are generalized for qudits
of arbitrary dimension by studying the group of unitary local permuta-
tions. We introduce the concept ofjoint performance parameterηthat
allows us the comparison of distillation protocols with different values
of fidelity, probability of success and number of Bell pairs used alto-
gether.We analyze several distillation protocols assisted with twirling
operations as the dimensionDof qudits vary. We find that the best
performance according toηis not achieved for qubits (D= 2), but
for qutrits (D=3)andn= 3 input pairs of Bell states. We propose
and study an extension of the Quantum Privacy Amplification protocols
that work for arbitrary dimensionD, being more efficient whenDis a
prime number.
Keywords.Entanglement, Distillation, Qudits, Quantum Privacy Amplification
1. Introduction
We show that the analysis of entanglement distillation protocols [1], [2] for qudits
of arbitrary dimensionDbenefits from applying basic concepts from number the-
ory [3]. This is so because the set modulusZ
D:=Z/DZcan be identified with
theelementsofabasisofC
D
. We characterize the group of local permutations
over tensor products of Bell states as an affine symplectic matrix group. Then we
find that a partition ofZ
n
D
into divisor classes is related to the invariant proper-
ties of mixed Bell diagonal states under this group. We construct a very general
class of recursion protocols by means of unitary operations implementing these
local permutations. We analyze these distillation protocols depending on whether
we use twirling operations in the intermediate steps or not, and we study them
both analytically and numerically with Monte Carlo methods. When twirling is
present, we are considering generalizations of the original distillation protocol. In
the absence of twirling operations, we construct extensions of quantum privacy
amplification algorithms valid for secure communications with qudits of any di-
mensionD.WhenDis a prime number, we show that distillation protocols are
optimal both qualitatively and quantitatively.
Quantum Information Processing
D.G. Angelakis et al. (Eds.)
IOS Press, 2006
© 2006 IOS Press. All rights reserved.
30

Figure 1.Alice and Bob sharenBell pairs of qudits. It is shown the pattern of indices in the
tensor product of pairs.
2. Bell States for Qudits and the ModuleZ
n
D
Why qudits after all? They are more resistant to noise than qubits. In quan-
tum cryptography, they are more secure against eavesdropping attacks. Moreover,
they violate more strongly local reality. A qudit is any quantum system of finite
dimensionD. The elements of a given orthogonal basis can be denoted|xηwith
x=0,...,D−1. This set of numbers is naturally identified with the elements of
the set modulus
Z
D:=Z/DZ. (1)
Z
n
D
is a module and not a vector space. This is becauseZ Dis a field only for D
prime. For example, for D = 4 we have:
2·(0,2) = (0,0). (2)
The computational basis for n shared pairs and Bell (product) basis is [3]
|ijη:=
n

k=1
|ikjkη,
|ijη
B:=
n

k=1
|ikjkη
B=D

n
2
η
k∈Z
n
D
ϕ(i·k)|kk−jη, (3)
H. Bombin and M.A. Martin-Delgado / Unitary Local Permutations on BDS of Qudits31

Figure 2.Schematic representation of the relationship among the unitary groups of local oper-
ations and the group of permutations in theorem 3.1.
whereϕ(x):=exp(
2πi
D
x)andi,j∈Z
n
D
. Bell diagonal states (BDS) are mixed
states of the form
ρ
(n)
=
η
x∈Z
2n
D
p
(n)
x|xη
Bx|. (4)
Fig. 1 pictures the notation of shared pairs by Alice and Bob.
In order to classify the anomalous vectors, consider the set of divisors of D:
div(D):={d∈N:d|D}. (5)
This set inherits the ordering ofN, and we shall use this property to introduce a
suitable gcd function inZ
D:
Definition 2.1For everyS⊂Z
Dwe define the greatest common divisor ofS,or
gcd(S),tobethegreatestd∈div(D)such that
D
d
s=0,∀s∈S.
The nomenclature was chosen because for anyd∈div(D)andx∈Zwe have
d|x⇐⇒D|
D
d
x⇐⇒
D
d
x=0 (modD), (6)
and then for any set of integersX
gcd(X)=max{d∈div(D):d|x∀x∈X}, (7)
whereXis the corresponding set inZ D.
Vectors overZ
Dare n-tuples of elements inZ D, and so we extend the gcd
function to act overZ
n
D
in the natural way, that is, ifv=(v 1,... vn), gcd(v):=
H. Bombin and M.A. Martin-Delgado / Unitary Local Permutations on BDS of Qudits32

Figure 3.This is a representation of a generic heterotropic state (25D= 6 (single pair
of qudits). Each cell represents the probabilityp
ijof a (Bell) diagonal element (4
given by the row and the column of the cell. The cells with the same colour have the same
probability. There is one colour for each of the divisors of 6: 1 is associated to white, 2 to blue,
3toyellowand6tored..
gcd({v 1,...,vn}). Now we can consider an equivalence relation inZ
n
D
governed
by the equality under the gcd function. The corresponding partition consists in
the sets:
C
d(D, n):={v∈Z
n
D
:gcd(v)=d},d∈div(D)(8)
The most important of these sets isC
1(D, n), since it contains those vectors
vfor with{v}is linearly independent. Later we will need its cardinality when
considering properties of local unitary operators acting on diagonal Bell states.
Thus, it is useful to define:
φ
n(D):=

1if D=1
C
1(D, n)ifD>1
(9)
For the particular case ofn=1,φ
1(x) corresponds to Euler’s totientφ-function
[4]. Euler’sφ-function appears naturally in number theory since it gives for a nat-
ural numbern, the cardinality of the set{m=1,...,n−1:gcd(m, n)=1}.That
H. Bombin and M.A. Martin-Delgado / Unitary Local Permutations on BDS of Qudits33

is,φ1(n) is the total number of coprime integers (or totatives) below or equal to
n. For example, there are eight totatives of 24, namely,{1,5,7,11,13,17,19,23},
thusφ
1(24) = 8. Fornν = 1, we have therefore introduced a generalization of
Euler’s totient function for elements inZ
n
D
. The following lemma gives us how to
compute the cardinalities of the setsC
d(D, n), which shall naturally arise in our
analysis of distillation protocols.
Lemma 2.2For everyn∈N,D∈N−{1}andd∈div(D):
1.
φ
n(D)=D
n
ϑ
p|D
pprime
p
n
−1
p
n
(10)
2.
C
d(D, n)=φ n(
D
d
) (11)
3.
η
d
Γ
∈div(D)
φn(d
γ
)=D
n
(12)
The proof of this lemma can be found in [3].
3. Unitary Local Permutations on Bell Diagonal States
The main constraint Alice and Bob have to face when they intend to distill qu-
dits is that they can perform only local operations. If we consider only unitary
operations, we are led to the groupU
locof local unitary operations. Its elements
are all of the form
U=U
A⊗UB. (13)
We study the subgroupU
Bloc, defined as the group of local unitary operations
which are closed over the space of Bell diagonal states (4).
Since the mapping of Bell states must be one to one, the action of anyU∈
U
Bloc(D, n) involves a permutation of the Bell states:
UρU

=
η
x∈Z
2n
D
p
(n)
x|π(x)η
Bπ(x)|, (14)
whereπ:Z
2n
D
→Z
2n
D
is a permutation. So we introduceP
loc(D, n), thegroup of
local permutations, as the set of permutations overZ
2n
D
implementable over Bell
states by local (unitary) means.
We need to define several groups. Consider the family of unitary operators
u
x(x∈Z
2n
D
) over Bob’s part of the system such that by definition
H. Bombin and M.A. Martin-Delgado / Unitary Local Permutations on BDS of Qudits34

Figure 4.This is a representation of the two permutations generating the groupP Sfor a pair
of qudits. Each cell represents an element ofZ
2
D
and the arrows shows the action of the permu-
tation.
1⊗u

x
|0η
B:=|xη
B, (15)
where conjugation is taken respect to the computational basis. With this operators
at hand, we construct the groupU
Binv(D, n) with the elementsU x:=u x⊗ux
∗.
We claim that it is a subgroup ofU
Bloc(D, n). An explicit calculation shows that
the action of its elements is
U
x|yη
B=ϕ(x
t
Ωy)|yη
B, (16)
where Ω∈M
2n×2n (ZD)is
Ω:=

01
−10

. (17)
So the special feature ofU
Binvis that forρ
(n)
Bell diagonalU xρ
(n)
U

x

(n)
,which
means that its elements implement the identity permutation.
We also define two subgroups of the group of permutations overZ
2n
D
.The
translation groupP
T(D, n) contains the permutations of the form
π
a(x)=x+a, (18)
witha∈Z
2n
D
,andthesymplectic groupP
S(D, n)containsinturnthosewhose
action is
π
M(x)=Mx, (19)
H. Bombin and M.A. Martin-Delgado / Unitary Local Permutations on BDS of Qudits35

whereM∈M 2n×2n (ZD) is such that
M
t
ΩM=Ω. (20)
In Fig. 4 we picture the action of some of its generators.P
S(D, n) is a finite
non-simple group. A suitable generator set for this group is presented in [3].
Now, we can stablish our main result that plays an important role in the
distillation protocols for qudits to be devised later on.
Theorem 3.1
1.P
locis the semidirect product ofP
SandP
T:
P
loc(D, n)=P
T(D, n)ΓP
S(D, n) (21)
2.Lethbe the natural homomorphism fromU
BlocontoP
loc, then its kernel
is
kerh=U
Binv(D, n)⊗U(1), (22)
whereU(1)denotes the global phase.
We prove this theorem in [3]. This theorem is summarized in Fig. 2.
4. Invariant States under Local Permutations: Heterotropic States
The twirling operation or depolarizing channel is defined for finite groups as
follows:
T
U(ρ):=
1
U
η
U∈U
UρU

. (23)
The groupP
Scan also be successfully used in twirling operations. ForρBell
diagonal states, we have:
T
PS(ρ):=
1
P
S
η
π∈PS
U(π)ρU(π)

. (24)
To perform the sum we need to know which are the states invariant under the
action of the group.
Theorem 4.1For everyx,y∈Z
2n
D
,gcd(x)=gcd(y)if and only if there exists a
permutation inP
S(D, n)with associated matrixMsuch thatx=My.
We prove this theorem in [3].
Now, let us recall the partition inZ
2n
D
associated to the function gcd (see (8)).
We define the related states as heterotropic states (see Fig. 3)
H. Bombin and M.A. Martin-Delgado / Unitary Local Permutations on BDS of Qudits36

