Statistical and Fuzzy Approaches to

Data Processing, with Applications to

Econometrics and Other Areas: In
Honor of Hung T. Nguyen's 75th
Birthday Vladik Kreinovich
Visit to download the full and correct content document:
https://textbookfull.com/product/statistical-and-fuzzy-approaches-to-data-processing-
with-applications-to-econometrics-and-other-areas-in-honor-of-hung-t-nguyens-75th-b
irthday-vladik-kreinovich/
More products digital (pdf, epub, mobi) instant
download maybe you interests ...

A First Course in Fuzzy Logic 4th Edition Hung T.

Nguyen

https://textbookfull.com/product/a-first-course-in-fuzzy-
logic-4th-edition-hung-t-nguyen/

Optimization and Its Applications in Control and Data

Sciences In Honor of Boris T Polyak s 80th Birthday 1st
Edition Boris Goldengorin (Eds.)

https://textbookfull.com/product/optimization-and-its-
applications-in-control-and-data-sciences-in-honor-of-boris-t-
polyak-s-80th-birthday-1st-edition-boris-goldengorin-eds/

Exploring the World with the Laser: Dedicated to

Theodor Hänsch on his 75th birthday 1st Edition Dieter
Meschede

https://textbookfull.com/product/exploring-the-world-with-the-
laser-dedicated-to-theodor-hansch-on-his-75th-birthday-1st-
edition-dieter-meschede/

Mathematical Analysis With Applications In Honor of the

90th Birthday of Constantin Corduneanu Ekaterinburg
Russia July 2018 Sandra Pinelas

https://textbookfull.com/product/mathematical-analysis-with-
applications-in-honor-of-the-90th-birthday-of-constantin-
corduneanu-ekaterinburg-russia-july-2018-sandra-pinelas/
Econometrics of panel data : methods and applications
First Edition Biørn

https://textbookfull.com/product/econometrics-of-panel-data-
methods-and-applications-first-edition-biorn/

Approaches to the History and Dialectology of Arabic

Papers in Honor of Pierre Larcher Manuel Sartori

https://textbookfull.com/product/approaches-to-the-history-and-
dialectology-of-arabic-papers-in-honor-of-pierre-larcher-manuel-
sartori/

An Introduction To Statistical Learning With

Applications In R Gareth James

https://textbookfull.com/product/an-introduction-to-statistical-
learning-with-applications-in-r-gareth-james/

Approaches to Byzantine Architecture and its Decoration

Studies in Honor of Slobodan ■ur■i■ Mark J. Johnson

https://textbookfull.com/product/approaches-to-byzantine-
architecture-and-its-decoration-studies-in-honor-of-slobodan-
curcic-mark-j-johnson/

Diffuse Algorithms for Neural and Neuro Fuzzy Networks

With Applications in Control Engineering and Signal
Processing Boris A. Skorohod (Auth.)

https://textbookfull.com/product/diffuse-algorithms-for-neural-
and-neuro-fuzzy-networks-with-applications-in-control-
engineering-and-signal-processing-boris-a-skorohod-auth/
Studies in Computational Intelligence 892

Vladik Kreinovich Editor

Statistical and
Fuzzy Approaches
to Data Processing,
with Applications
to Econometrics and
Other Areas
In Honor of Hung T. Nguyen’s
75th Birthday
Studies in Computational Intelligence

Volume 892

Series Editor
Janusz Kacprzyk, Polish Academy of Sciences, Warsaw, Poland
The series “Studies in Computational Intelligence” (SCI) publishes new develop-
ments and advances in the various areas of computational intelligence—quickly and
with a high quality. The intent is to cover the theory, applications, and design
methods of computational intelligence, as embedded in the fields of engineering,
computer science, physics and life sciences, as well as the methodologies behind
them. The series contains monographs, lecture notes and edited volumes in
computational intelligence spanning the areas of neural networks, connectionist
systems, genetic algorithms, evolutionary computation, artificial intelligence,
cellular automata, self-organizing systems, soft computing, fuzzy systems, and
hybrid intelligent systems. Of particular value to both the contributors and the
readership are the short publication timeframe and the world-wide distribution,
which enable both wide and rapid dissemination of research output.
The books of this series are submitted to indexing to Web of Science,
EI-Compendex, DBLP, SCOPUS, Google Scholar and Springerlink.

More information about this series at http://www.springer.com/series/7092

Vladik Kreinovich
Editor

Statistical and Fuzzy

Approaches to Data
Processing, with Applications
to Econometrics and Other
Areas
In Honor of Hung T. Nguyen’s 75th Birthday

123
Editor
Vladik Kreinovich
Department of Computer Science
University of Texas at El Paso
El Paso, TX, USA

ISSN 1860-949X ISSN 1860-9503 (electronic)

Studies in Computational Intelligence
ISBN 978-3-030-45618-4 ISBN 978-3-030-45619-1 (eBook)
https://doi.org/10.1007/978-3-030-45619-1
© The Editor(s) (if applicable) and The Author(s), under exclusive license to Springer Nature
Switzerland AG 2021
This work is subject to copyright. All rights are solely and exclusively licensed by the Publisher, whether
the whole or part of the material is concerned, specifically the rights of translation, reprinting, reuse of
illustrations, recitation, broadcasting, reproduction on microfilms or in any other physical way, and
transmission or information storage and retrieval, electronic adaptation, computer software, or by similar
or dissimilar methodology now known or hereafter developed.
The use of general descriptive names, registered names, trademarks, service marks, etc. in this
publication does not imply, even in the absence of a specific statement, that such names are exempt from
the relevant protective laws and regulations and therefore free for general use.
The publisher, the authors and the editors are safe to assume that the advice and information in this
book are believed to be true and accurate at the date of publication. Neither the publisher nor the
authors or the editors give a warranty, expressed or implied, with respect to the material contained
herein or for any errors or omissions that may have been made. The publisher remains neutral with regard
to jurisdictional claims in published maps and institutional affiliations.

This Springer imprint is published by the registered company Springer Nature Switzerland AG
The registered company address is: Gewerbestrasse 11, 6330 Cham, Switzerland
Foreword

It is always with great pleasure that we celebrate someone’s birthday and it is,
indeed, all the more pleasant for me to celebrate Hung T. Nguyen’s 75th birthday, a
colleague and dear friend I have known for almost 50 years.
In 1970, Hung was appointed at the University of Lille (France) where I was an
associate professor. Both of us had already presented a (Third Cycle) doctoral
thesis, Hung defending his in Paris while I defended mine in Lille.
That year, we had the chance to meet Joseph Kampé de Fériet, a university
professor from Lille who had taken retirement in 1964 and whose research work on
turbulence had been internationally known and appreciated.
He mentioned to us a new theory of information whose foundations he had
rethought in 1966 with two Italian mathematicians, Pietro Benvenuti from the
Institute of Applied Mathematics in Rome, and Bruno Forte from the University of
Pavia. Considering that the notion of information is more primitive than that of
probability, the new theory challenged the probability basis and suggested a more
general version.
The research topic was new. Professor Kampé de Fériet felt very enthusiastic to
work with two young researchers in Lille and offered to develop that new theory
under his direction.
We thus spent five years working together, with this research making it possible
for us to obtain the degrees of Doctorat d’État es Sciences Mathématiques.
Our careers then followed different paths, Hung went to the USA, namely to
California (Berkeley), Massachusetts, and New Mexico, and later to Chiang Mai in
Thailand, while I remained in Lille, developing cooperations in the field of statis-
tics, first with Brazil and later with all South American states.
However, we have had many opportunities to meet each other, notably while
attending congresses and conferences exemplified by: Bangkok (2000, 2001),
Hanoi (2001), Chiang Mai (2003), Houston (2004), Kunming (2005), Chengdu
(2006), Taipei (2005, 2006), Guillin (2009), and Saigon (2012, 2013). What is
more, apart from these scientific events, we have often shared friendly family

v
vi Foreword

meetings, at home (Lille, Las Cruces, Chiang Mai) but also in Japan, Vietnam,
Thailand, in northern Italy (Lake Como and Venice), in Prague, on a cruise down
the Nile and also in Paris where Hung, a talented tennis player himself, attended the
Roland Garros tournaments.
In light of all these landmarks, I therefore wish a memorable birthday to
Prof. Hung T. Nguyen, a distinguished colleague and a great friend, hoping that he
will remain in good health and will keep developing his numerous research ideas.

Claude Langrand
Professor Emeritus
University of Lille
Lille, France

Bibliography

1. J. Kampé De Fériet, B. Forte, Information et probabilité. C. R. Acad. Sci. Paris 265, Sér.A,
110–114, 142–146, 350–353 (1967)
2. J. Kampé De Fériet, P. Benvenuti, Sur une classe d’informations. C. R. Acad. Sci. Paris 269,
Sér.A, 97–101 (1969)
3. J. Kampé De Fériet, B. Forte, P. Benvenuti, Forme générale de l’opération continue d’une
information, C. R. Acad. Sci. Paris 269, Sér.A, 529–534 (1969)
4. J. Kampé De Fériet, T. H. Nguyen, Temps d’entrée d’un processus stochastique et mesure de
l’information. C. R. Acad. Sci. Paris 275, Sér.A, 721–725 (1972)
5. C. Langrand, Constructions de m-précapacités. C. R. Acad. Sci. Paris 275, Sér.A, 1243–1246
(1972)
6. N.-T. Hung, Mesures d’informations, capacités positives et sous-mesures. C. R. Acad. Sci.
Paris 275, Sér.A, 441–443 (1972)
7. C. Langrand, N.-T. Hung, Sur les mesures intérieures de l’information et les r-précapacités.
C. R. Acad. Sci. Paris 276, Sér.A, 927–930 (1972)
8. C. Langrand, Mesures extérieures d’information. C. R. Acad. Sc. Paris 276, Sér.A, 703–706
(1973)
9. J. Kampé De Fériet, N.T. Hung, Mesure de l’information, temps d’entrée et dimension de
Hausdorff. C. R. Acad. Sci. Paris 276, Sér.A, 807–811 (1973)
10. C. Langrand, Information généralisée; estimation et sélection, Thèse de doctorat es sciences
mathématiques, Lille, 1973
11. J. Kampé De Fériet, La théorie généralisée de l’Information et la mesure subjective de
l’Information Actes des Rencontres de Marseille-Luminy, in Théories de l’information,
Lectures Notes in Mathematics, vol. 398 (Springer-Verlag, 1974), pp. 1–35, 5–7 juin 1973
12. C. Langrand, Précapacités fortes et mesures d’information, in Lectures Notes in Mathematics,
vol. 398, (Springer-Verlag, 1974), pp. 36–48
13. N. T. Hung, Sur les mesures d’information de type Inf, in Lectures Notes in Mathematics,
vol. 398 (Springer-Verlag, 1974), pp. 62–75
Foreword vii