ρ
d:=
1
Cd
η
x∈C d
|xη
Bx|. (25)
These are the invariant states we were searching for. Thus, ifρis Bell diagonal:
T
PS(ρ)=
η
d∈div(D)
Tr(ρ
dρ)
Tr(ρ

d)
ρ
d (26)
As a corollary, ifDis prime there are just two Bell diagonal invariant states
ρ
1=
1
D
2n
−1
ω
1−|0η
B0|
ξ
, (27)
ρ
D=|0η
B0|, (28)
and thus the result of the twirling operation is an isotropic state, which is the
simplest example of an heterotropic state.
5. Permutation Based Distillation
In the distillation protocols we consider, which are iterative, each iteration cycle
can be decomposed in the following steps:
1.At start, Alice and Bob sharenqudit pairs of dimensionDand state matrix
ρ
(n)
.
2.They apply by local means one of the permutationsπ
M∈P
S(D, n)in(19).
3.They measure the lastn−mqudit pairs, both of them in their computational
basis.
4.If the results of the measurement agree for each of the measured pairs, they
keep the firstmpairs (in the stateρ
(m)
). Else, they discard them.
Fig. 7 shows a representation of the iteration cycle of a generic distillation proto-
col.
In most situations, the initialnpairs are independent and have equal state
matricesρ. In these cases
ρ
(n)

⊗n
. (29)
In general (form>1) this does not guarantee thatρ
(m)
will be a product state,
however, and thus it is preferable to consider the most general case.
Primes are special due to the appearance of one undesired attractor in the
distillation process for each of the proper divisors ofD. These attractors are
related to invariant states underP
loc(D, n).
Protocols withm>1 are useless for low fidelities, but useful in other domains
[3].
The recursion relations associated to these distillation protocols can be com-
puted analitically and do not depend on non diagonal elements:
H. Bombin and M.A. Martin-Delgado / Unitary Local Permutations on BDS of Qudits37

2 4 6 8 10 12 14 16 18 20
0
0.005
0.01
n = 2
n = 4
n = 3
D
η
Figure 5.Values of the coefficientηfor the considered twirled-assisted protocols withn=2,3,4.
Initial fidelity is close to
1
D
.
0.4 0.5 0.6 0.7 0.8 0.9 1
10
−12
10
−10
10
−8
10
−6
10
−4
10
−2
10
0
F
ϒ
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1
0
0.2
0.4
0.6
0.8
1
F
V 2
5
3
4
6
Figure 6.Left: Comparison of the yield respect to the twirling based protocol (D= 3). Right:
Normalized volume of distilled Bell diagonal states. Non prime numbers behave much worse.
If we callV
Mthe space generated by the lastn−mrows ofM(the matrix
associated toπ
M) the probability of obtaining the desired measure is
P=
δ
x∈V

M
p
(n)
x, (30)
and the recurrence relation for the probabilities is
p
(m)
x
=
1
P
δ
y∈V M
p
(n)
Ωy+M
−1
¯x
, (31)
wherex∈Z
2m
D
and¯x∈Z
2n
D
is
H. Bombin and M.A. Martin-Delgado / Unitary Local Permutations on BDS of Qudits38

¯x:= (x 1,...,xm,0,...,0
˜ !
n−m
,xn+1,...,xn+m,0,...,0
˜ !
n−m
). (32)
We must distinguish between twirling assisted protocols and protocols without
twirling.
For twirling assisted protocols, before each step, states are made isotropic
using a suitable twirling operation. This reduces greatly the number of parameters
describing the protocol. Low fidelity states cannot be distilled using hashing [5].
They deserve special attention. It is posible to define the joint performanceη,
which allows a direct comparison of protocols. This is shown in Fig. 5 [3].
For protocols without twirling, the Quantum Privacy Amplification algo-
rithm [2] can be generalized to qudits by carefully choosing the permutation
π∈P
loc(D, n) [3]. We get the following recursion relation:
p
γ
ij
=
1
P
η
k∈Z D
pi+k,−i−j−k pk,j−k, (33)
P=
1
D
η
˜
k∈Z
D

k
˜
k
2. (34)
Numerical results are shwon in Fig. 6 [3].
References
[1] C.H. Bennett, G. Brassard, S. Popescu, B. Schumacher, J.A. Smolin, W.K. Woot-
ters, “Purification of Noisy Entanglement and Faithful Teleportation via Noisy
Channels”,Phys.Rev.Lett.76, 722-725 (1996).
[2]D.Deutsch,A.Ekert,R.Jozsa,C.Macchiavello,S.Popescu,A.Sanpera,“Quan-
tum Privacy Amplification and the Security of Quantum Cryptography over Noisy
Channels”,Phys.Rev.Lett.77, 2818-2821 (1996).
[3] H. Bombin, M.A. Martin-Delgado,
“Entanglement Distillation Protocols and Number Theory”; quant-ph/0503013.
Phys. Rev.Aaccepted.
[4] J.H. Conway, R.K. Guy, “Euler’s Totient Numbers.” inTheBookofNumbers.New
York: Springer-Verlag, pp. 154-156, 1996.
[5] K. G. H. Vollbrecht and M. M. Wolf, “Efficient distillation beyond qubits”, Phys.
Rev. A67, 012303 (2003)
H. Bombin and M.A. Martin-Delgado / Unitary Local Permutations on BDS of Qudits39

Figure 7.Pictorial represenation of the iteration cycle of a generic distillation protocol based on
the local permutations of the groupP
loc(D, n).
H. Bombin and M.A. Martin-Delgado / Unitary Local Permutations on BDS of Qudits40

Exploring the Variety of Random
Documents with Different Content

en toch wordt ook daar de stuwing gevoeld. Maar poëzie, muziek,
beeldende- en bouwkunst van de laatste tientallen jaren spreken
overal van herleving van den zin voor het religieuze.
Dat zijn teekenen die ons moeten leeren, dat en in welken zin de
godsdienst in de 20e eeuw nog iets voor ons is of weder iets voor
ons is geworden. Met het oog daarop is het zeker niet misplaatst in
een boekverzameling als de Wereldbibliotheek, die toch een beeld
van het geestesleven onzes tijds wil zijn, een geschrift op te nemen,
dat onze aandacht weder bij den godsdienst en de godsdiensten
bepaalt. Menigeen wil van den godsdienst niet weten omdat hij er
nu eenmaal niet van weet. Anderen ijveren in den politieken
partijstrijd voor hunne godsdienstige beginselen, zonder zich
voldoende rekenschap te hebben gegeven van de vragen die alles
beheerschen en den waren aard der tegenover elkander staande
ideeën. En niets is bedriegelijker dan de leuzen, die in het
strijdgewoel worden gehoord. Toen de schoolstrijd hier te lande het
heftigst woedde, werd de leus der antithese: Christendom of
Paganisme gehoord. Zij verborg echter de ware tegenstelling, die
gelegen was in de verschillende wijze, waarop men zich de
verhouding tusschen godsdienst en zedelijkheid dacht, en het
andere even diepgaand verschil in de denkbeelden over de
verhouding tusschen gelooven en weten. Midden in het partijgewoel
komt men er niet toe zulke vragen rustig onder de oogen te zien.
Maar nu maken wij ons van al die verwarrende practische kwesties
los en willen trachten onder voorlichting van wetenschappelijk
denken kalm en klaar, sine ira et studio juist die algemeene vragen
eens te beschouwen, wier belang voor ieder denkend mensch
evenredig is aan haar algemeen belang.
Maar nu rijst dan de vraag of zulk een rustig onbevooroordeeld
onderzoek juist in deze vragen en op dit gebied wel mogelijk is. Wij
behoeven ons niet te ontveinzen dat in onzen omgang in het
dagelijksch leven ieder gesprek over dingen den godsdienst
betreffend, stilzwijgend verboden en vrijwel uitgesloten is. Dat is,
denk ik, een van die preventieve maatregelen der „goede zeden”, die

zoowel goede als slechte gronden heeft. Eenerzijds vreest men de
latente meeningsverschillen en den strijd daarover; anderzijds is
men beangst aan iets heiligs te raken: zoo kunnen wij het wellicht
formuleeren. Maar zou het daarom dan ook niet fijner en
beschaafder staan en tevens voorzichtiger zijn, met allen eerbied dit
onderwerp maar te laten rusten? Het wil mij voorkomen dat men
tegenwoordig al te veel menschen en al te veel dingen „beschaafd”
noemt. Dit soort angstvallige „beschaafdheid” is vaak niet veel meer
dan verlegenheid; en in elk geval, met haar komt de wereld niet
vooruit. Evenals voorzichtigheid maar al te vaak in lafheid ontaardt,
zoo blijkt tegenover het leven zulk een houding vaak een masker
voor hulpeloosheid en onbekwaamheid.
Wij moeten, dunkt mij, wel erkennen, als wij eerlijk willen zijn, dat
dit angstvallig zwijgen over religieuze vragen in de zoogenaamd
beschaafde kringen slechts een soort verstoppertje spelen is,
misschien in de eerste plaats een verstoppertje spelen met en voor
zich zelf: men schaamt zich om te zeggen wat men denkt, en te
toonen hoe weinig men denkt, en daarom zwijgt men maar liever
geheel. Maar in de tweede plaats: wie of wat dwingt ons dan om
over dergelijke vragen te spreken zonder piëteit? Ja, er is een
schijnpiëteit, die het oude vereert alleen omdat het oud is, zooals
het bekende woord van Schiller zegt: „Was grau vor Alter ist, das ist
ihm göttlich”; maar zulk soort van respect voor „wat altijd bestaan
heeft en altijd terugkomt, voor wat morgen waarde moet hebben
omdat het heden waarde heeft” is der wetenschap onbekend. Wie
dat van mij mocht verlangen, hem zou ik niet kunnen tevreden
stellen. Evenals zich het blauwe hemelgewelf met zijn gouden
sterren aan den astronoom, die toch niet genoeg heeft aan de
kinderlijk vrome gedachte van het lied „hoe lieflijk praalt gij
morgenster”, anders voordoet dan aan ’t ongewapend oog van een
onkundigen beschouwer, zoo zullen ook de godsdienstige dingen zich
anders voordoen bij wetenschappelijk onderzoek dan in de kerk en
onder de preek. Maar gelijk de hemel schoon en verheven blijft ook
voor hem, voor wien hij zich oplost in een rijk van wetmatig zich
bewegende werelden en zonnenstelsels, zoo blijft de godsdienst