14. N.T. Hung, Mesures d’information sur les ensembles ordonnés. C. R. Acad. Sci. Paris 278,
Sér.A, 1139–1142 (1974)
15. C. Langrand, Composabilité d’une mesure d’information limite de mesures d’informations
composables, C. R. Acad. Sci. Paris 279, Sér.A, 727–730 (1974)
16. N.T. Hung, Mesures d’information, ensembles flous et espaces topologiques aléatoires, Thèse
de doctorat es sciences mathématiques, Lille, 1975
Preface

It is my privilege and my honor to start this volume devoted to our good friend,
great colleague, and wonderful teacher Hung T. Nguyen.
I have first met Professor Hung T. Nguyen twenty years ago, in 1990, shortly
after my coming to the USA. Since then, we have published many joint papers, two
joint books, and numerous joint edited books. During these twenty years, I continue
to be impressed by his enthusiasm, by his productivity, and by his ability to
transition to completely research areas.
His contributions are numerous, let me mention a few major ones. Papers
co-authored by Hung T. Nguyen revolutionized our understanding of conditional
probabilities—by showing that these probabilities can be interpreted as probabilities
of naturally defined events. He showed that both Dempster–Shafer belief theory and
fuzzy techniques, two of the most successful post-probabilistic approaches to
uncertainty, can be naturally interpreted in probabilistic terms—namely in terms of
random sets. He showed that a natural way to extend functions (and algorithms in
general) to fuzzy data can be expressed in terms of the ranges of these functions
over naturally defined sets. From statistics and fuzzy, he moved to econometrics,
where he is now developing and actively promoting quantum ideas.
In many of these areas, not only he got interesting results, he also wrote text-
books and monographs and lecture notes that describe these—and other—results in
a very clear and convincing way. These books helped several generations of
researchers understand complex topics and complex results.
In this volume, we, his friends and colleagues, present examples of our results
motivated by Hung's results and ideas. Happy birthday to Hung T. Nguyen. We are
looking forward toward many more years of his ideas and his successes.

ix
x Preface

HUNG T. NGUYEN

A. Awards

(1) Chair of Fuzzy Theory (1992–1993): Tokyo Institute of Technology, Japan

(2) Westhafer Award for Excellence in Research and Creativity (1999–2000): New
Mexico State University, USA
(3) Distinguished Lukacs Professor of Statistics (Spring 2002): Bowling Green
State University, Ohio, USA
(4) Distinguished Faculty Fellow (Summers 2002&2003): American Association
of Engineering Education/Summer Faculty Research Program, USA.

B. Selected Publication

1. A note on the extension principle for fuzzy sets. J. Math.Anal. and Appl. (64)
369–380 (1978)
2. On random sets and belief functions. J. Math. Anal. Appl. (65), 541–542 (1978)
3. Density estimation in a continuous-time, stationary markov process. Annals of
Statistics 7(2), 341–348 (1979)
4. Sur l’utilisation du temps local en statistique des processus (co-authored with T.
D. Pham). Comptes Rendus Acad. Sci. Paris A-290, 165–168 (1980)
5. Recursive estimation in diffusion model (co-authored with G. Banon). SIAM
J. Control and Optimization 19(5), 676–685 (1981)
6. Identification of non-stationary diffusion model by the method od sieves
(co-authored with T.D. Pham). SIAM J. Control and Optimization 20(5), 603–
611 (1982)
7. Estimation on change-point hazard rate model (co-authored with G.S. Rogers
and E.A. Walker). Biometrika 71(2), 299–304 (1984)
8. Strong consistency of maximum likelihood estimator in a change-point hazard
rate model (co-authored with T.D. Pham). J. Statistics 2(2), 203–216 (1990)
9. On the scoring approach to admissibility of uncertainty measures in expert
systems (co-authored with I.R. Goodman and G.S. Rogers). J. Math. Anal. and
Appl. 159(2), 550–594 (1991)
10. Intervals in boolean rings: approximation and logic. J. Foundations of
Computing and Decision Sciences 17(3), 131–138 (1992)
11. Bootstrapping the change-point in a hazard rate model (co-authored with T.D.
Pham). J. Inst. Statist. Math. 45(2) 331-340 (1993)
12. A history and introduction to the algebra of conditional events and probability
logic (co-authored with E.A. Walker). IEEE Trans. Man Systems and
Cyyernetics 24(2) 1671–1675 (1994)
Preface xi

13. Robust reasoning with rules that have exceptions (co-authored with D. Bamber
and I.R. Goodman). Annals of Math. and Artifical Intell.(45), 83–171 (2005)
14. Random and fuzzy sets in coarse data analysis (co-authored with B. Wu).
Comp.Statist. and Data Anal. (51) 70–85 (2006)
15. On some claims related to Choquet integral risk measures (co-authored with U.
Pham and H. Tran). Annals of Operations Research 195(1) 5–31 (2012)
16. Nonparametric estimation of a scalar diffusion model from discrete data: A
survey (co-authored with C. Gourieroux and S, Sriboonchitta). An- nals of
Operations Research 256(2), 203–219 (2017)

C. Books

1. Uncertainty Models for knowledge-Based Systems (co-authored with I.R.

Goodman). North Holland (1985)
2. Fundamentals of Mathematical Statistics Vol.I and Vol II (co-authored with G.
S. Rogers). Springer -Verlag University Text Book Series (1989)
3. Conditioning Inference and Logic for Intelligent Systems A Theory of
Measure-Free Conditioning (co-authored with I.R. Goodman and E.A.
Walker). North-Holland (1991)
4. Fundamental of Uncertainty Calculi with Applications to Fuzzy Inference
(co-authored with M. Grabisch and E.A. Walker). Kluwer Academic (1994)
5. Les Incertitudes dans les Systemes Intelligents (co-authored with B.
Bouchon-Meunier). Collection “Que Sais-Je?”, Presses Universitaires de
France (1996)
6. Mathematics of Data Fusion (co-authored with I.R. Goodman and R. Mahler).
Kluwer Academic (1997)
7. Applications of Continuous Mathematics to Computer Science (co-authored
with V. Kreinovich). Kluwer Academic (1997)
8. A First Course in Fuzzy and Neural Control (co-authored with N.R. Prasad, C.
L. Walker and E.A. Walker). Chapman and Hall/CRC Press (2000)
9. An Introduction to Random Sets. Chapman and Hall/CRC Press (2006)
10. Stochastic Dominance and Applications to Finance, Risk and Economics
(co-authored with S. Sriboonchitta, W.K. Wong and S. Dhompongsa).
Chapman and Hall/CRC Press (2009)
11. A First Course in Fuzzy Logic/Fourth Edition (co-authored with C.L. and E.A.
Walker). Chapman and Hall/CRC Press (2018)

El Paso, TX, USA Vladik Kreinovich

e-mail: vladik@utep.edu
Contents

An Enjoyable Research Journey on Uncertainty . . . . . . . . . . . . . . . . . . 1

Hung T. Nguyen
A Bayesian Dilemma . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 15
Donald Bamber
The Fell Compactification of a Poset . . . . . . . . . . . . . . . . . . . . . . . . . . . 31
G. Bezhanishvili and J. Harding
Is Time Fuzzy? . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 47
Bernadette Bouchon-Meunier
Hung Jury: The Verdict on Uncertainty . . . . . . . . . . . . . . . . . . . . . . . . 55
William M. Briggs
Applications of Type-2 Fuzzy Sets with Set Approximation
Approaches: A Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 67
Chia-Wen Chang and Chin-Wang Tao
A Remark on the Caristi’s Fixed Point Theorem and the Brouwer
Fixed Point Theorem . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 93
S. Dhompongsa and P. Kumam
Thick Sets, Multiple-Valued Mappings, and Possibility Theory . . . . . . . 101
Didier Dubois, Luc Jaulin, and Henri Prade
A Tacit Assumption Behind Lewis Triviality That Is Not Applicable
to Product Space Conditional Event Algebra . . . . . . . . . . . . . . . . . . . . . 111
I. R. Goodman and Donald Bamber
Arrondir le cercle... . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 127
Emmanuel Haven
Beyond p-Boxes and Interval-Valued Moments: Natural Next
Approximations to General Imprecise Probabilities . . . . . . . . . . . . . . . . 133
Olga Kosheleva and Vladik Kreinovich

xiii
xiv Contents

How to Reconcile Maximum Entropy Approach with Intuition:

E.g., Should Interval Uncertainty Be Represented by a Uniform
Distribution . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 145
Vladik Kreinovich, Olga Kosheleva, and Songsak Sriboonchitta
Time Series: How Unusual Local Behavior Can Be Recognized
Using Fuzzy Modeling Methods . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 157
Vilém Novák and Viktor Pavliska
Fuzzy-Based Methods in Data Analysis with the Focus
on Dimensionality Reduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 179
Irina Perfilieva
Perception and Reality . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 193
Nadipuram R. Prasad
The Falsificationist Foundation for Null Hypothesis
Significance Testing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 219
David Trafimow
A Wavelet Strategy to Buy Low and Sell High . . . . . . . . . . . . . . . . . . . 227
Lanh Tran
A Note on the Relation Between Liu’s Uncertain Measure
and Choquet Capacity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 237
Guo Wei, Zhiming Li, and Mei Cai
The Joint Distribution of the Discrete Random Set Vector
and Bivariate Coarsening at Random Models . . . . . . . . . . . . . . . . . . . . 247
Zheng Wei, Baokun Li, and Tonghui Wang
An Enjoyable Research Journey
on Uncertainty

Hung T. Nguyen

Abstract This is a story of research on uncertainty modeling from information

measures without probability to Choquet capacities, possibility measures, fuzzy
measures, imprecise probabilities, belief functions and finally, quantum probabil-
ity. The main part of the paper is devoted, almost entirely, to an invitation to quantum
probability for behavioral economics.