heilig, ook als men hem erkent als een menschelijk verschijnsel en
de motieven waaruit hij is voortgekomen, de veranderingen die hij
heeft ondergaan en de psychische krachten die in hem werken,
tracht te leeren kennen. Van profaneeren kan slechts dan sprake
zijn, als over het heilige oneerbiedig, zonder achting en piëteit,
zonder ernst en waardigheid wordt gesproken. En dat is het dan
toch ook eigenlijk wat wij in onze gesprekken en in gewoon
gezelschap vreezen en terecht vreezen: dat over godsdienstige
dingen oneerbiedig gesproken, gelachen, gespot wordt. Oneerbiedig,
dat is lichtvaardig en oppervlakkig; en juist de toon in onze
zoogenaamd beschaafde kringen is dikwijls maar al te lichtvaardig
en oppervlakkig. De wetenschap echter neemt het met de dingen
niet licht, maar ernstig en streng, blijft niet aan de oppervlakte,
maar tracht in de diepte door te dringen. En daarom is reeds het
wetenschappelijk doel van dit boekje een waarborg tegen al wat plat
is of kwetsen kan. Zoo is dan ook zeker de persoonlijke verzekering
overbodig, dat ik ernstig en nauwgezet daarbij te werk wil gaan.
Nu moeten wij echter, voor wij ons eerste onderwerp, de vraag naar
het wezen van den godsdienst, nader gaan bezien, eerst overwegen
hoe wij daarbij te werk zullen gaan, hoe wij het geheim der religie
kunnen ontsluieren. Welnu, daarbij dringt zich in onzen tijd, die
zoozeer staat in het teeken der natuurwetenschappen, als vanzelf
het antwoord naar voren: hiertoe is er slechts eén weg, die der
inductieve methode, d. w. z. men zoekt naar die kenteekenen die
alle godsdiensten in verleden en heden gemeen hebben, dan vinden
wij uit de vele en veelvuldige verschijnselen het eene voor allen
geldende gemeenschappelijke begrip. Daarbij moet dan voor alles de
nadruk gelegd worden op die godsdiensten, die ’t dichtst staan bij
het begin der religieuze ontwikkeling, dat is op den godsdienst in zijn
meest oorspronkelijke gedaante. Men zoeke alzoo bij alle
godsdiensten en vooral bij de eenvoudigste en oudste.
Stellig is dit een weg; maar hij is lang en onzeker. Want nu moeten
eerst alle de vele, de zeer vele godsdiensten doorloopen en
onderzocht, en op het al of niet onderling gelijke gelet worden. Maar

ik vrees dat dit reeds verre de kracht van eén mensch te boven gaat,
en in elk geval gaat het ver uit boven ons hedendaagsch weten. En
er is een nog veel grooter moeilijkheid: waar moeten wij dat
primitieve, waar het eenvoudige en oorspronkelijke vinden? Men
wijst op de onontwikkelde volksstammen in Centraal Afrika, op de
Groenlanders en Patagoniërs, de Polynesiërs en Melanesiërs: maar
wie staat ons borg dat de godsdienst van deze wilden niet veeleer
een pas ontstane in plaats van een oude is, een laat tot ontwikkeling
gekomene in plaats van een oorspronkelijke, een verminkte in plaats
van een die het wezenlijke tot uiting brengt? En bovendien is bij den
nauwen samenhang van mythologie en taal de kennis der
godsdiensten van deze onbeschaafde volkeren afhankelijk van de
nog vaak zeer onvolkomen kennis van taal en dialect, zoodat de
eene duisternis de andere moet helpen verdrijven. Of wel men
verwijst ons terecht naar de oudste sporen van den godsdienst der
historische volken en de berichten daarover. Maar juist van dit oude
resten ons nog slechts enkele fragmenten en spaarzame
overblijfselen en hun verklaring reeds geeft groote moeilijkheden. En
eindelijk, om dan het wezenlijke uit die gansche lange rij te vinden,
moet ik toch altijd al van tevoren weten wat belangrijk en wat
bijkomstig is, moet ik derhalve het wezen van den godsdienst reeds
ten naastenbij kennen, om het in de geschiedenis der godsdiensten
te kunnen terugvinden. De godsdienst is toch stellig in de eerste
plaats iets dat leeft en beleefd wordt in ’s menschen innerlijk,
derhalve een psychologisch gegeven. Nu roept weliswaar ook bijv.
de psychologie der massa de geschiedenis te hulp, maar toch steeds
slechts als iets secundairs. De wetten van ’s menschen zieleleven
vertoonen zich voor het oog van den onderzoeker juist in de
geschiedenis om zoo te zeggen slechts in groote trekken, en men zal
zich veel vaker gedwongen zien, om de geschiedenis psychologisch
te verklaren, dan men in staat zal zijn uit de geschiedenis een
psychologie te construeeren.
[2]
Derhalve zoowel psychologie als zelfwaarneming? Zeker, ook dat.
Maar, zoo komt dan weer de vraag, wie waarborgt ons, dat wij
daarbij niet persoonlijke en toevallige ervaringen van bijzonderen

aard met het wezen der zaak verwisselen, en dat wij op die wijze
wel ervaren wat mijn en uw en zijn godsdienst, maar niet wat de
godsdienst zelf is?
En zoo blijft dan nog slechts een derde weg open. Wij zoeken den
godsdienst daar waar hij ons het best bekend is, bij ons zelven;
maar niet bij ons als afzonderlijke meer of minder religieus
aangelegde menschen, maar in dien vorm en gestalte, waarin wij
hem in onze omgeving, in ons volksleven en op onze hoogte van
beschaving sinds de dagen onzer kindsheid als een objectief
verschijnsel met eigen invloed en macht hebben waargenomen en
dan ook subjectief en individueel onze verhouding tegenover hem
hebben vastgesteld en ons met hem hebben verstaan. Daarbij komt
het niet uitsluitend op het eigen religieuze leven van den
onderzoeker aan: daarvan moeten wij ons blijkens het voorafgaande
vrij en onafhankelijk maken. Naast het eigen beleven staat hier het
„hypothetische nabeleven” van datgene, wat anderen om ons heen
beleefd hebben en waarin wij ons met behulp van phantasie en
analogie zeer goed kunnen verplaatsen. Want het gaat hier om
persoonlijke ervaring op den bodem van een godsdienst, die ons
historisch geheel bekend en psychologisch geheel kenbaar en
doorzichtig is, die niet primitief en oorspronkelijk en toch in ontstaan
en wording te begrijpen is, en die een hoogtepunt van ontwikkeling
te zien geeft, waarop zich de godsdienst naar alle mogelijke zijden
ontvouwt en zijn leven ten toon spreidt. Men noemt dat wel het
„normbegrip” van den godsdienst. Dat is mij te neo-kantiaansch en
te theologisch, is mij ook te zeer vooropgestelde theorie.
[3]
Niet om
normaal of abnormaal, om juist of onjuist gaat het hier; maar wat
wij op deze wijze zoeken en vinden, dat is het karakteristiek
religieuze in zijn voor ons meest verstaanbaren vorm.
[4]
Zoo vermijden wij twee uitersten. Eenerzijds het overbrengen van
een zuiver natuurwetenschappelijke beschouwingswijze op een
gebied van geesteswetenschappen: daar geldt de inductieve
methode om tot natuurwetten en algemeene begrippen te komen,
hier gaat het om analyse, om psychologisch ontleden, om het

nasporen van geestelijke verschijnselen en processen van den wortel
af, om het inzicht hoe de mensch daartoe gekomen is en welken zin
en beteekenis een dergelijk zielkundig gebeuren voor hem heeft. En
aan den anderen kant vermijden wij de theologische vóor-
onderstelling alsof in het Christendom zonder meer reeds aanstonds,
vóor alle onderzoek, de norm en maatstaf gegeven waren, waaraan
alle andere godsdiensten moeten worden afgemeten om dan als
valsch of minderwaardig te worden afgewezen. Dit zou wellicht het
resultaat van onze onderzoekingen kunnen blijken te zijn, hoewel ik
niet geloof dat wij tot een dergelijke uitkomst zullen geraken; maar
hier mag het in geen geval reeds van te voren willekeurig worden
aangenomen.
[5]
Tevens houden wij hiermede echter den
geschiedkundigen weg voor ons open, welks waarde wij niet willen
ontkennen; want de godsdienst is een historisch verschijnsel en
slechts als zoodanig te verstaan. Wij zullen derhalve dan ook dezen
weg niet versmaden, zullen ons mede naar hun historischen
ontwikkelingsgang een oordeel over de verschillende godsdiensten
moeten vormen, dus niet van meet af aan, daar waar het de vraag
geldt naar het wezen van allen godsdienst, maar eerst waar het gaat
om de verschillende vormen, om de belichaming van godsdienst in
geloofsvoorstellingen, cultus en leven.
Welnu dan—wat is godsdienst? Het woord religie kan het ons niet
duidelijk maken; want of wij het van religare afleiden en daaronder
het gebonden en verplicht zijn verstaan, of het van relegere in
tegenstelling tot neglegere afleiden, en opvatten als aandachtige
zorg en eerbiedigen schroom, etymologieën kunnen ons nu eenmaal
nimmer leeren wat voor ons een woord heden ten dage beteekent.
Nu is gelukkig in de laatste jaren gebleken dat de resultaten der
onderzoekingen aangaande het begrip godsdienst veel meer
overeenstemmen dan men, te oordeelen naar de zoo verschillend
klinkende definities
[6]
, wel zou meenen. Dat religie iets specifiek
menschelijks is, daarover zijn wij het allen wel eens.
[7]
Wanneer
echter de mensch alleen religie heeft en deze derhalve in den
mensch zijn oorsprong en zetel vindt, wanneer de
religionsphilosophie in de eerste plaats religionspsychologie is, dan

moet ook de geestesfunctie zijn te vinden, waardoor zij gedragen
wordt. En dat deze het gevoel is, ook dat wordt tegenwoordig
nauwelijks meer bestreden, al mogen wij er dan wel terstond
bijvoegen, dat daarbij niet meer gedacht mag worden aan die
verouderde indeeling van de menschelijke psyche in drie
onderdeelen waarvan het gevoel er een is, maar dat hier sprake is
van die grondkracht onzer ziel, die als uit de diepte omhoog streeft
naar de oppervlakte en het gansche leven omvatten en beheerschen
wil.
[8]
In de praktijk wordt dat weliswaar nog niet zoo algemeen erkend.
De waarde die in zaken van godsdienst aan bepaalde formules en
formuleeringen, aan geloofsbelijdenissen en hun inhoud wordt
toegekend, doet ons duidelijk inzien, dat het intellectualisme van
vroeger tijd nog bij lange na niet is overwonnen en dat juist in ’t
christendom de „orthodoxie”, de rechtgeloovigheid veelal nog als de
kern van den godsdienst wordt aangezien, waarbij dus alle nadruk
valt op een theoretische overtuiging, een voor waar houden. Nu zal
wel de verhouding tusschen gelooven en weten eerst in het
volgende hoofdstuk meer in het bijzonder besproken worden, maar
nu reeds kan het zijn nut hebben zich de dingen een weinig helder
en klaar voor oogen te stellen om de onjuistheid van dit
intellectualisme te doorzien. Immers als dat het bij ’t rechte eind
had, dan zou de historicus die de zuiverste kennis van het religieuze
verleden, de philosoof die de klaarste gedachten over God en zijne
verhouding tot de wereld, de logische denker die het bestgeordende
samenstel van godsdienstige voorstellingen en gedachten bezat en
kende, ook de meest religieuze en vrome mensch zijn. Men behoeft
zich dat maar in te denken of het uit te spreken om de onjuistheid
ervan in te zien. En geen orthodox theoloog zal er dan ook ernstig
aanspraak op willen maken, om wegens zijn theologische
geleerdheid of zijn inzicht in geloofszaken, voor vroom te worden
gehouden. Door onontwikkelden en armen, door onwetende
visschers is de christelijke godsdienst in de wereld gebracht, door
hen is hij het eerst aangenomen. Een godsdienst der „armen van
geest”, der onontwikkelden en ongeleerden is hij in ’t begin geweest.

Mocht het heden anders zijn gesteld, dan zou dat een bewijs zijn dat
hij zichzelf ontrouw is geworden.
Aanbevelenswaardiger dan deze intellectualistische opvatting, schijnt
die andere te zijn, die den godsdienst verlegt naar het gebied van
den wil en derhalve naar dat van ’s menschen handel en wandel.
„De zuivere, onbevlekte vroomheid voor God den Vader is deze:
weduwen en weezen in hun verdrukking bezoeken, zich zelven
onbesmet bewaren van de wereld”—zoo wordt deze meening op
heel eenvoudige wijze door den Jacobusbrief (1
27) vertolkt. Maar ook
hier wederom is slechts weinig bezinning noodig om in te zien, dat
daarin toch de geheele godsdienst niet kan bestaan.
[9]
Ook de
ongodsdienstige mensch kan dat volbrengen, en de godsdienstige
kan het nalaten. Goed doen is geen speciale vroomheidsuiting: men
kan, gelijk wij nog nader zullen zien, goed zijn zonder vroom te zijn
en men kan ook in quiëstistische vroomheidsstemming zijne handen
in den schoot leggen en niets doen, zooals de pilaarheiligen en
kluizenaars in het Christelijk of Boeddhistisch monnikwezen ons
toonen; of men kan zelfs in religieus fanatisme, dus uit louter
vroomheid in allerlei opzicht kwaad en zonde bedrijven en
goedkeuren. En wat die goede werken betreft, waar het op zou
moeten aankomen,—uit welke gezindheid, uit welke motieven ze
ontstaan zijn, dat alleen beslist, zoowel over hun waarde voor den
dader, als ook daarover, of het daden uit vroomheid of uit
zedelijkheid of wel producten van menschelijke ijdelheid zijn. Of zijn
er misschien naast die zedelijke nog specifiek religieuze uitingen en
daden, met name datgene wat wij onder cultus verstaan: het
kerkgaan, het offeren, het gebed, het deelnemen aan de
sacramenten en dergelijke? Maar wie zou niet weten dat dergelijke
uiterlijke dingen en het deelnemen aan al deze uiterlijke kerkelijke
handelingen, niet het wezen van den godsdienst uitmaken. Men
hoore slechts de oud-testamentische profeten, hoe verachtelijk zij
zich uitspreken over het offer en hun Jahwe laten zeggen: „Ik heb
lust tot weldadigheid en niet tot offer, en tot de kennis Gods, meer
dan tot brandofferen.” (Hosea 6
6). Dat een onwaardig eten en
drinken voorkomt bij het mysterie des avondmaals, daarop wijst

reeds de apostel Paulus in zijn brief aan de Korinthiërs (Kor. 11
27). En
dat er onder ons maar al te velen zijn die uit mode, karakterlooze
lafheid of berekening, uit oogendienst of baatzucht elken Zondag ter
kerke gaan, weten wij allen. Wij zouden velen ronduit dat diep
verontwaardigde „gij huichelaars” kunnen toevoegen, dat Jezus den
Pharisaeërs en Schriftgeleerden in het aangezicht slingerde
(Matthaeus 23
17). En het zou heden niet minder waar zijn van
duizenden en nogmaals duizenden, die Heere, Heere zeggen en zich
naar zijnen naam noemen. Derhalve het handelen alleen beslist niet,
op de motieven komt het aan, en motieven zijn steeds gevoelens; in
het rijk der gevoelens of populair uitgedrukt, in ons gemoedsleven
hebben wij dus ook den zetel der religie te zoeken.
[10]
Waarbij dan
intusschen reeds hier niet, en ook hier weder niet vergeten mag
worden dat de godsdienst als religiositeit en vroomheid meer is dan
zaak van den enkeling, dat hij tegelijk ook een stuk geschiedenis,
een deel van het geschiedkundig leven is en zoo tegenover den
enkelen mensch staat als objectieve macht, zoodat dan een ieder
weer zich tot zijn godsdienst als tot zulk een werkelijk bestaande
levensmacht in een innerlijke betrekking heeft te stellen.
Zien wij nu echter uit naar dat bepaalde gevoel, dat wij bij den
enkelen mensch als religieus aanduidden, dan kan ik het heden nog
niet anders of beter kenschetsen dan Schleiermacher het gedaan
heeft. Nadat het eerst scheen alsof hij, uitgaande van een
eigenaardige opvatting van het gevoel en de plaats die het inneemt
in ons zieleleven, alle gevoel als religieus zou moeten beschouwen,
omschrijft hij het later nauwkeuriger als „slechthinniges
Abhängigkeitsgefühl”. Volstrekt afhankelijkheidsgevoel, gevoel van
eindigheid en oneindigheid: dat sluit niet, gelijk ons langzamerhand
zal blijken, het gansche wezen der religie in, maar het is het
uitgangspunt, het fundament van den godsdienst, de meest
oorspronkelijk religieuze gemoedsstemming en de grondstemming
van den religieuzen mensch.
[11]
Aan de wijze, waarop dit gevoel ontstaat, leeren wij het zuiverst
zien, waarin het bestaat. Want hiervan moeten wij ons natuurlijk

bewust blijven, dat het niet van ’t begin af ten volle in den mensch
aanwezig is en ook niet in allen even vroeg ontwaakt; religieus
opgevoede kinderen bidden reeds lang voordat bij hen van
werkelijken godsdienst sprake is, kennen veel eerder den zedelijken
plicht der gehoorzaamheid dan het vrome bewustzijn hunner
eindigheid en afhankelijkheid; en dat is ook de reden waarom het bij
vele menschen nooit geheel ontwaakt, misschien juist daarom ook
niet, wijl men het van den aanvang af in een conventioneelen vorm
kleedt en dan uit overschatting van dien vorm de zaak zelve verstikt
en verdrukt. Bij alle normale menschen is de aanleg en het
vermogen aanwezig om godsdienst bij zich te ontwikkelen, maar niet
allen bezitten werkelijken godsdienst, evenmin als alle menschen
den zin voor schoonheid en den drang naar waarheid kennen en
weten in zich tot uiting en volle ontwikkeling laten komen. Maar juist
omdat het zoo natuurlijk is, dat zulke gevoelens ontwaken, zulke
stemmingen bij ons opkomen, zijn zeker de meeste menschen
tijdelijk, nu en dan religieus en vroom gestemd, dikwijls zonder dat
zij zelve weten dat zij het zijn en hoezeer zij het zijn, zonder dat zij
zich bewust worden dat juist datgene wat zij nu gevoelen, religieus
en vroom is. Aan den anderen kant echter hoede men zich voor den
overdreven eisch, dat ieder moment in ’s menschen leven vroom
behoort te zijn en door godsdienst begeleid en gedragen moet
worden. Dat dit iets onnatuurlijks is, is aangetoond door den
theoloog Hase, die tegenover het uitsluitend hunkeren naar de
eeuwige dingen, voor onzen tijd veeleer in de innige verbinding van
den oud-Griekschen en den christelijken geest, in het samengaan
van zin voor de wereld, die iets anders is dan wereldzin, en zin voor
het eeuwige en goddelijke, gelijk de vroomheid die kweekt, de
voorwaarde zag voor levensschoonheid en levensharmonie.
[12]
Hoe komt dan nu dat vrome gevoel tot ontwaken? Daarvoor zijn
geen bepaalde en bijzondere oorzaken aan te wijzen. Het is veeleer
een persoonlijke gemoedstoestand en levenservaring die het
uitgangspunt worden kan en worden moet! En toch heeft de
volkswijsheid reeds lang de waarheid gezien, d. i. het gewoonste en
meest voorkomende geval, het natuurlijke en noodzakelijke

gebeuren begrepen, als zij in het spreekwoord zegt: nood leert
bidden. Een groot verdriet, wellicht de dood van een die ons lief
was, of misschien meer stoffelijke nood: wanneer een landbouwer
staat bij zijn door den hagel neergeslagen velden en al zijn arbeid
vernietigd ziet, of zelfs een veel kleinere teleurstelling, een plan dat
mislukte, een verwachting die onvervuld bleef, of bij gansche
volkeren tijden van zware bezoeking, een oorlog bijvoorbeeld,—daar
overvalt het ons plotseling, dat gevoel dat wij zwak en machteloos,
dat wij hulpeloos, eindig en afhankelijk zijn, dat wij eenzaam en
alleen staan midden in die groote wijde wereld, en dat de loop van
die wereld en de loop der natuur en het rad van het noodlot zonder
mededoogen of erbarmen over ons heen gaan. „Mit unserer Macht
ist nichts gethan, wir sind gar bald verloren,” zong Luther en dat is
ook onze stemming, als ons het noodlot voert tot de grenzen van
ons kunnen en onze macht, en met harde slagen ons die grenzen tot
bewustzijn doet komen. En zooals het gaat met onze macht, zoo
voelen wij ook de beperktheid van al onze geestelijke vermogens, de
zwakheid onzer zedelijke kracht en van ons verstand: ook daar
stooten wij maar al te vaak op de beperktheid, de enge grenzen van
ons kunnen en kennen, ons weten, ons begrip, ook daar komen wij
maar al te vaak tot het besef van onze eindigheid en ons
onvermogen. En wanneer de mensch staat hoog op de bergen, in
het gebied van de eeuwige sneeuw, hij alleen te midden der
geweldige, om menschen en menschelijk doen zich niet
bekommerende natuur, of wanneer hij opziet naar dien oneindigen
sterrenhemel, die zich over hem welft en weet dat de aarde die hem
draagt slechts een nietig deel is van een dier ontelbare
wereldsystemen daar boven en het armzalige menschenkind slechts
een atoom daarvan is, atoom van een atoom, dan komt het over
hem, dat gevoel van verlatenheid en eenzaamheid, de angst en
vrees, de kleinheid en nietigheid, de oneindige eindigheid van zijn
bestaan en zijn wezen.
En dat zou dan godsdienst zijn? Natuurlijk niet—nog niet, en toch ja!
Dit moet aanwezig zijn, hoe dan ook, maar het moet niet het eenige
zijn. Dat gevoel van eindigheid en afhankelijkheid zou ons