Keywords Behavioral economics · Decision theory · Quantum behavioral

economics · Quantum entropy · Quantum finance · Quantum mechanics ·
Quantum probability · Uncertainty modeling

1 Introduction

It is all about uncertainty in its various faces we face in various aspects of life, either
in natural or social sciences. Of course, we start out with the notion of chance as
uncertainty. Recognizing that chance can be measured, various “ideas” about chance
surfaced (see [7]) together with associated quantitative modeling of chance, called
probability (although “Probability is the most important concept in modern science,
especially as nobody has the slighest notion what it means”, Bertrand Russell, 1929).
And of course we talk about Kolmogorov’s book, written in 1933, which provides
the foundations for statistics. As mentioned in [17], Kolmogorov may write another
kind of book if he was aware of a book written just a year earlier ([36], but written in
German, in 1932) by von Neumann in which quantum probability was formulated,
since quantum probability is just “a bit” more general than “standard” probability
that Kolmogorov formulated in his 1933 book.

H. T. Nguyen (B)
Department of Mathematical Sciences, New Mexico State University,
Las Cruces, NM 88003, USA
e-mail: hunguyen@nmsu.edu
Faculty of Economics, Chiang Mai University, Chiang Mai, Thailand

© The Editor(s) (if applicable) and The Author(s), under exclusive license 1
to Springer Nature Switzerland AG 2021
V. Kreinovich (ed.), Statistical and Fuzzy Approaches to Data Processing, with Applications
to Econometrics and Other Areas, Studies in Computational Intelligence 892,
https://doi.org/10.1007/978-3-030-45619-1_1
2 H. T. Nguyen

Our journey will be from Kolmogorov to von Neumann (a strange one!). While the
main purpose of this paper is providing a friendly invitation to quantum probability for
applied econometricians (since other “introductions” to quantum probability are only
written in journals of physics, for physicists or mathematicians), we will, however,
trace back the journey we went through from Kolmogorov in the 1970s, passing by
various kinds of uncertainty modelings (e.g., capacities, fuzzy measures, and belief
functions), before, finally reaching quantum probability!
But let us elaborate a bit on why the jump from Kolmogorov to von Neumann?
As stated clearly in [18]: “The proposed modification of the probability calculus
appears more natural if we distinguish between probability calculus and probability
theory. With calculus we denote the mathematical formalism devoid of any interpre-
tation of this formalism. With theory we refer to the application of this calculus to
various situations involving the occurrences of observed phenomena”. This spirit is
the cause of various proposals concerning non-additive uncertainty measures since
von Neumann’s work (in 1944) on behavioral economics (more specifically in deci-
sion analysis). In the context of quantum mechanics, as Feynman [8] spelled out
clearly, while the meaning of probability is the same, nature computes probabilities
of observed events differently, i.e., its probability calculus is different than that of
Kolmogorov. A theory and calculus of probability is “suitable” for a given context
(e.g., a sub-atomic level, or social sciences) if the application of the calculus to that
context is “consistent” with observations, such as effectivity in predicting the prob-
abilities of actually occuring events. As we will spell out in Sect. 4, this is precisely
why we borrow quantum probability calculus from quantum mechanics to model
uncertainty in human decision-making.
At the end of the journey (!), we arrive at quantum probability (a noncommuta-
tive generalization of Kolmogorov probability theory), the main topic of this paper.
Considering it as a new topic, at least for applied econometricians (and not for proba-
bilists since they have [22]!), I will proceed, in subsequent sections, to elaborate, in a
“pedalogical” order, on the “what”, “why” and “How”. In doing so, we keep in mind
that while there are compelling reasons why quantum probability calculus should
replace Kolmogorov probability calculus in behavioral economics (as exposed in the
literature since many years), we still see little efforts from econometricians to apply
it to real-world problems. The reason for this phenomenon is this. Almost all efforts
on promoting quantum probability seem to be for “econophysics” (where physics is
the main theme), and not for Economics! For example, a nice article like [17] was
not published in a economic (or statistics) journal. Research efforts seem to flourish
outside of the economic community, such as [5, 16, 34, 37], while others seem not
attract attention of econometricians because of “unclear” mathematical presentations
(in fact, “complications deter”) such as [1, 3, 14, 32, 33], recalling that serious statis-
ticians and econometricians have no problems with Kolmogorov probability theory
as the foundation of statistical science! If quantum probability is to be useful and
appropriate outside of physics, e.g., in behavioral economics, the theory and calculus
of quantum probability should be made as simple as possible (but not simpler!), at
least at the introductory level. What seems to be really needed is an introductory text
in the form of “A First Course in Quantum Probability for Economists”!
An Enjoyable Research Journey on Uncertainty 3

In the sequent, first we elaborate on some “popular” uncertainty measures in

Sect. 2. The subsequent Sects. 3, 4, and 5 are devoted to an “invitation to quantum
probability for applied statisticians and econometricians”.

2 Some Uncertainty Measures

Measuring and computing probabilities (of uncertain events) provide information.

Without asking (like Joseph Kampe’ de Feriet) whether information comes before
or after probability, let’s concentrate on the theory and calculus of probability as
formulated by Kolmogorov in 1933. Among tremendous impacts on sciences from
this formalism, let’s focus on the birth of quantitative economics in 1944 with von
Neumann’s classical work on game theory and economic behavior, in which, under
the assumption of rationality of economic agents, Kolmogorov probability is used
to model the uncertainty of decision-makers. Of course that assumption is false, as
later, Stephen Hawking said [15]: “Economics is also an effective theory, based on
the notion of free will plus the assumption that people evaluate their possible alter-
native courses of action and choose the best. That effective theory is only moderately
successful in predicting behavior because, as we all know, decisions are often not
rational or are based on a defective analysis of the consequences of the choice. That
is why the world is in such a mess”. The argument for assuming rationality is this.
Rational behavior can be predicted, irrational behavior cannot!
In fact, if von Neumann’s work is about “economic behavior”, it should describe
economic agents’ behavior in making decisions under uncertainty. However, it is
surprizing that he used Kolmogorov probability as behavioral probability rather than
his own quantum probability. If he used quantum probability as behavioral probability
instead, based on the “analogy” of intrinsic randomness of particles’ motion with
human’s free will (rational or not), we could have then behavioral economics ever
since! As a technical note, the main ingredient in von Neumann’s decision theory is
the concept of expected utility, an integral operator which operates on functions which
form a commutative algebra, whereas in quantum probability, it is the trace operator,
replacing integral, which operates on linear operators (of a Hilbert space) which
forms a noncommutative algebra. As we will see, this extension from commutative
algebras to noncommutative algebras is essential for behavioral economics. In other
words, we could have the most appropriate behavioral economics without going
through all efforts since M. Allais in 1953, and the work [19] which opened the
doors to the application of empirical psychology to economics flourishing the actual
behavioral economics. However, such efforts are not in vain, since, as a “side effect”
(!), they revealed different useful uncertainty measures as we proceed to mention
next.
Kampe’ de Feriet’s information measures without probability led to a formal
connection with (Choquet) capacities in potential theory. By themselves, capacities
are not uncertainty measures per se. However, they can be used, as generalized
probability measures in, say, financial risk management. This can be seen as follows.
4 H. T. Nguyen

Since a financial (loss) variable X , defined on a probability space (Ω, A , P), can
be heavy-tailed, the notion of value-at-risk is more realistic to use rather than the
variance. If F(.) is the distribution function of X , then its value-at-risk, at level
α ∈ (0, 1) is its quantile: V a Rα (X ) = V a Rα (F) = F −1 (α). But,

∞
−1
F (α) = 1(1−α,1] (1 − F(t))dt =
0

∞ ∞
gα (1 − F(t))dt = (gα ◦ P)(X > t)dt
0 0

where gα (.) : [0, 1] → [0, 1], gα (x) = 1(1−α] (x). The non-additive set-function
ν(.) = gα ◦ P(.) : A → [0, 1] is non-decreasing (A ⊆ B =⇒ ν(A) ≤ ν(B)), and
ν(∅) = 0, ν(Ω) = 1 which  ∞ is called a capacity. Thus, V a Rα (F) has a Choquet
integral representation 0 ν(X > t)dt. Likewise, the coherent risk measure “tail
 1 −1
value-at-risk” T V a Rα (F) = 1−α 1
α F (t)dt has the Choquet integral representa-
tion with capacity ν(.) = gα ◦ P(.), with gα (x) = min{1, 1−α x
}.
In 1965, L. A. Zadeh considered a type of “uncertain evidence” (called fuzzy
data). In general, as spelled out in [7], an uncertain evidence is an observation which
could be imprecise, ambiguous, or fuzzy (i.e., no sharply defined boundaries). For
example, an uncertain evidence could be the realization of a random set (see [28]) as
a coarse data. Note that if we have an “evidence”, we are going to use it to make some
decisions (based on it), although, as we all know, an evidence alone is not sufficient
to reach a reasonable decision, exemplified by the use of P-value alone in frequentist
testing of hypotheses. The specific kind of uncertain evidence, so to speak, that Zadeh
considered is fuzziness in observations, mostly expressed in a natural language.
Viewing events as subsets of some set Ω, a fuzzy subset of Ω (a fuzzy event) is a
generalization of an ordinary (crisp) subset of Ω. Whereas an ordinary subset A ⊆ Ω
has members all of full membership (1), a fuzzy subset of Ω could contain elements
with partial membership (between 0 and 1). A direct extension is not possible. An
indirect one is “simple”: Each subset A is equivalent to its membership (indicator)
function 1 A in the sense that ω ∈ A ⇐⇒ 1 A (ω) = 1. Now 1 A (.) : Ω → {0, 1}. To
define fuzzy subsets of Ω, it suffices to extend the range of indicator functions
from{0, 1} to the whole unit interval [0, 1], so that a function μ(.) : Ω → [0, 1]
represents (or in fact, defines) a fuzzy subset of Ω with the interpretation that the
value μ(ω) ∈ [0, 1] is the membership degree (full or partial) of ω ∈ Ω in the fuzzy
set under consideration. With its membership function specified, a “fuzzy evidence”
provides some information. Clearly, fuzziness is different than randomness. The
membership function is not a probability distribution! Fuzziness is another kind of
uncertainty (lack of total knowledge). In fact, fuzziness and randomness can coexist.
An Enjoyable Research Journey on Uncertainty 5

At a practical level, how to “manipulate” fuzzy data? i.e., how to carry out compu-
tations with functions defined on fuzzy sets? It turns out that such computations can
be carried out by using interval mathematics, via a result in [23], known as Nguyen’s
theorem, see also [10, 11].
In the context of fuzzy theory (see e.g., [29]), there are various associated uncer-
tainty measures. Zadeh’s possibility measure [40] is a special capacity. As a gener-
alization of probability measure, a first attempt to define conditional possibility dis-
tribution was [24]. See also [4, 26]. More generally, Sugeno’s fuzzy measures were
considered for subjective evaluation problems. A fuzzy measure μ(.) is a capacity
with the interpretation that the value μ(A) of a crisp subset A ⊆ Ω, being a number
in [0, 1], is the subjective evaluation expressing, say, the degree of matching of the
observed A with some intended target. The measure is called “fuzzy” by analogy
with membership function of a fuzzy set (viewing the set-function μ(.) as a fuzzy
subset of 2Ω ).
With the interpretation that fuzziness in observed data is a type of “uncertain
evidence”, this is no wonder why Zadeh was immediately interested in Shafer’s
“mathematical theory of evidence” in 1976 [35], inspired from Demspter’s upper
and lower probabilities [6] (and robust Bayesian statistics, imprecise probabilities,
see e.g. [38]). The framework in which the so-called belief functions were developed
is this. Let U be a finite set, a function F(.) : 2U → [0, 1] is called a belief function
when it has the following properties: F(∅) = 0, F(U ) = 1, and, for k ≥ 2, and
A j , j = 1, 2, . . . , k subsets of U :

F(∪kj=1 A j ) ≥ (−1)|I |+1 F(∩i∈I Ai )
∅=I ⊆{1,2,...,k}

where |I | denotes the cardinality of the set I .