terneerdrukken en verstikken, als het ons ook niet tevens boven zich
zelf ophief, ons van het eindige losmaakte en zich verbreedde tot het
verlangend en verlossend gevoel van het oneindige. Uit de engte
naar buiten in de ruimte, over het kleine heen tot het groote,
opwaarts van het eindige tot het oneindige, van zwakheid tot sterkte
en kracht,—dat is het verlangen, dat ons in zulke uren van nietigheid
en eindigheid doortrilt. In deze tweeheid, in dezen overgang van het
gevoel van beklemming tot dat van verheffing, van gedruktheid tot
vrijheid en opbruisende kracht, van zwakheid tot sterk en moedig
zijn, ligt de betrekking van het religieuze tot het aesthetische gevoel
van het verhevene: onlust en lust in beiden. Vandaar ook in den
godsdienst steeds weer de wisseling van zondebesef en genade, van
hemel en hel, van jubelen en klagen. Zij behooren als pool en
tegenpool bijeen, al overweegt bij de verschillende menschen, en bij
denzelfden mensch in verschillende tijden en stemmingen nu het
eene, dan het andere.
Terwijl zoo het hart, daareven nog zoo angstig en klein, zoo
moedeloos en neerslachtig, verlangend zich verheft en zoekt naar
ruimte en vrijheid, is nu naast het gevoel van eindigheid en
afhankelijkheid ook dat der oneindigheid ontwaakt. Ook dit vertoont
zich niet in een enkele, maar in duizend vormen, voor alles, zooals
ons Feuerbach
[13]
heeft getoond, als wensch, vaak als recht
egoïstische wensch; en daarin ligt tevens het uitgangspunt voor alle
pathologische en bijgeloovige vormen van religie; daarin echter ook
het recht tot bevrediging der gemoedsbehoeften. O, dat er toch een
oneindige macht was tegenover uwe zwakheid en onmacht! roept de
hulpelooze mensch vol zielsverlangen uit; dat zij een wonder
verrichtte ten uwen behoeve; dat de dooden weer levend werden;
hadt gij toch vleugels—lichamelijke, die u droegen naar de wijde
verten en u toonden de gansche wereld,—geestelijke vleugels die u
voerden over de afgronden van het noodlot en u verlossing brachten
en redding in alle nooden des levens, uw eigen zonde en alle
gevaren die u omringen! Of, als er dan toch maar een hart was dat
met u meevoelde, dat in ’t over u komende leed zich uwer
erbarmde; een rechtvaardig rechter tegenover de

onrechtvaardigheid der menschen; een helper en verlosser, een
Algoede en Almachtige! Doch ’t moge dan in zulke bewuste
wenschen worden belichaamd of niet, al of niet worden
uitgesproken, het verlangen is er, een bovenaardsch, een oneindig
zielsverlangen, dat geheel van zelf wordt tot een verlangen naar het
oneindige, en toch ontwaakt te midden van het eindige en in ’t
gevoel van eigen eindigheid, en er daarom ook volledig mede
samensmelt, het opheft ver boven het aardsche en alledaagsche uit
en bezielt met een te voren ongekende sterkte en kracht.
[14]
En dat is dan godsdienst? dat moet de godsdienst zijn in zijn kern en
wezen? Ja; maar het is nog steeds niet alles, niet de geheele
godsdienst. Want zoo het gevoel in het menschelijk zieleleven het
diepste en oorspronkelijke is, waaruit al het andere zich ontwikkelt
en opgroeit, moet ook dit religieuze gevoel een rijk leven van
voorstellingen van allerlei aard scheppen en zijn leven uiten in alle
daaruit voortgekomen en daardoor beheerschte handelingen, waarin
het zich omzet en tot uiting komt. Die stemming van afhankelijkheid
en zielsverlangen voert daarom noodwendig tot een geloof en tot
een eeredienst. Daarin vindt het alles waarop het zich richt en wat
het wil bezitten, zijn steun en houvast, zijne bescherming, zijn
toetssteen en waarborg. Zoo ontstaan religieuze voorstellingen en
geloofsovertuigingen die van deze machtige grondstemming der ziel
uit het geheele wilsleven richten en verheffen en den mensch doen
leven en handelen uit de kracht van dat geloof.
Daarover, over het rijke, naar buiten zich uitende leven der religie
zullen wij spreken in de beide volgende hoofdstukken.

 

TWEEDE HOOFDSTUK
Religieuze voorstellingen en denkbeelden. Gelooven en
Weten.
 
elooven en weten, daarbij zullen wij ’t eerst onze aandacht
hebben te bepalen. Men zou er wellicht over kunnen
twisten, of niet een behandeling van den cultus aan die van
het geloof vooraf behoorde te gaan. Nog niet lang geleden
n.l. is men bij religionsphilosophische onderzoekingen juist met den
cultus als het oorspronkelijke en eerste in den godsdienst begonnen.
[15]
Ik ben echter van oordeel, dat deze wijze van behandeling niet is
door te voeren: voor men tot een God kan bidden of hem een offer
kan brengen, moet men toch eerst een God bezitten en van hem
weten, en daarom blijft toch altijd de eerste vraag: hoe ontstaat de
Godsvoorstelling, of in ’t algemeen het godsdienstig geloof?
Nu schijnt het alsof wij in weinige stappen daartoe kunnen komen:
afhankelijkheidsgevoel—afhankelijk van wat en van wien? van God;
zielsverlangen naar het oneindige—en dit oneindige? natuurlijk God.
Maar ik vrees dat wij daarmede slechts een naam hebben
gewonnen, niets meer, en het is ons niet om een naam—„een naam
is niets dan klank en rook,”—maar om de zaak te doen. De vraag:
vanwaar eigenlijk dat gevoel van afhankelijkheid? wien of wat
hebben wij onder dat oneindige te verstaan? komt mij voor veel te
theoretisch en te verstandelijk te zijn.
[16]
Zoo vraagt en overlegt de
vrome mensch niet. Op den causaliteitsdrang laat zich het
godsdienstig geloof niet grondvesten. Andere krachten dan die van
het bespiegelend denken zijn daarin werkzaam: gevoel is alles,
daarmede is het begonnen, en zoo brengt nu ook het gevoel het

denken in beweging en blijft daarin leven en werken. Dit
gevoelmatige denken noemen wij fantasie. Dat er fantasie bij in het
spel is, ligt reeds in onze afleiding van ’t religieuze gevoel zelf
besloten. Met ons weten is het hier uit, hebben wij gezegd, wij staan
aan de grenzen van ons kennen en begrijpen. Een zielsverlangen
voert ons verder, en de vleugels die wij noodig hebben om verder te
komen, schept de fantasie en draagt ons daarmede boven den
afgrond uit naar de hoogten, ver weg over alle grenzen en
beperking.
[17]
Zij lost ’s levens raadsels op, leert ons wereld en leven
verstaan, toont ons het oneindige in een beeld, en plant haar banier
op alle hoogten en in alle diepten der wereld en des levens.
Maar nu dringt zich daartegenover als met ongeduld de vraag naar
voren: „Zoo is u dan godsdienst en godsdienstige voorstelling slechts
verdichting; want alle werk der fantasie is poëzie; maar hoe staat
het dan met de waarheid van ons geloof? Wordt zoo niet alles illusie
en zelfbedrog?” Al komt die vraag eigenlijk nog te vroeg, wij kunnen
ons toch reeds voorloopig door twee tegenvragen eenigermate
geruststellen. Dat de religieuze voorstellingen der Grieken
verdichting zijn, daaraan twijfelt wel niemand, evenmin aan het
dichterlijk karakter der Indische mythologie, zooals wij die kennen
uit de heilige hymnen aan Indra en Varuna, aan Soma en Agni; sinds
Herder ook wel nauwelijks nog één ontwikkelde aan den geest der
Hebreeuwsche poëzie in het oude Testament, in ’t bijzonder aan het
dichterlijk karakter van het diepzinnige Paradijsverhaal. Maar
waarom zou dan alleen aan ónze religieuze voorstellingen de
fantasie geen deel mogen hebben? Wij vieren toch allen Kerstmis:
wie zou nu aan de poëzie van het Kerstfeest aanstoot nemen, omdat
ook hier vrome verdichting en sage het heilige tafereel hebben
geteekend?
Maar dan in de tweede plaats, is verdichting dan zooveel als
onwaarheid en leugen? Men behoeft heusch niet eerst aan het
moderne realisme te herinneren, dat de waarheid tot een wet der
poëzie wil maken, om in te zien dat juist de hoogste en laatste
waarheden ons niet door de natuurwetenschappen of de

geschiedenis, maar integendeel door de groote dichters
geopenbaard worden: zij zijn de groote zieners der menschheid en
verklaren haar den zin van haar bestaan. Is dan de schildering die
Schiller in Wallenstein geeft van menschelijke vrijheid en van de
macht van het noodlot, of Goethe in den Faust van den goeden
mensch, die zich in zijn duisteren levensdrang toch van den rechten
weg bewust blijft, om van Lessings juist hier zoo treffenden Nathan
niet te spreken,—zijn dergelijke schilderingen nu onwaar, omdat zij
in schoonen vorm zijn gekleed en in dichterlijke beelden hunne
gedachten over den mensch en het leven tot onze ziel brengen?
Daar wordt ons echter een uitweg gewezen die langen tijd in de
theologie als het ei van Columbus werd beschouwd: voor den
godsdienst, zoo zegt men, zijn denken en voorstellen slechts
secundair; de godsdienst kent slechts gevoelsoordeelen, die door het
gevoel worden te voorschijn geroepen en aan gemoedsbehoeften
beantwoorden, en alleen op het gemoed opbouwend en stemmend
werken: het zijn waardeeringsoordeelen.
[18]
Maar daarmede wordt
de eigenlijke hoofdvraag voorbijgezien, de vraag waarop het
godsdienstig geloof allen nadruk legt en leggen moet: of aan die
beelden en symbolen, die gevoels- en waardeeringsoordeelen ook
waarheid toekomt, of de werkelijkheid er aan beantwoordt, of zij
derhalve ook „zijnsoordeelen” mogen heeten. Natuurlijk beaamt het
geloof dit: niet dat zij waarde hebben, maar dat zij waar zijn, dat is
de hoofdzaak. Maar heeft het daartoe recht? Zijn zij werkelijk waar?
Dat moet dan de philosophie bewijzen. Maar kan zij dat? En
bovendien, komt zij dan daarmede niet toch weer boven het geloof
te staan, en wordt dit daardoor niet eenigermate onder haar voogdij
en contrôle gesteld? In deze lastige situatie treedt nu de religions-
philosophische opvatting van ’t geloof naar voren en toont ons den
eenig mogelijken uitweg uit het labyrinth der hier zich kruisende en
slingerende paden. Daarbij kan zij uitgaan van de bekende definitie
uit den Hebreërbrief: „Het geloof is een vast vertrouwen op dat wat
men hoopt, en niet twijfelen aan dat wat men niet ziet.” (Hebr. 11
1)
De voorwerpen des geloofs zijn onzienlijk, dat is dus tevens
onkenbaar en onbewijsbaar. Dat is in de wijsbegeerte door niemand

scherper gezien en uitdrukkelijk aangetoond dan door Kant, toen hij
aanwees dat zelfs de drie grondideeën der Aufklärung
[19]
: God,
vrijheid en onsterfelijkheid nimmer als kennis der werkelijkheid zijn
te beschouwen en hun bestaan zich ook voor het denken nooit
afdoende en zonder innerlijke tegenspraak laat bewijzen. Maar
wanneer dan nu de wijsbegeerte er van moet afzien, als waar te
bewijzen wat het geloof gelooft, wil dat dan zeggen dat het geloof
minderwaardig en zijne oordeelen geen zijnsoordeelen zijn? Kant zelf
was zoo ver van deze meening verwijderd, dat hij veeleer de
overtuiging koesterde dat hij met de vernietiging der oude
metaphysica en hare schijnbewijzen eerst voorgoed het geloof plaats
verschaft en grond onder de voeten gegeven had. En inderdaad, een
slecht bewijs doet altijd afbreuk aan de zaak zelf. Men houdt haar
zelve voor weerlegd, ook wanneer slechts de betrekkelijk toevallige
manier waarop men haar bewijzen wilde, onhoudbaar is gebleken.
Juist wanneer, èn omdat het geloof in de onzienlijke dingen nooit in
een weten kan worden omgezet, juist dan en ook alleen dan en
daarom blijft het in alle eeuwigheid wat het is:—Geloof.
Nu kan weliswaar ook de wetenschap van den bodem van haar
kennis uit, komen tot de aanvaarding van iets onzienlijks, ook door
haar is, in de philosophie, de stap van wereldverklaring tot
wereldbegrip steeds op nieuw gedaan. Maar hier spreekt men dan
ook in tegenstelling tot „weten” van „enkel gelooven”. Omdat het
zich niet laat bewijzen, blijft het voor het weten slechts hypothese,
en als hypothese draagt het de onzekerheid, de mogelijkheid dat het
ook wel niet zoo, dat het ook anders zou kunnen zijn, draagt het in
een woord den twijfel in zich.
Maar heeft dat dan ook maar een spoor van gelijkenis met datgene,
wat de godsdienstige mensch zijn geloof noemt en in zijn geloof
bezit? Wordt hier niet juist datgene ingesloten en verondersteld, wat
ginds uitgesloten is? Het wetenschappelijk geloof is onlosmakelijk
verbonden met den twijfel aan zich zelf en aan de juistheid van wat
het hypothetisch heeft aangenomen; twijfelen en zoeken op gevaar
van dwaling af is zijn recht en zijn plicht, zijn noodzakelijke

levensvoorwaarde: „dubito ergo sum”, twijfel bewijst het bestaan der
wetenschap. Het godsdienstig geloof daarentegen is boven allen
twijfel verheven; voor het geloof wordt twijfel tot zonde, dwaling tot
schuld, het is in waarheid een vast en onwrikbaar vertrouwen, dat
niet twijfelt aan wat geen menschelijk oog ooit gezien, geen
menschelijk oor ooit vernomen heeft. De wetenschap schrijdt
rusteloos voort, voor haar geldt geen stilstand en geen einde, voor
haar onderzoek toom noch teugel; het geloof daarentegen rust in
zijn klaar en zeker bezit, is conservatief, houdt wat het heeft en vindt
daarin zijn taak en zijn doel. Bedenken wij dan nog, dat het
onzichtbare als voorwerp van wetenschap iets gansch onbepaalds,
iets duisters en problematieks, voor den godsdienst echter vol leven,
aanschouwelijk vast en klaar, „als ziende der Onzienlijke” en tegelijk
iets gansch persoonlijks is, waarbij de geloovige vertrouwt dat het
zijn zielsverlangen naar ’t oneindige zal stillen. Dáar denken en zucht
om te begrijpen, hier hart en gevoel; dáar metaphysiek, hier
mystiek; en derhalve dáar het denken van een enkeling of een kleine
school, hier de gemeenschappelijke overtuiging eener groote
gemeente en van alle geloovigen: „wij gelooven allen aan éénen
God”. Maar daarmede komen wij dan toch weer tot de
aantrekkelijkste taak der Religionsphilosophie, het zoeken van een
antwoord op de vraag naar het psychologisch ontstaan van alle
geloof en de daarmee verbonden kracht des vertrouwens.
De wereld is mijne voorstelling, alles is slechts als inhoud van mijn
bewustzijn mij gegeven; maar uit gewaarwordingen en
voorstellingen schept zich de menschelijke geest een buitenwereld,
wijl hij wat in hem is naar buiten verlegt, zijne bewustzijnservaringen
projecteert op dingen buiten zich. Zoo ontstaat het geloof aan een
buitenwereld, dat ons aller deel is en dat toch slechts een geloof is.
Op dit proces van objectiveeren en projecteeren maakt natuurlijk het
religieuze afhankelijkheidsgevoel, het vrome verlangen naar het
oneindige geen uitzondering; ook het oneindige wordt geprojecteerd
en buiten het ik verlegd: zoo ontstaat het geloof aan een oneindige
macht buiten en boven mij, aan een God, aan mìjn God.

Ik leg den nadruk op dit „aan mìjn God”, want dat is juist het
onderscheid tusschen theïstische philosophie en godsdienst. De
eerste zegt: er is een God, en is daarmede niet religieus; de laatste
zegt: gij zijt mìjn God, en is daarbij niet wijsgeerig.
[20]
Dat oneindige
waarnaar de vrome verlangt, is zijn God; hij staat met Hem in
onmiddellijke persoonlijke betrekking, verwacht van Hem
bevrediging voor zich en—daarin heeft ons Feuerbach den weg
gewezen—voor zijne wenschen, hij verwacht van den Oneindige
oneindige bevrediging en eeuwige zaligheid; daarom ook behooren
Godsgeloof en onsterfelijkheidsgeloof in zoo vele godsdiensten
bijeen.
Zullen wij echter in een persoonlijke verhouding tot God kunnen
komen, dan moet Hij ook dienovereenkomstig gedacht en
voorgesteld, Hij moet persoonlijk opgevat worden. Ook dat is niet
speciaal den godsdienst eigen; ons gansche denken is
anthropomorphistisch. Het kind dat de tafel slaat waaraan het zich
gestooten heeft, en zijn pop behandelt alsof het zijn gelijke was,
denkend en voelend als hij zelf, volgt geheel en onbevangen deze
zucht tot personificeeren en anthropomorphiseeren. Weliswaar wordt
langzamerhand het menschelijke en al te menschelijke meer en
meer afgestooten, maar wij houden toch niet op het doode en
levenlooze te bezielen en te vervormen naar analogie met onze
menschelijke persoonlijkheid; ja zelfs in ons streng wetenschappelijk
denken sluipen nog zulke anthropomorphe voorstellingen binnen als
die van kracht en wet, die beiden van den mensch op de natuur en
de stof zijn overgedragen. En zoo wordt dan ook het oneindige
gepersonificeerd en menschelijk gedacht, zoo schept zich de mensch
de goden, de godheid naar zijn beeld.
Het duidelijkste zien wij dat in de Grieksche mythologie, waar poëzie
en beeldende kunst hebben samengewerkt om die goddelijke
menschheidsidealen vorm te geven en dan steeds heerlijker, steeds
schooner en hooger vormen te scheppen. Dat het echter een
algemeen menschelijke trek is, dat toont ons ook de taal, die
overvloeit van personificaties en dus van den aanvang af een

mythologischen trek, iets als een natuurlijke neiging vertoont tot het
vormen van mythen.
Waar echter de mythologie en de mythenscheppende phantasie het
oneindige op deze wijze vermenschelijkt, kan zij het gevaar niet
ontgaan, het te trekken in de sfeer van het eindige. Dat beantwoordt
dan ook geheel aan het godsdienstig gevoel, dat, gelijk wij zagen,
eindig en oneindig tevens is, omdat het zoowel het eindige als het
oneindige tot inhoud heeft. Maar daarin ligt ook het tegenstrijdige
dat alle mythologie eigen is: een eindig gedacht oneindige! En aan
die tegenstrijdigheid knoopt zich dan ook vast het eerste conflict
tusschen gelooven en weten. Zooals bekend is, waren het in
Griekenland moreele bedenkingen die den eersten stoot gaven. De
dichters verhaalden van de goden dingen, die hun zedelijke reinheid
en hoogheid in twijfel brachten: de goden liegen en bedriegen,
plegen echtbreuk en zijn naijverig en afgunstig. Het is merkwaardig
dat het volk—en niet alleen het Grieksche volk!—over het algemeen
aan dergelijke dingen geen aanstoot heeft genomen en neemt; een
feit dat alleen te verklaren is eenerzijds uit een zekere zedelijke
afstomping en gebrek aan ontwikkeling, anderszijds uit een
begrijpelijke afwezigheid van alle kritiek tegenover het religieuze en
goddelijke, die, door het overnemen der overlevering, uit gewoonte
en piëteit nog versterkt wordt. En daarom begint ook dit conflict
nooit als een algemeen verzet, gedragen door het bewustzijn van
een heel volk, maar steeds eerst bij den enkeling, die daardoor komt
te staan tegenover zijn volksgenooten en het algemeen geloof of de
religieuze voorstellingen van zijn volk. Maar daarover later meer.
De religieuze voorstellingen staan aanvankelijk niet geïsoleerd naast
andere, maar omvatten oorspronkelijk het geheel van onze denk-
wereld, zijn den mensch alles. Historisch is dat buiten kijf: in de
Mythologie der Grieken lag opgesloten wat zij van de wereld en
speciaal van de natuur wisten en dachten, mythologie was tevens
hun wetenschap en hun philosophie. En dat is ook volkomen
begrijpelijk: hoe vreemder de mensch nog staat tegenover de hem
omringende natuur, hoe minder hij haar kent en in staat is te