Note that if F(.) is a probability measure on 2U , then the above inequality is
an (H. Poincaré) equality, so that belief functions are simply slightly generalized
probability measures.
The intent of the theory was to provide another kind of uncertainty measure
different than randomness. It was quickly pointed out in [25] that it is not so, i.e.,
a belief function F(.) is, in fact, a bona fide probability distribution function, not
of a random variable, but of a random set (i.e., a random element whose values are
subsets of some set), so that randomness is around!
This can be seen simply as follows. Given a belief function F(.), consider the
set-function f (.) : 2U → [0, 1] defined by

f (A) = (−1)|A\B| F(B)
B⊆A

 f (.) is a bona fide probability density function on 2 , and we have F(A) =

U
then
B⊆A f (B), i.e., F(.) acts as the probability distribution function with density f (.).
More specifically, given F(.), there exists a random set S : (Ω, A , P) → 2U ,
such that F(A) = P(S ⊆ A). Note that in random set theory (see [28]), the partial
6 H. T. Nguyen

order ⊆ on subsets replaces ≤ on numbers. Thus, like fuzzy set theory which can be
studied via interval mathematics, belief function theory can be studied in the context
of random set theory (which is based on Kolmogorov probability theory).
In summary, we got quite a number of non-additive uncertainty measures, gener-
alizing Kolmogorov probability measures, say, by the 90s. In closing his text on the
theory of choice in 1988, Kreps [20] had this to say “These data provide a continu-
ing challenge to the theorist, a challenge to develop and adapt the standard models
so that they are more descriptive of what we see. It will be interesting to see what
will be in a course on choice theory in ten or twenty years time”. Now, over 30
years, what do we see as possible advances on choice theory? Since choice theory is
clearly part of “economic behavior”, and in view of the current flourishing approach
to economics, namely behavioral economics, we could answer Kreps as: An update
course on choice theory, for the 21st century, should contain quantum probability
since only quantum probability captures the noncommutativity of “observables” that
all the other uncertainty measures did not.
But how to formalize a noncommutative probability theory? Well, as an old say-
ing goes “There is nothing new under the sun”, we got it for free from quantum
mechanics!

3 What Is Quantum Probability

Upfront, quantum probability (borrowed from quantum mechanics, including the

term “quantum” which is used here simply to distinguish it from conventional notion
of probability) is simply a generalization of Kolmogorov probability theory. We
spell out next what to be “generalized” (whereas “why” such a generalization will
be explained in the next Sect. 4).
Following David Hilbert’s advice (“what is clear and easy to grasp attracts us,
complications deter”), here is a “clear and easy” exposition to grasp! Essentially,
quantum probability refers to a noncommutative generalization of Kolmogorov the-
ory. Specifically, random variables in Kolmogorov’s framework are functions so that
they are commutative with respect to multiplication. We wish to extend random vari-
ables to a noncommutative setting. For “easy to grasp”, it suffices to consider the
simplest case, namely a finite probability space (Ω, A , P) where Ω is a finite set,
say, Ω = {1, 2, . . . , n}, A is its power set, and P(.) : A → [0, 1] is a probability
measure with probability density function ρ(.) : Ω → [0, 1], ρ(ω) = P({ω}).
Objects related to (Ω, A , ρ) are functions f (.) : Ω → R, such as random vari-
ables, the probability density ρ(.), and events A ∈ A via their indicator functions
1 A (.) : Ω → {0, 1}.
We wish to extend the commutative algebra of functions to a noncommutative
algebra.
For those familiar with fuzzy set theory, remember how Zadeh generalized crisp
sets to fuzzy sets? A subset A of Ω cannot be extended to a fuzzy subset of Ω
directly. However, it can be if we do it indirectly, i.e., looking for an equivalent
An Enjoyable Research Journey on Uncertainty 7

representation which can be extended. A ⊆ Ω is equivalent to its indicator function

1 A (.) : Ω → {0, 1}. Extending the range {0, 1} to [0, 1] we obtain generalized mem-
bership functions which are used as definition of fuzzy subsets. We face the same
situation here in extending Kolmogorov probability space to quantum probability
space. We seek a feasible indirect way to achieve this.
First, when Ω = {1, 2, . . . , n}, such functions f (.) are identified with the trans-
pose vector ( f (1), f (2), . . . , f (n)) ∈ Rn . In turn, each vector X = (X (1), X (2),
. . . , X (n)) ∈ Rn is equivalent to the diagonal n × n matrix [X ] whose diagonal
terms are X (1), X (2), . . . , X (n). We have transformed functions to (special) matri-
ces. Now a n × n matrix is a linear map (called operator) from Rn to Rn .
If X is a random variable, then we say that X is “represented” by [X ]. In particular,
an event A ∈ A , is represented by [1 A ] whose diagonal terms are 0 or 1. Such
a diagonal matrix is of course symmetric, i.e., [1 A ] = [1 A ] , but also idempotent,
i.e., [1 A ]2 = [1 A ], so that [1 A ] is a (special) projection (operator) onto some closed
subspace of the (Hilbert) space Rn .
The probability density function ρ(.) is represented by the diagonal matrix [ρ]
whose diagonal terms are nonnegative and sum up to one (its trace is one). Such an
operator is “positive” in the sense that for any x ∈ Rn , the scalar product x, [ρ]x≥ 0
noting that arbitrary positive operators are symmetric.
All these diagonal matrices are of course symmetric: random variables “are” (diag-
onal) symmetric matrices, events are (special) projections, and probability densities
are (special) positive operators with unit trace. Note that in this equivalent repre-
sentation of standard (Kolmogorov) probabilistic objects, we obtain also equivalent
computational procedures such as: for A ∈ A ,

P(A) = ρ(ω) = tr ([ρ][1 A ])
ω∈A

where tr (.) denotes the trace operator operating on matrices, and


EX = X (ω)ρ(ω) = tr ([ρ][X ])
ω∈Ω

i.e., the trace operator replaces expectation operator (thus replacing “integral” in
continuous case).
As observations on random variables are necessary for predictions (say, in physics
or in psychological experiments to discover “laws”), as well as for statistics in general
(!), we see that, when a random variable X is represented by [X ], its range (i.e., its
possible values) is the set of its diagonal entries of [X ], which is the set of its
eigenvalues (its spectrum), recalling that the spectrum on any symmetric (or more
generally, self adjoint) matrix (or operator) lies in R (eigenvalues of self adjoint
operators are real-valued). These observations motivate what we continue now to
arrive as quantum counterpart of Kolmogorov probability.
Now diagonal matrices (as special case of symmetric matrices) form a commu-
tative subagebra of the noncommutative algebra of all arbitrary symmetric matrices.
8 H. T. Nguyen

It is this noncommutative algebra of (arbitrary) symmetric matrices which extends

Kolmogorov probability space to quantum probability space. Specifically, a finite
quantum probability space is a triple (Rn , P(Rn ), ρ), where P(Rn ) denotes the
space of all (orthogonal) projections on Rn (i.e., projections p onto closed subspaces
of Rn , characterized by p = p 2 = p ∗ (transpose)), and ρ(.) is a “density matrix”, i.e.,
an arbitrary positive operator with unit trace. Of course, any Kolmogorov probabil-
ity spaces are examples of quantum probability spaces! And of course, we leave
the general quantum probability spaces for “advanced readers” to consult texts such
as [30].
After spelling out “why quantum probability spaces are appropriate for behavioral
economics?” in the next Sect. 4, we will elaborate on quantum probability calculus
in Sect. 5.

4 Why Do We Need Quantum Probability in Economics?

Upfront, we are going to motivate (in fact, to explain) the need to use quantum proba-
bility as behavioral probability to model decision-making behavior in social sciences,
especially in economics which, in turn, will suggest that the actual flourishing field
of behavioral economics should be quantum behavioral economics.
In neoclassical economics (or more generally, in conventional social sciences) the
uncertainty involved (in human decision-making) is measured and manipulated by
standard probability (SP) of Kolmogorov. As stated in Sect. 2, various modifications
of SP were proposed based on empirical facts that indicated that SP cannot explain
certain behavior of decision-makers. This is consistent with the spirit of natural
sciences (e.g., physics) where a good model should be able to predict well what are
to be observed, exemplified by the replacement of classical mechanics by quantum
mechanics in the first quarter of the last century. There are empirical facts in human
decision context which call for further extension of SP to obtain a better behavioral
probability formalism (see e.g., [3, 19]). These empirical facts were known for quite
some time, but we reproduce them here to give a clear motivation. Specifically,
these are empirical facts related to the so-called “conjunction fallacy”, “disjunction
fallacy”, “order effect” and “preference reversal” which cannot be “explained” if SP
is used to model uncertainty, i.e., the associated empirical facts are not consistent
with SP. Just like in physics–when a model (e.g., a suggestive “law”) does not predict
well the observations, it should be not only abandoned, but also replaced by some
other model which can explain what we see–it seems natural to do the same thing in
social sciences, including economics.
Roughly speaking, experiments on human decision-making provide empirical
probabilities such as Pr(A&B) > Pr(A) (probability measures are not monotone
increasing), Pr(Aor B) = Pr(A) + Pr(B) (the additivity, or the “law of total prob-
ability” fails), Pr(A&B) = Pr(B& A) (the commutativity fails). Thus, since eco-
nomic decisions involve humans, these facts should ring the bell! Moreover, as far
An Enjoyable Research Journey on Uncertainty 9

as modeling of economic/financial data is concerned, should we take seriously the

investigation in [34]?:
A natural explanation of extreme irregularities in the evolution of prices in
financial markets is provided by quantum effects

And, as far as statistics with quantum probability is concerned, see e.g.,

[12, 27, 31].
It turns out that the mentioned fallacies are very similar to what physicists see in
quantum mechanics (which is intrinsically random), exemplified by the well-known
two-slit experiment (see, e.g., [8]), in which the observed probabilities do not obey
the calculus of SP. Since quantum mechanics is intrinsically random, physicists need
to figure out ways to compute probabilities of quantum events (e.g., measurements).
With the knowledge of the Schrodinger equation (counterpart of Newtonian law of
motion in classical mechanics), a “quantum probability calculus” was discovered
and applied successfully in explaining (and predicting) quantum phenomena, as we
know today. In fact, psychologists have shown that quantum probability theory (as
we outlined in Sect. 3, and will be elaborating more in the next Sect. 5) can explain the
mentioned fallacies. Moreover, it seems that the “order effect” could be a testimony
for the quantum nature of human judgement. All the above form the basis for consid-
ering seriously the use of quantum probability as behavioral probability, to improve
behavioral economics. In a sense, such a shift from neoclassical economics to quan-
tum behavioral economics could be compared with the extension from Newtonian
mechanics to quantum mechanics.