beheerschen, des te afhankelijker is hij in werkelijkheid van dit hem
onbekende, en hoe afhankelijker en eindiger, des te hulpeloozer en
machteloozer gevoelt hij zich daar tegenover. De natuur zelf is voor
hem het oneindige, en wat hij over haar peinst en denkt, dat wordt
ook deel van zijne voorstelling van ’t Oneindige. En omgekeerd
dienen hem de door zijne phantasie gewekte beelden van ’t
Oneindige, om den zin en de beteekenis van de natuur in al hare
verschijnselen te leeren verstaan. Zoo is het dan het goddelijke in de
natuur zelf, dat hem moeiten en belemmeringen van allerlei aard
bereidt, en daarom kan hij ook slechts van dat goddelijke dat zich in
haar uit, hulp, redding en bescherming hopen en deelachtig worden.
En gelijk zich aldus achter al het natuurlijke een goddelijke macht
verbergt, zoo gaan ook wereldopvatting en mythologie, natuurkennis
en Godsvoorstelling voor den mensch op dezen trap van
ontwikkeling naast elkander hand in hand. Zoo verbinden zich steeds
meer ook naar hun inhoud, gelooven en weten tot een
samenhangende religieuze wereldbeschouwing, waarin ook
brokstukken kennis aangaande de wereld godsdienstig gekleurd en
in dienst van religieuze phantasievoorstellingen en
gemoedsbehoeften gesteld worden.
Zoo beheerscht de godsdienst in den aanvang de wetenschap en
neemt haar in zich op. Maar terwijl hij op deze wijze zelf een
religieus weten wordt, komt hij toch ook weder te staan onder de
wetten en voorwaarden der wetenschap, en dat leidt tot
eigenaardige moeilijkheden en verwikkelingen. In het Christendom
zien wij duidelijk hoe zich dat ontwikkeld en verwikkeld, welken loop
het genomen heeft. Gedwongen om de aanvallen der heidenen, en
daaronder scherpzinnige heidenen als Celsus, met geestelijke
wapens af te weren, werd het Christendom tot apologetiek en kon
als zoodanig niet vermijden tegenover de Grieksche philosophie post
te vatten,
[21]
en daarbij ging het zooals het steeds pleegt te gaan:
meer dan éen brokstuk van deze Grieksche philosophie, een groot
deel der Grieksche wereldbeschouwing werd in het Christendom
opgenomen; nog geheel afgezien van den invloed van dat
eigenaardige hellenistische synkretisme, dat door zijn taal reeds de

N.T.-ische litteratuur hellenistisch heeft gekleurd. Daar kwam nog bij
de noodzakelijkheid van godsdienstig onderricht: de religieuze
voorstellingen werden een godsdienstleer, en in een stad van zoo
hooge geestelijke ontwikkeling als Alexandrië werd de
katechetenschool van zelf een philosophische school: het
Christendom werd een philosophie. In de middeleeuwen, toen de
dogmen voorgoed waren vastgesteld, zocht men ze tot een systeem
te vereenigen en dit naar alle zijden te ontvouwen; zoo ontstonden
de groote scholastische stelsels van een Albertus Magnus en Thomas
van Aquino. De geloofsinhoud stond voor deze scholastische
theologen onomstootelijk vast, de andere wetenschappen waren de
dienaressen, bepaaldelijk de philosophie was de dienstmaagd dezer
koningin. De Theologie was alles. Maar reeds tegen ’t einde der
Middeleeuwen begon de Scholastiek zelf, als een tweede Penelope,
het kleed weer uiteen te halen, waaraan zij zoolang geweven had. In
het zoogenaamde Nominalisme kwam de tweestrijd tusschen den
geloofsinhoud en het denken tot uiting, de onbewijsbaarheid des
geloofs werd ingezien, en zoo kwam men tot die eigenaardige leer
van de dubbele waarheid, een soort dubbel boekhouden, waarbij in
de theologie waar kan zijn, wat voor de philosophie onwaar is, en
omgekeerd. Men heeft daarin vaak een oneerlijk spel willen zien,
alsof het den Nominalisten met dezen uitweg geen ernst was en zij
daarmede slechts den schijn van ketterij en ongeloof wilden
vermijden. Ik meen dat men hen daarmede, althans in ’t begin,
onrecht doet; die leer van de dubbele waarheid was veeleer een
teeken des tijds, was slechts de eerste onbeholpen en onzuivere
uitdrukking voor het nu langzamerhand tot bewustzijn komend
conflict tusschen gelooven en weten, dat ons van dezen tijd af niet
meer met rust laat.
Een ander redmiddel, reeds in de middeleeuwen aangewend maar
eerst in de Protestantsche theologie tot volle ontwikkeling gekomen,
was het supranaturalisme met zijn leer, dat er in den godsdienst en
het godsdienstig geloof dingen zijn die, gelijk boven de natuur—
vandaar de naam—, ook boven het verstand verheven zijn, dingen
die niet tegen de rede in-, maar boven de rede uit gaan. Dat was

hier de oplossing. Het geheimzinnige en onbegrijpelijke wordt nu tot
een toevluchtsoord voor den godsdienst; hier heeft hij recht van
spreken, voor ’t overige echter beslist het verstand. Maar zoo
eenvoudig als zij er uitzien, zijn dergelijke grensbepalingen op den
duur toch onhoudbaar: wie zal beslissen waar het gebied des
verstands eindigt en het rijk van het bovenredelijke aanvangt? Wie
kan, wie zal het zoekend en vorschend verstand paal en perk
stellen? Zoo zag zich dan ook inderdaad het supranaturalisme
gedwongen van zijn gereserveerd gebied het eene stuk na het
andere aan het verstand over te leveren en aan zijn rechtspraak te
onderwerpen. De rede kan zich de aanmatiging van zulke
bovenredelijke waarheid niet laten welgevallen; alles moet haar zijn
rechten kunnen bewijzen om als redelijk te worden erkend,—dat is
de eisch van het Rationalisme, en zijn doel en ideaal is dan ook een
redelijke godsdienst. Zoolang dit rationalisme op den
voorstellingsinhoud van den godsdienst kritiek oefent, zijne
historische bestanddeelen onderzoekt op hun geschiedkundige
juistheid en geloofwaardigheid, de houdbaarheid toetst van de
wijsgeerige onderstellingen der geloofsleer, is het in zijn recht. Maar
het begaat onrecht en tast mis wanneer het de beteekenis van dit
geloof voor het religieuze gevoel miskent en vergeet, en een
godsdienst zoekt te construeeren, die uitsluitend het denken moet
bevredigen, en van het recht of het aandeel der fantasie daarin niets
wil weten. Want daarmee plaatst het zich op een intellectualistisch
standpunt, dat wij in het vorige hoofdstuk reeds als dat der
rechtgeloovigheid hebben leeren kennen en verworpen hebben.
Wanneer dan zoo alle pogingen tot verzoening blijken schipbreuk te
lijden, is dat dan niet een bewijs voor de juistheid van het woord van
Strausz: „laat dan de geloovige den denker maar rustig zijn gang
laten gaan, en evenzoo omgekeerd; wij gunnen hun gaarne hun
geloof, en zij moeten het ons maar onze philosophie doen”?
[22]
Als
dat nu ook maar mogelijk was! Evenwel wanneer wij er over
nadenken wáarom dat onmogelijk is, juist dan kunnen wij misschien
verder komen. „De geloovige”—dat wijst reeds op éen van de
redenen. De godsdienst voert met noodwendigheid tot religieuze

voorstellingen, en aan deze wordt—geloofd. De dichter weet dat de
beelden zijner fantasie vrij gevonden zijn, wijl hij zelf ze gevormd en
in ’t leven geroepen heeft. De scheppingen der religieuze fantasie
echter zijn veelmeer onbewust, mythen worden niet gemaakt maar
ontstaan zooals de taal en voor een deel met de taal. En omdat zij
over ’t heilige handelen, worden ze al spoedig zelve voor heilig
gehouden, wordt hun een heilige oorsprong toegekend. Bovendien,
zij zijn de uitdrukking van een oneindig zielsverlangen, dat zij nu in
den vorm van een voorstelling, een beeld willen weergeven; daarom
wordt dat verlangen er ook door bevredigd en vindt er zijn doel in,
lossen daarin en daardoor alle ellende van het bestaan, alle
raadselen van leven en wereld zich op. Daarom blijft het hart er aan
gehecht, klinkt er de volle toon van het gevoel in, en hebben zij een
geheel eenige gemoedswaarde. Maar het hart is conservatief, veel
conservatiever dan het critische verstand: wat het eenmaal heeft
gegrepen dat houdt het vast en laat het niet meer los. En dan komt
daar nog bij de macht der traditie, een nu bijna tweeduizendjarige
traditie van de kerk met hare heilige overlevering en de bij ieder
afzonderlijk weder nieuwe, bezielende macht der indrukken van
jeugd en kindsheid. Wat wij van onze jeugd af als het beste en
schoonste, als het hoogste en heiligste hebben leeren kennen, en
wat ons de liefste en beste menschen, onze ouders, onze moeder
hebben ingeprent, dat blijft en houdt ons vast, zelfs tegen onzen wil,
zelfs als wij het niet vasthouden willen. Zoo kunnen wij begrijpen
hoe het religieuze geloof met het innigste van ons wezen, met ons
diepste zielsverlangen en ons karakter samenhangt, hoe het vaste
overtuiging is en onwrikbaar vertrouwen, dat niet twijfelt, maar
verzekerd is van zijn volstrekte waarheid.
In deze zelfverzekerdheid van het godsdienstig geloof ligt echter ook
zijn tot op zekere hoogte noodzakelijke onverdraagzaamheid
besloten. Als het waar en absoluut waar is, wat de vrome gelooft,
dan is al het andere onwaar en alleen onwaar, en ieder die anders
gelooft dan hij is hem een ongeloovige. Als mijn geloof mijn
hartsverlangen stilt, mij volkomen bevredigt en rustig stemt, mij
gelukkig en zalig maakt, dan is onzalig en van alle zaligheid

verstoken, al wie niet gelooft gelijk ik. En als dit mijn geloof mij
heilig en vol heilige verzekerdheid is, dan is ieder die daaraan twijfelt
en tornt, onheilig, oneerbiedig, zondig. Dat uit zich dan al spoedig in
gedachten over den waren God tegenover de valsche goden van
anderen, in de leer van de eeuwige verdoemenis der heidenen en in
het exclusieve: „extra ecclesiam nulla salus” (buiten de kerk geen
heil) of in de verschrikkingen der godsdienstoorlogen en de
gruwelen der kettervervolgingen. Zoo laat in de werkelijkheid de
eene geloovige den anders-geloovige en in deze dingen
andersdenkende, de geloovige den ongeloovige niet met rust, niet
vreedzaam zijns weegs gaan.
Maar nog om een gansch andere, meer innerlijke en zakelijke reden
moet het geloof wel met het weten en de wetenschap in botsing
komen. De religieuze voorstellingen worden door de fantasie
gewekt; maar ook deze mythenvormende fantasie en de bij haar
zich aansluitende vorming van dogmata is een kind van haar tijd en
staat onder den invloed en de inwerking der haar omringende
wereldbeschouwing. Hoe leerstelliger die godsdienstige
voorstellingen zijn en worden, des te meer nemen zij van de
denkbeelden des tijds in zich op. Zoo is de Christelijke dogmatiek,
gelijk reeds gezegd is, slechts te verstaan en te verklaren zoo wij
haar zien als gegroeid op den bodem der Grieksch-Hellenistische
Philosophie: een fantasievolle wereldbeschouwing, beantwoordend
aan de behoeften van een hooggespannen godsdienstig gevoel,
opgetrokken uit bouwsteenen der toenmalige wetenschap en
philosophie, zoo staat zij voor ons in haar heilige oorkonden, in haar
eerste belijdenisgeschriften en hare vroegste formuleeringen, in haar
oudste aan apologie of onderricht gewijde litteratuur. Nu weten wij
wel, de behoeften van hart en gemoed blijven in hun wezen voor
alle tijden dezelfde; maar de bouwsteenen, waarmede de religieuze
fantasie werkt en schept, zij gelden slechts tijdelijk en hebben geen
eeuwige waarde, zij verouderen en verliezen allengs hun aanspraak
op waarheid, juistheid en algemeengeldigheid, verliezen
langzamerhand hun waarde.

En dan in de tweede plaats: bij die godsdiensten die in ’t verleden
wortelen, historische personen tot stichters en dragers, historische
feiten tot uitgangspunt en grondslag hebben, heeft ook in de
opvatting der gebeurtenissen de fantasie een groote rol gespeeld.
Die gebeurtenissen en hunne overlevering, zij zijn naar de behoeften
des harten veranderd en verdicht, en bovenal het beeld van den
stichter zelf heeft zich in het liefhebbend hart en niet in het critisch
verstand, door geloof en niet door weten gevormd. Welnu, deze in
het geheel van de godsdienstige voorstellingen mede opgenomen
historische gegevens zijn wetenschappelijk te onderzoeken en
moeten geverifiëerd worden.
Zoo komt de wetenschap van twee kanten, ongezocht en niet gewild
en zonder zucht tot strijd, met het geloof in aanraking en in
tegenspraak. Men wil van de heilige dingen en die heilige personen
nog meer weten, doet nasporingen en vindt, dat veel van wat tot
dusver verhaald en geloofd werd—sage en legende is en dat de
heilige oorkonden geen vertrouwbare geschiedbronnen zijn. En aan
den anderen kant gaat de wereldsche wetenschap steeds voort op
haren weg, en komt daarbij allengs tot het inzicht, dat, wat de
vrome gelooft, niet te bewijzen is; komt verder bij de verklaring van
natuur en wereld tot gansch andere resultaten, dan die voor duizend
en tweeduizend jaar geleden. De wetenschappelijke beschouwingen
en verklaringen der Grieken omtrent aard en bouw der wereld is
onhoudbaar gebleken, en toch vormen zij de fundamenten waarop
de groote dogmata van het Christelijk geloof zijn opgetrokken. De
Christelijke opvatting der verhouding van God en wereld berust op
wat Aristoteles leerde omtrent hemel en aarde en hun betrekking tot
elkaar; en zelfs de verhalen van een hemel- en hellevaart van
Christus worden beheerscht door—ja zijn ook alleen mogelijk bij een
geocentrische wereldbeschouwing, zooals die der Grieksche
astronomen en natuur-philosophen. Maar deze oude
wetenschappelijke veronderstellingen zijn weggevallen, ons
wereldbeeld is een ander geworden; en zoo komen wij nu voor een
conflict te staan, het conflict tusschen religieuze voorstelling en
moderne wereldbeschouwing. Wel klampt zich het geloof nog een

tijdlang vast aan de hoop, dat de wetenschap het mis heeft, en het
beroept zich daarbij op het onzekere en wisselende der
wetenschappelijke hypothesen en op het hypothetische en
fragmentarische van alle menschelijk weten, maar op den duur is
deze troost toch niet voldoende, en men kan zich ten slotte niet
langer ontveinzen, dat de wetenschap anders weet dan het geloof
gelooft.
En naast deze groote, uitsluitend zakelijke tegenstelling komen dan
nog wederzijdsch allerlei oorzaken van meer persoonlijken aard het
conflict verscherpen. Het geloof ziet in de wetenschap zijnen vijand
en in haar arbeid een oneerbiedig, ja goddeloos bedrijf, en daarom is
ten allen tijde het geloof jegens de wetenschap onverdraagzaam
geweest en heeft haar vaak tot het uiterste vervolgd. Maar
anderszijds beschouwt de wetenschap juist daarom het geloof als
een beletsel voor haar vrije ontwikkeling en vooruitgang, voor de
erkenning en verbreiding van haar resultaten en gedachten, en komt
zoo veelal tot een hoogmoedige minachting van het geloof, alsof dit,
wijl het momenteel met oude steenen bouwt, zelf iets verouderds en
onnoodigs geworden is, den volgroeiden mensch en zijn rijpere
gedachten onwaardig; of tot een koel verstandelijk negeeren der in
religieuze voorstellingen naar bevrediging zoekende en snakkende
behoeften des harten.
In dezen strijd neemt dan een tijdlang de theologie de
bemiddelaarsrol op zich. In den aanvang bestemd om de religieuze
voorstellingen tot een systeem bijeen te voegen, moet zij deze nu
met de voortschrijdende wetenschap in overeenstemming brengen
en verzoenen, en ook voor het denken rechtvaardigen. Maar,
hoelang dat haar ook moge gelukken, door anders uit leggen en er
iets anders in te leggen, vaak door concessies van allerlei aard, aan
’t einde gaat het haar niet anders dan alle overige wetenschappen,
zij zelf wordt „weten”, wordt historische en philosophische
wetenschap en komt nu ook harerzijds in conflict met de religieuze
voorstellingen, waaraan zij haar arbeid gewijd heeft. Zoo hebben wij
het dan ook gezien, beleven het en zien het telkens weer om ons

heen, dat juist de wetenschappelijke theologen in de kerk het
heftigst worden aangevallen, en hun arbeid als schadelijk voor den
godsdienst en het geloof benadeeld, wordt bestreden en verketterd.
Zoo schijnt er dan nu nog slechts één uitweg open te staan: een
ieder vorme en verheldere zijn godsdienstige voorstellingen naar den
stand en de hoogte der huidige wetenschap en neme telkens
opnieuw er in op wat deze als nieuwe winst meent te mogen
boeken! Maar daarbij moeten wij toch al dadelijk herinneren aan het
reeds genoemde conservatieve karakter van het godsdienstig geloof.
En bovendien, godsdienst is niet alleen zaak van den enkeling, maar
altijd ook een historisch verschijnsel, door die tijden heen gevormd
en gegroeid, hij is een zaak der gemeenschap. Daarom treedt ook
altijd dat conflict eerst bij slechts enkelen op; zij komen door den
vooruitgang van hun weten tot het besef dat de godsdienstige
voorstellingen zijn achtergebleven, en staan dan als beoordeelaars
daartegenover. Maar eerst staan zij alleen tegenover de groote
massa der geloovigen en tegenover het geloof eener kerk. Een
tijdlang dan zijn zij, deze vooruitstrevenden als eenlingen, de
martelaars der wetenschap,—soms is dat, gelijk de brandstapel van
Giordano Bruno bewijst, een werkelijk martelaarsschap geweest;
onder beter gesternte, in tijden van zachtere zeden is het ’t
tragische lot der vereenzaamden en uitgestootenen. Allengs echter
dringt de wetenschap en hare resultaten in breedere kringen door;
en dat brengt allerlei gevaren met zich. Hoe grooter de macht der
kerk, en hare machtsmiddelen zijn, des te meer zullen zwakke
karakters hun breuk met de kerkelijke dogmata voor zich houden en
verbergen, zij zullen een intellectueel offer brengen en met den
mond belijden wat zij in hun hart niet meer gelooven. En al mogen
dan deze zwakke en onzelfstandige karakters daarom nog niet alle
als huichelaars te veroordeelen zijn, het is toch niet te ontkennen,
dat in zulke tijden de waarachtigheid en de zin voor waarheid en
eerlijkheid in wijden kring schade lijdt en daardoor zoowel het weten
der heimelijk wetenden als het geloof der niet meer geloovenden
benadeeld en onwaarachtig wordt. Enkele volslagen huichelaars en
vele halve karakters—dat zijn in het strijdperk van gelooven en

Welcome to our website – the perfect destination for book lovers and
knowledge seekers. We believe that every book holds a new world,
offering opportunities for learning, discovery, and personal growth.
That’s why we are dedicated to bringing you a diverse collection of
books, ranging from classic literature and specialized publications to
self-development guides and children's books.
More than just a book-buying platform, we strive to be a bridge
connecting you with timeless cultural and intellectual values. With an
elegant, user-friendly interface and a smart search system, you can
quickly find the books that best suit your interests. Additionally,
our special promotions and home delivery services help you save time
and fully enjoy the joy of reading.
Join us on a journey of knowledge exploration, passion nurturing, and
personal growth every day!
ebookbell.com