5 How to Use Quantum Probability in Economics?

Just like with Kolmogorov probability, we need to learn how to carry out quantum
probability calculus in order to investigate economic issues! However, we do not
have (yet) anything similar to standard probability calculus from which statistical
applications follow, although quantum probability is as old as Kolmogorov proba-
bility. Perhaps the reason is that most, if not all, writings on quantum probability are
for physics and cognitive science, rather for economics. Hopefully, with the flour-
ishing field of behavioral economics, and the aggressive push in favor of quantum
probability in it by a large community of researchers, we will see soon textbooks
on quantum probability calculus for economic applications. Until then, a flavor of it
could be useful as a starting point.
Recall that standard probability calculus is built on a probability space (Ω, A , P).
Similarly, quantum probability calculus is based upon a quantum probability space
(H, P(H ), ρ) where H is a separable, complex Hilbert space, P(H ) is the set
of all projectors on H , and ρ is a positive operator, with unit trace, on H . To
be simple (avoiding unnecessary mathematical complications at this stage), as in
10 H. T. Nguyen

Sect. 2, it suffices to consider a finite probability space (Ω, A , P), with, say,
Ω = {1, 2, . . . , n}, so that its extension to quantum probability space (H, P(H ), ρ)
is simply (Rn , P(Rn ), ρ).
A random variable X (called also an “observable”) is “represented” by a symmetric
matrix [X ] (diagonal or not) whose spectrum σ (X ) (set of its real eigenvalues) is the
range of X .
As such, we need to specify its probability distribution on σ (X ) or σ ([X ]). For this,
we will rely upon the spectral decomposition of symmetric matrices. Let σ ([X ]) =
{λ1 , λ2 , . . . , λn } ⊆ R, and u 1 , u 2 , . . . , u n (in Rn ) be its corresponding eigenvectors.
Let P j = u j u ∗j (the “outer product” on the finitely dimensional Hilbert space Rn ,
where u ∗j is the transpose on u j ). Each P j is an operator (a matrix) on Rn , orthogonal
projection onto the closed subspace of Rn spaned by the eigenvector u j , which is
in fact a (orthogonal) projector since P j = P j2 = P j∗ . The spectral decomposition of
[X ] is this.
⎡ ⎤⎡ ∗⎤
λ1 u1  n n
[X ] = [u 1 , u 2 , . . . , u n ] ⎣ . ⎦ ⎣ . ⎦ = λ j u j u ∗j = λ j Pj
λn u ∗n j=1 j=1

noting that this decomposition is the counterpart of the “decomposition” of a ran-

dom variable (in classical setting) as a linear combination of indicator (elementary)
variables.
As stated in Sect. 3, quantum events, i.e., A ⊆ Ω, are represented by projectors
[1 A ], so that P(A) in (Rn , P(Rn ), ρ) means tr (ρ[1 A ]), noting that the trace operator
tr (.) is commutative. More generally, the expected value, under ρ (a density matrix,
i.e., a positive operator with unit trace) of an observable [X ] is E ρ [X ] = tr (ρ[X ]).
Note however that quantum events, as projectors, might not be commutative in gen-
eral: for p, q ∈ P(Rn ), pq is an event (i.e., a projector) if and only if they commute
since ( pq)∗ = pq = q ∗ p ∗ = qp, so that the problem of “joint distribution” arises
(see later). 
Now for B ∈ B(R), the event (X ∈ B) is the projector ζ (B) = λ∈B Pλ , where
Pλ is the event “X takes the value λ”. This set-function ζ (.) : B(R) → P(Rn )
takes values in P(Rn ), so that it is a “projection-valued” mapping. In fact, it is
a projection-valued measure, called a spectral measure. With the above spectral
decomposition of a symmetric matrix [X ], we have a spectral measure associated
with each observable X , denoted as ζ X (.). Thus, the probability distribution of [X ]
on (R, B(R)) is μ X (.) = tr (ρζ X (.)), where μ X (B) = tr (ρζ X (B)) is the probability
that X takes a value in B, when the density matrix is ρ.
Remark. In general, i.e., for general quantum probability spaces, μ X (.) plays the
role of a probability distribution for the observable X , so that
 
E ρ (X ) = E ρ ([X ]) = λμ X (dλ) = λtr (ρζ X (dλ) = tr (ρ[X ])
R R
An Enjoyable Research Journey on Uncertainty 11


Indeed, ρ = j ai (u j u ∗j ) for any orthonormal
 basis {u j } of an arbitrary, separa-
ble, complex Hilbert space H , with a j > 0, j ai = 1. For B ∈ B(R), we have

 
μρ,X (B) = tr (ρζ X (B)) = ρζ X (B)u j , u j  = a j ζ X (B)u j , u j 
j j


Thus, since [X ] = xζ X (d x), the spectral decomposition of [X ], we have
  
E ρ (X ) = xμρ,x (d x) = a j xζ X (d x)u j , u j  =
j R

 
a j [X ]u j , u j  = a j ρ[X ]u j , u j  = tr (ρ[X ])
j j

When the density matrix ρ is induced by a

normalized “wave function” ψ ∈ H
(in the Schrodinger’s equation) on R3 as ρ = j a j |u j u j |, with a j = |c j |2 , c j =
ψ, u j , we have

E ψ (X ) = ψ ∗ (x)[X ]ψ(x)d x = [X ]ψ, ψ
R3

We use Dirac’s notations in the above analysis just to have an occasion to explain
now what they are and help the readers when they read the literature!
The case where H = Rn makes things clear. For x ∈ H , we rewrite it as |x
(called a “ket”). It is a “column vector”, such as x = (x1 , x2 , . . . , xn )∗ , whereas we
write < x| (a “bra”) to designate an element of the dual of H , i.e., a map from H
to R, so that x|(y) = x|(|y >) = x|y is the “inner (scalar) product” on H , such
as x|y = x, y (getting a “bracket”) = nj=1 x j yi . On the other hand, the “outer
product” on H is a “multiplication” of a column vector x ∈ Rn with a “row vector”
y (i.e., the transpose y ∗ of the column vector y ∈ Rn ) resulting in a matrix (i.e., an
operator on Rn ) is denoted as |xy| which acts on Rn as, for z ∈ Rn , (|xy|)(z) =
y, zx ∈ Rn . Of course, on Rn , |xy| is simply x y ∗ (multiplication of two matrices:
x is a n × 1 matrix, and y ∗ is a 1 × n matrix).
With respect to obtaining the amplitude |ψ|2 of the wave function ψ in the
Schrodinger’s equation, a “practical” method is using Feynman’s path integral (see
[1, 9]).
Recall that the standard interpretation of quantum mechanics is this. The dynamics
of a particle with mass m is “described” by a wave function ψ(x, t), where x ∈ R3
is the particle position at time t, which is the solution of the Schrodinger’s equation
(counterpart of Newton’s law of motion of macroobjects):

∂ψ(x, t) h2
ih =− Δx ψ(x, t) + V (x)ψ(x, t)
∂t 2m
12 H. T. Nguyen

As such, particles in motion do not have trajectories (in their phase space), or put
it more specifically, their motion cannot be described (mathematically) by trajecto-
ries (because of the Heisenberg’s uncertainty principle). The probability amplitude
|ψ(x, t)|2 is used to make probabilistic statements about the particle motion. There
is another, not “official”, interpretation of quantum mechanics in which it is possi-
ble to consider trajectories for particles, called Bohmian mechanics. This mechanics
formulation is suitable to use as models for, say, financial data (where time series
data are like “trajectories” of moving objects). In fact, they could be considered as
“useful models” in G. Box’s sense since they display an extra term which can be
used to represent the missing human factor.
Now as the “density matrix” ρ on a general quantum probability space plays
the role of an ordinary probability density function f (whose ordinary entropy is
− f (x) log f (x)d x), its quantum entropy (as defined by von Neumann [36]) is
−tr (ρ log ρ). As maximum (Kolmogorov) entropy principle provides equilibrium
models in statistical mechanics or other stochastic systems, it also enters financial
econometrics as the most diversified portfolio selection, see e.g. [2, 13].
Let’s specify von Neunman’s quantum entropy in a simple case, e.g., when H =
Cn .
For a density matrix ρ (extension of a probability density function) on
(Cn , P(Cn )), ρ log ρ is a n × n self adjoint matrix (operator) which is defined
as follows (by using spectral theorem). The spectral theorem says this. Since ρ is a
self adjoint operator on Cn , there exists an orthonormal basis of Cn , {u 1 , u 2 , . . . , u n }
consisting of eigenvectors of ρ, with associated eigenvalues {λ1 , λ2 , . . . , λn } (the
spectrum of ρ).  If we let P j be the projector onto the closed subspace spanned by
u j , then ρ = nj=1 λ j P j .
For g(.): R → R, g(x) = x log x, the (self adjoint) operator g(ρ) = ρ log ρ is
defined by nj=1 g(λ j )P j whose trace is


n 
n
tr (ρ log ρ) = u j , (ρ log ρ)u j  = λ j log λ j
j=1 j=1


so that the quantum entropy of ρ is −tr (ρ log ρ) = − nj=1 λ j log λ j which depends
only on the eigenvalues of ρ.
One more thing we like to mention about quantum probability calculus is the
noncommutativity of observables. When events p, q ∈ P(H ) do not commute, pq is
not an event, so that the problem of “joint distribution” of non commuting observables
X and Y , arises. By considering them separately, we can, in principle get their
“marginal distributions”. In classical SP, models for joint distribution of (X, Y ) can
be obtained by using copulas. The situation in quantum probability theory is different:
there is no notion of copulas in it. For this issue, see e.g., [21, 39].
An Enjoyable Research Journey on Uncertainty 13

References

1. B.E. Baaquie, Quantum Finance: Path Integrals and Hamiltonians for Options and Interest
Rates (Cambridge University Press, Cambridge, 2007)
2. A. Bera, S. Park, Optimal portfolio diversification using maximum entropy principle. Econ.
Rev. 27, 484–512 (2008)
3. J.R. Busemeyer, P.D. Bruza, Quantum Models of Cognition and Decision (Cambridge Univer-
sity Press, Cambridge, 2012)
4. L. Coroianu, R. Fuller, Nguyen type theorem for extension principle based on a joint possibility
distribution. Intern. J. Approx. Reason. 95, 22–35 (2018)
5. P. Darbyshire, Quantum physics meets classical finance. Phys. World 25–29 (2005)
6. A. Dempster, Upper and lower probabilities induced by a multivalued mapping. Ann. Math.
Statist. 38, 325–339 (1967)
7. P. Diaconis, B. Skyrms, Ten Great Ideas About Chance (Princeton University Press, Princeton
and Oxford, 2018)
8. R. Feynman, The concept of probability in quantum mechanics, in Berkeley Symposium on
Mathematical Statistics (1951), pp. 533–541
9. R. Feynman, A. Hibbs, Quantum Mechanics and Path Integrals (Dover, New York, 1965)
10. R. Fuller, T. Keresztfalvi, On generalization of Nguyen’s theorem. Fuzzy Sets Syst. 41(3),
371–374 (1991)
11. R. Fuller, On generalization of Nguyen’s theorem: a short survey of recent developments. Adv.
Soft Comput. Intell. Robot. Control., 183–190 (2014)
12. A. Gelman, M. Bethancourt, Does quantum uncertainty have a place in everyday applied
statistics. Behav. Brain Sci. 36(3), 285 (2013)
13. A. Golan, G. Judge, D. Miller, Maximum Entropy Econometrics (Wiley, New York, 1996)
14. E. Haven, A. Khrennikov, Quantum Social Science (Cambridge University Press, Cambridge,
2013)
15. S. Hawking, L. Mlodinow, The Grand Design (Bantam Books, London, 2010)
16. B. Herzog, Quantum models of decision-making in economics. J. Quantum Inf. Sci. 5, 1–5
(2015)
17. R. Hudson, A short walk in quantum probability. Phil. Trans. R. Soc. A 376, 20170226 (2017)
18. J.M. Jauch, The quantum probability calculus. Synthese 29, 131–154 (1974)
19. D. Kahneman, A. Tversky, Prospect theory: an analysis of decision under risk. Econometrica
47, 263–292 (1979)
20. D.M. Kreps, Notes on the Theory of Choice (Westview Press, 1988)
21. J.M. Malley, A. Fletcher, Joint distribution and quantum nonlocal models. Axioms 3, 166–176
(2014)
22. P.A. Meyer, Quantum Probability for Probabilists (Springer, Berlin, 1995)
23. H.T. Nguyen, A note on the extension principle for fuzzy sets. J. Math. Anal. Appl. 64, 369–380
(1978)
24. H.T. Nguyen, On conditional possibility distributions. Fuzzy Sets Syst. 1(4), 299–309 (1978)
25. H.T. Nguyen, On random sets and belief functions. J. Math. Anal. Appl. 65, 531–542 (1978)
26. H.T. Nguyen, On the entropy of random sets and possibility distributions, in The Analysis of
Fuzzy Information, ed. by J. Bezdek (CRC Press, 1987), pp. 45–156
27. H.T. Nguyen, Toward improving models for decision making in economics. Asian J. Econ.
Bank. 3(01), 1–19 (2019)
28. H.T. Nguyen, An Introduction to Random Sets (Chapman and Hall/CRC Press, Boca Raton,
2006)
29. H.T. Nguyen, C.L. Walker, E.A. Walker, A First Course in Fuzzy Logic, 4th edn. (Chapman
and Hall/CRC Press, Boca Raton, 2019)
30. K.R. Parthasarathy, An Introduction to Quantum Stochastic Calculus (Springer, Basel, 1992)
31. L. Pasca, A critical review of the main approaches on financial market dynamics modeling. J.
Heterodox Econ. 2(2), 151–167 (2015)
14 H. T. Nguyen

32. S. Patra, A quantum framework for economic science: new directions. Economics 2019–20
(2019)
33. E.W. Piotrowski, J. Sladkowski, Quantum games in finance. Quant. Financ. 4(6), 61–67 (2004)
34. S. Segal, I.E. Segal, The Black-Scholes pricing formula in the quantum context. Proc. Nat.
Acad. Sci. USA 95, 4072–4075 (1998)
35. G. Shafer, A Mathematical Theory of Evidence (Princeton University, University Press, Prince-
ton, 1976)
36. J. von Neumann, Mathematical Foundations of Quantum Mechanics, New edn. (Princeton
University Press, Princeton, 2018)
37. V. Vukotic, Quantum economics. Panoeconomicus 2, 267–276 (2011)
38. P. Walley, Statistical Reasoning with Imprecise Probabilities (Chapman and Hall, London,
1991)
39. V.I. Yukalov, D. Sornette, Quantum probabilities of composite events in quantum measurements
with multimode states. Lazer Phys. 23, 105502 (2013)
40. L.A. Zadeh, Fuzzy sets as a basis for a theory of possibility. Fuzzy Sets Syst. 1, 3–28 (1978)
A Bayesian Dilemma

Donald Bamber

Abstract An example is given in which a Bayesian reasoner fails to learn the obvi-
ous. That failure calls into question whether Bayesian epistemology is complete and
correct as it stands.

Keywords Bayesian epistemology · Bayesian reasoning · Intuitive reasoning ·

Rationality · Degree of belief · Epistemic Bayesian · Pragmatic Bayesian

1 Intuitive Rationality and Bayesian Rationality

Usually the conclusions that people arrive at intuitively agree approximately with
conclusions arrived at by Bayesian reasoning.1 Thus, intuitive rationality and
Bayesian rationality tend to agree.
However, there are circumstances where conclusions that are judged rational by
intuitive standards are judged grossly irrational by Bayesian standards and, con-
versely, conclusions judged rational by Bayesian standards are judged grossly irra-
tional by intuitive standards.
For example, suppose that a large number N trials of an experiment have been
run. On each trial, it is observed whether a particular event A occurred or not. Across
all N trials, the event A never occurred. After observing such empirical results, most
intuitive reasoners would assert that they had only a very small degree of belief—
close to zero—that A would occur on Trial N + 1. Moreover, they would regard it to

D. Bamber (B)
Cognitive Sciences Department, University of California, Irvine, USA
e-mail: dbamber@uci.edu
1 By “Bayesian reasoning” is meant the type of reasoning advocated by epistemic Bayesians, that
is by adherents of Bayesian epistemology. (As discussed in [3], epistemic Bayesians are much
different from pragmatic Bayesians.) In Bayesian epistemology, one should update one’s beliefs
by conditioning them on observations. For more detailed information, see [8, Chap. 2] and [10,
Chap. 4], the latter book having been reviewed by me in [2]. For a briefer, less complete account,
see my paper [3].

© The Editor(s) (if applicable) and The Author(s), under exclusive license 15
to Springer Nature Switzerland AG 2021
V. Kreinovich (ed.), Statistical and Fuzzy Approaches to Data Processing, with Applications
to Econometrics and Other Areas, Studies in Computational Intelligence 892,
https://doi.org/10.1007/978-3-030-45619-1_2
16 D. Bamber

be grossly irrational to have a degree of belief close to one-half that A would occur
on Trial N + 1.
However, there are circumstances where, after observing no occurrences of A in
N trials (N large), a Bayesian reasoner would have degree of belief around one-half
that A would occur on Trial N + 1. Moreover, the Bayesian reasoner would regard
as grossly irrational a near-zero belief that A would occur on Trial N + 1.
How can this happen? The next section will explain.

2 A Case of Counterintuitive Bayesian Reasoning

2.1 The Robotic Black Box

Suppose that B.R., a Bayesian reasoner, happens upon a black box. Although nothing
inside the box can be seen, an informant provides the following information:
• On the exterior of the black box, there is pushbutton and two side-by-side display
windows.
• Inside the black box, there is a robot having one arm, a hand, and an eye.
• On the floor of the black box lie a gold coin and a silver coin.
• When the pushbutton on the outside of the box is pressed, the robot picks up the
two coins in its hand and tosses them in the air so that they tumble about as they
fall.
• After the coins land on the floor, the robot observes whether the gold coin landed
heads or tails. Then the robot puts either the printed word “heads” or the word
“tails”, as appropriate, in the left display window.
• Next the robot examines the silver coin on the floor and puts either the printed
word “heads” or the word “tails”, as appropriate, in the right display window.

2.2 Two-Coins Model

B.R. decides to perform an experiment on the black box. A single trial of the exper-
iment will consist of pressing the pushbutton on the black box and observing the
results in the display window. The experiment will consist of running many such
trials.
After hearing the informant’s description of the inner workings of the black box,
B.R formulates his/her beliefs about the results of the experiment. To describe those
beliefs, we will need some notation.
Another random document with
no related content on Scribd:
marry? For him it would be all very well, but what about her? Mrs.
Burgess, a good sort and no chatterbox, in whom he had confided,
thought this absence of his in England, ostensibly to see lawyers
might serve to make Daisy reconsider, think what it meant. It was a
question of her position, Mrs. Burgess said; the social barrier; giving
up her children. She’d be a widow with a past one of these days,
draggling about in the suburbs, or more likely, indiscriminate (you
know, she said, what such women get like, with too much paint). But
Peter Walsh pooh-poohed all that. He didn’t mean to die yet.
Anyhow she must settle for herself; judge for herself, he thought,
padding about the room in his socks smoothing out his dress-shirt,
for he might go to Clarissa’s party, or he might go to one of the Halls,
or he might settle in and read an absorbing book written by a man he
used to know at Oxford. And if he did retire, that’s what he’d do—
write books. He would go to Oxford and poke about in the Bodleian.
Vainly the dark, adorably pretty girl ran to the end of the terrace;
vainly waved her hand; vainly cried she didn’t care a straw what
people said. There he was, the man she thought the world of, the
perfect gentleman, the fascinating, the distinguished (and his age
made not the least difference to her), padding about a room in an
hotel in Bloomsbury, shaving, washing, continuing, as he took up
cans, put down razors, to poke about in the Bodleian, and get at the
truth about one or two little matters that interested him. And he would
have a chat with whoever it might be, and so come to disregard
more and more precise hours for lunch, and miss engagements, and
when Daisy asked him, as she would, for a kiss, a scene, fail to
come up to the scratch (though he was genuinely devoted to her)—
in short it might be happier, as Mrs. Burgess said, that she should
forget him, or merely remember him as he was in August 1922, like a
figure standing at the cross roads at dusk, which grows more and
more remote as the dog-cart spins away, carrying her securely
fastened to the back seat, though her arms are outstretched, and as
she sees the figure dwindle and disappear still she cries out how she
would do anything in the world, anything, anything, anything....
He never knew what people thought. It became more and more
difficult for him to concentrate. He became absorbed; he became
busied with his own concerns; now surly, now gay; dependent on
women, absent-minded, moody, less and less able (so he thought as
he shaved) to understand why Clarissa couldn’t simply find them a
lodging and be nice to Daisy; introduce her. And then he could just—
just do what? just haunt and hover (he was at the moment actually
engaged in sorting out various keys, papers), swoop and taste, be
alone, in short, sufficient to himself; and yet nobody of course was
more dependent upon others (he buttoned his waistcoat); it had
been his undoing. He could not keep out of smoking-rooms, liked
colonels, liked golf, liked bridge, and above all women’s society, and
the fineness of their companionship, and their faithfulness and
audacity and greatness in loving which though it had its drawbacks
seemed to him (and the dark, adorably pretty face was on top of the
envelopes) so wholly admirable, so splendid a flower to grow on the
crest of human life, and yet he could not come up to the scratch,
being always apt to see round things (Clarissa had sapped
something in him permanently), and to tire very easily of mute
devotion and to want variety in love, though it would make him
furious if Daisy loved anybody else, furious! for he was jealous,
uncontrollably jealous by temperament. He suffered tortures! But
where was his knife; his watch; his seals, his note-case, and
Clarissa’s letter which he would not read again but liked to think of,
and Daisy’s photograph? And now for dinner.
They were eating.
Sitting at little tables round vases, dressed or not dressed, with their
shawls and bags laid beside them, with their air of false composure,
for they were not used to so many courses at dinner, and
confidence, for they were able to pay for it, and strain, for they had
been running about London all day shopping, sightseeing; and their
natural curiosity, for they looked round and up as the nice-looking
gentleman in horn-rimmed spectacles came in, and their good
nature, for they would have been glad to do any little service, such
as lend a time-table or impart useful information, and their desire,
pulsing in them, tugging at them subterraneously, somehow to
establish connections if it were only a birthplace (Liverpool, for
example) in common or friends of the same name; with their furtive
glances, odd silences, and sudden withdrawals into family jocularity
and isolation; there they sat eating dinner when Mr. Walsh came in
and took his seat at a little table by the curtain.
It was not that he said anything, for being solitary he could only
address himself to the waiter; it was his way of looking at the menu,
of pointing his forefinger to a particular wine, of hitching himself up to
the table, of addressing himself seriously, not gluttonously to dinner,
that won him their respect; which, having to remain unexpressed for
the greater part of the meal, flared up at the table where the Morrises
sat when Mr. Walsh was heard to say at the end of the meal, “Bartlett
pears.” Why he should have spoken so moderately yet firmly, with
the air of a disciplinarian well within his rights which are founded
upon justice, neither young Charles Morris, nor old Charles, neither
Miss Elaine nor Mrs. Morris knew. But when he said, “Bartlett pears,”
sitting alone at his table, they felt that he counted on their support in
some lawful demand; was champion of a cause which immediately
became their own, so that their eyes met his eyes sympathetically,
and when they all reached the smoking-room simultaneously, a little
talk between them became inevitable.
It was not very profound—only to the effect that London was
crowded; had changed in thirty years; that Mr. Morris preferred
Liverpool; that Mrs. Morris had been to the Westminster flower-show,
and that they had all seen the Prince of Wales. Yet, thought Peter
Walsh, no family in the world can compare with the Morrises; none
whatever; and their relations to each other are perfect, and they
don’t care a hang for the upper classes, and they like what they like,
and Elaine is training for the family business, and the boy has won a
scholarship at Leeds, and the old lady (who is about his own age)
has three more children at home; and they have two motor cars, but
Mr. Morris still mends the boots on Sunday: it is superb, it is
absolutely superb, thought Peter Walsh, swaying a little backwards
and forwards with his liqueur glass in his hand among the hairy red
chairs and ash-trays, feeling very well pleased with himself, for the
Morrises liked him. Yes, they liked a man who said, “Bartlett pears.”
They liked him, he felt.
He would go to Clarissa’s party. (The Morrises moved off; but they
would meet again.) He would go to Clarissa’s party, because he
wanted to ask Richard what they were doing in India—the
conservative duffers. And what’s being acted? And music.... Oh yes,
and mere gossip.
For this is the truth about our soul, he thought, our self, who fish-like
inhabits deep seas and plies among obscurities threading her way
between the boles of giant weeds, over sun-flickered spaces and on
and on into gloom, cold, deep, inscrutable; suddenly she shoots to
the surface and sports on the wind-wrinkled waves; that is, has a
positive need to brush, scrape, kindle herself, gossiping. What did
the Government mean—Richard Dalloway would know—to do about
India?
Since it was a very hot night and the paper boys went by with
placards proclaiming in huge red letters that there was a heat-wave,
wicker chairs were placed on the hotel steps and there, sipping,
smoking, detached gentlemen sat. Peter Walsh sat there. One might
fancy that day, the London day, was just beginning. Like a woman
who had slipped off her print dress and white apron to array herself
in blue and pearls, the day changed, put off stuff, took gauze,
changed to evening, and with the same sigh of exhilaration that a
woman breathes, tumbling petticoats on the floor, it too shed dust,
heat, colour; the traffic thinned; motor cars, tinkling, darting,
succeeded the lumber of vans; and here and there among the thick
foliage of the squares an intense light hung. I resign, the evening
seemed to say, as it paled and faded above the battlements and
prominences, moulded, pointed, of hotel, flat, and block of shops, I
fade, she was beginning, I disappear, but London would have none
of it, and rushed her bayonets into the sky, pinioned her, constrained
her to partnership in her revelry.
For the great revolution of Mr. Willett’s summer time had taken place
since Peter Walsh’s last visit to England. The prolonged evening was
new to him. It was inspiriting, rather. For as the young people went
by with their despatch-boxes, awfully glad to be free, proud too,
dumbly, of stepping this famous pavement, joy of a kind, cheap,
tinselly, if you like, but all the same rapture, flushed their faces. They
dressed well too; pink stockings; pretty shoes. They would now have
two hours at the pictures. It sharpened, it refined them, the yellow-
blue evening light; and on the leaves in the square shone lurid, livid
—they looked as if dipped in sea water—the foliage of a submerged
city. He was astonished by the beauty; it was encouraging too, for
where the returned Anglo-Indian sat by rights (he knew crowds of
them) in the Oriental Club biliously summing up the ruin of the world,
here was he, as young as ever; envying young people their summer
time and the rest of it, and more than suspecting from the words of a
girl, from a housemaid’s laughter—intangible things you couldn’t lay
your hands on—that shift in the whole pyramidal accumulation which
in his youth had seemed immovable. On top of them it had pressed;
weighed them down, the women especially, like those flowers
Clarissa’s Aunt Helena used to press between sheets of grey
blotting-paper with Littré’s dictionary on top, sitting under the lamp
after dinner. She was dead now. He had heard of her, from Clarissa,
losing the sight of one eye. It seemed so fitting—one of nature’s
masterpieces—that old Miss Parry should turn to glass. She would
die like some bird in a frost gripping her perch. She belonged to a
different age, but being so entire, so complete, would always stand
up on the horizon, stone-white, eminent, like a lighthouse marking
some past stage on this adventurous, long, long voyage, this
interminable (he felt for a copper to buy a paper and read about
Surrey and Yorkshire—he had held out that copper millions of times.
Surrey was all out once more)—this interminable life. But cricket was
no mere game. Cricket was important. He could never help reading
about cricket. He read the scores in the stop press first, then how it
was a hot day; then about a murder case. Having done things
millions of times enriched them, though it might be said to take the
surface off. The past enriched, and experience, and having cared for
one or two people, and so having acquired the power which the
young lack, of cutting short, doing what one likes, not caring a rap
what people say and coming and going without any very great
expectations (he left his paper on the table and moved off), which
however (and he looked for his hat and coat) was not altogether true
of him, not to-night, for here he was starting to go to a party, at his
age, with the belief upon him that he was about to have an
experience. But what?
Beauty anyhow. Not the crude beauty of the eye. It was not beauty
pure and simple—Bedford Place leading into Russell Square. It was
straightness and emptiness of course; the symmetry of a corridor;
but it was also windows lit up, a piano, a gramophone sounding; a
sense of pleasure-making hidden, but now and again emerging
when, through the uncurtained window, the window left open, one
saw parties sitting over tables, young people slowly circling,
conversations between men and women, maids idly looking out (a
strange comment theirs, when work was done), stockings drying on
top ledges, a parrot, a few plants. Absorbing, mysterious, of infinite
richness, this life. And in the large square where the cabs shot and
swerved so quick, there were loitering couples, dallying, embracing,
shrunk up under the shower of a tree; that was moving; so silent, so
absorbed, that one passed, discreetly, timidly, as if in the presence of
some sacred ceremony to interrupt which would have been impious.
That was interesting. And so on into the flare and glare.
His light overcoat blew open, he stepped with indescribable
idiosyncrasy, leant a little forward, tripped, with his hands behind his
back and his eyes still a little hawklike; he tripped through London,
towards Westminster, observing.
Was everybody dining out, then? Doors were being opened here by
a footman to let issue a high-stepping old dame, in buckled shoes,
with three purple ostrich feathers in her hair. Doors were being
opened for ladies wrapped like mummies in shawls with bright
flowers on them, ladies with bare heads. And in respectable quarters
with stucco pillars through small front gardens lightly swathed with
combs in their hair (having run up to see the children), women came;
men waited for them, with their coats blowing open, and the motor
started. Everybody was going out. What with these doors being
opened, and the descent and the start, it seemed as if the whole of
London were embarking in little boats moored to the bank, tossing
on the waters, as if the whole place were floating off in carnival. And
Whitehall was skated over, silver beaten as it was, skated over by
spiders, and there was a sense of midges round the arc lamps; it
was so hot that people stood about talking. And here in Westminster
was a retired Judge, presumably, sitting four square at his house
door dressed all in white. An Anglo-Indian presumably.
And here a shindy of brawling women, drunken women; here only a
policeman and looming houses, high houses, domed houses,
churches, parliaments, and the hoot of a steamer on the river, a
hollow misty cry. But it was her street, this, Clarissa’s; cabs were
rushing round the corner, like water round the piers of a bridge,
drawn together, it seemed to him because they bore people going to
her party, Clarissa’s party.
The cold stream of visual impressions failed him now as if the eye
were a cup that overflowed and let the rest run down its china walls
unrecorded. The brain must wake now. The body must contract now,
entering the house, the lighted house, where the door stood open,
where the motor cars were standing, and bright women descending:
the soul must brave itself to endure. He opened the big blade of his
pocket-knife.

Lucy came running full tilt downstairs, having just nipped in to the
drawing-room to smooth a cover, to straighten a chair, to pause a
moment and feel whoever came in must think how clean, how bright,
how beautifully cared for, when they saw the beautiful silver, the
brass fire-irons, the new chair-covers, and the curtains of yellow
chintz: she appraised each; heard a roar of voices; people already
coming up from dinner; she must fly!
The Prime Minister was coming, Agnes said: so she had heard them
say in the dining-room, she said, coming in with a tray of glasses.
Did it matter, did it matter in the least, one Prime Minister more or
less? It made no difference at this hour of the night to Mrs. Walker
among the plates, saucepans, cullenders, frying-pans, chicken in
aspic, ice-cream freezers, pared crusts of bread, lemons, soup
tureens, and pudding basins which, however hard they washed up in
the scullery seemed to be all on top of her, on the kitchen table, on
chairs, while the fire blared and roared, the electric lights glared, and
still supper had to be laid. All she felt was, one Prime Minister more
or less made not a scrap of difference to Mrs. Walker.
The ladies were going upstairs already, said Lucy; the ladies were
going up, one by one, Mrs. Dalloway walking last and almost always
sending back some message to the kitchen, “My love to Mrs.
Walker,” that was it one night. Next morning they would go over the
dishes—the soup, the salmon; the salmon, Mrs. Walker knew, as
usual underdone, for she always got nervous about the pudding and
left it to Jenny; so it happened, the salmon was always underdone.
But some lady with fair hair and silver ornaments had said, Lucy
said, about the entrée, was it really made at home? But it was the
salmon that bothered Mrs. Walker, as she spun the plates round and
round, and pulled in dampers and pulled out dampers; and there
came a burst of laughter from the dining-room; a voice speaking;
then another burst of laughter—the gentlemen enjoying themselves
when the ladies had gone. The tokay, said Lucy running in. Mr.
Dalloway had sent for the tokay, from the Emperor’s cellars, the
Imperial Tokay.
It was borne through the kitchen. Over her shoulder Lucy reported
how Miss Elizabeth looked quite lovely; she couldn’t take her eyes
off her; in her pink dress, wearing the necklace Mr. Dalloway had
given her. Jenny must remember the dog, Miss Elizabeth’s fox-
terrier, which, since it bit, had to be shut up and might, Elizabeth
thought, want something. Jenny must remember the dog. But Jenny
was not going upstairs with all those people about. There was a
motor at the door already! There was a ring at the bell—and the
gentlemen still in the dining-room, drinking tokay!
There, they were going upstairs; that was the first to come, and now
they would come faster and faster, so that Mrs. Parkinson (hired for
parties) would leave the hall door ajar, and the hall would be full of
gentlemen waiting (they stood waiting, sleeking down their hair)
while the ladies took their cloaks off in the room along the passage;
where Mrs. Barnet helped them, old Ellen Barnet, who had been with
the family for forty years, and came every summer to help the ladies,
and remembered mothers when they were girls, and though very
unassuming did shake hands; said “milady” very respectfully, yet had
a humorous way with her, looking at the young ladies, and ever so
tactfully helping Lady Lovejoy, who had some trouble with her
underbodice. And they could not help feeling, Lady Lovejoy and Miss
Alice, that some little privilege in the matter of brush and comb, was
awarded them having known Mrs. Barnet—“thirty years, milady,”
Mrs. Barnet supplied her. Young ladies did not use to rouge, said
Lady Lovejoy, when they stayed at Bourton in the old days. And Miss
Alice didn’t need rouge, said Mrs. Barnet, looking at her fondly.
There Mrs. Barnet would sit, in the cloakroom, patting down the furs,
smoothing out the Spanish shawls, tidying the dressing-table, and
knowing perfectly well, in spite of the furs and the embroideries,
which were nice ladies, which were not. The dear old body, said
Lady Lovejoy, mounting the stairs, Clarissa’s old nurse.
And then Lady Lovejoy stiffened. “Lady and Miss Lovejoy,” she said
to Mr. Wilkins (hired for parties). He had an admirable manner, as he
bent and straightened himself, bent and straightened himself and
announced with perfect impartiality “Lady and Miss Lovejoy ... Sir
John and Lady Needham ... Miss Weld ... Mr. Walsh.” His manner
was admirable; his family life must be irreproachable, except that it
seemed impossible that a being with greenish lips and shaven
cheeks could ever have blundered into the nuisance of children.
“How delightful to see you!” said Clarissa. She said it to every one.
How delightful to see you! She was at her worst—effusive, insincere.
It was a great mistake to have come. He should have stayed at
home and read his book, thought Peter Walsh; should have gone to
a music hall; he should have stayed at home, for he knew no one.
Oh dear, it was going to be a failure; a complete failure, Clarissa felt
it in her bones as dear old Lord Lexham stood there apologising for
his wife who had caught cold at the Buckingham Palace garden
party. She could see Peter out of the tail of her eye, criticising her,
there, in that corner. Why, after all, did she do these things? Why
seek pinnacles and stand drenched in fire? Might it consume her
anyhow! Burn her to cinders! Better anything, better brandish one’s
torch and hurl it to earth than taper and dwindle away like some Ellie
Henderson! It was extraordinary how Peter put her into these states
just by coming and standing in a corner. He made her see herself;
exaggerate. It was idiotic. But why did he come, then, merely to
criticise? Why always take, never give? Why not risk one’s one little
point of view? There he was wandering off, and she must speak to
him. But she would not get the chance. Life was that—humiliation,
renunciation. What Lord Lexham was saying was that his wife would
not wear her furs at the garden party because “my dear, you ladies
are all alike”—Lady Lexham being seventy-five at least! It was
delicious, how they petted each other, that old couple. She did like
old Lord Lexham. She did think it mattered, her party, and it made
her feel quite sick to know that it was all going wrong, all falling flat.
Anything, any explosion, any horror was better than people
wandering aimlessly, standing in a bunch at a corner like Ellie
Henderson, not even caring to hold themselves upright.
Gently the yellow curtain with all the birds of Paradise blew out and it
seemed as if there were a flight of wings into the room, right out,
then sucked back. (For the windows were open.) Was it draughty,
Ellie Henderson wondered? She was subject to chills. But it did not
matter that she should come down sneezing to-morrow; it was the
girls with their naked shoulders she thought of, being trained to think
of others by an old father, an invalid, late vicar of Bourton, but he
was dead now; and her chills never went to her chest, never. It was
the girls she thought of, the young girls with their bare shoulders, she
herself having always been a wisp of a creature, with her thin hair
and meagre profile; though now, past fifty, there was beginning to
shine through some mild beam, something purified into distinction by
years of self-abnegation but obscured again, perpetually, by her
distressing gentility, her panic fear, which arose from three hundred
pounds’ income, and her weaponless state (she could not earn a
penny) and it made her timid, and more and more disqualified year
by year to meet well-dressed people who did this sort of thing every
night of the season, merely telling their maids “I’ll wear so and so,”
whereas Ellie Henderson ran out nervously and bought cheap pink
flowers, half a dozen, and then threw a shawl over her old black
dress. For her invitation to Clarissa’s party had come at the last
moment. She was not quite happy about it. She had a sort of feeling
that Clarissa had not meant to ask her this year.
Why should she? There was no reason really, except that they had
always known each other. Indeed, they were cousins. But naturally
they had rather drifted apart, Clarissa being so sought after. It was
an event to her, going to a party. It was quite a treat just to see the
lovely clothes. Wasn’t that Elizabeth, grown up, with her hair done in
the fashionable way, in the pink dress? Yet she could not be more
than seventeen. She was very, very handsome. But girls when they
first came out didn’t seem to wear white as they used. (She must
remember everything to tell Edith.) Girls wore straight frocks,
perfectly tight, with skirts well above the ankles. It was not becoming,
she thought.
So, with her weak eyesight, Ellie Henderson craned rather forward,
and it wasn’t so much she who minded not having any one to talk to
(she hardly knew anybody there), for she felt that they were all such
interesting people to watch; politicians presumably; Richard
Dalloway’s friends; but it was Richard himself who felt that he could
not let the poor creature go on standing there all the evening by
herself.
“Well, Ellie, and how’s the world treating you?” he said in his genial
way, and Ellie Henderson, getting nervous and flushing and feeling
that it was extraordinarily nice of him to come and talk to her, said
that many people really felt the heat more than the cold.
“Yes, they do,” said Richard Dalloway. “Yes.”
But what more did one say?
“Hullo, Richard,” said somebody, taking him by the elbow, and, good
Lord, there was old Peter, old Peter Walsh. He was delighted to see
him—ever so pleased to see him! He hadn’t changed a bit. And off
they went together walking right across the room, giving each other
little pats, as if they hadn’t met for a long time, Ellie Henderson
thought, watching them go, certain she knew that man’s face. A tall
man, middle aged, rather fine eyes, dark, wearing spectacles, with a
look of John Burrows. Edith would be sure to know.
The curtain with its flight of birds of Paradise blew out again. And
Clarissa saw—she saw Ralph Lyon beat it back, and go on talking.
So it wasn’t a failure after all! it was going to be all right now—her
party. It had begun. It had started. But it was still touch and go. She
must stand there for the present. People seemed to come in a rush.
Colonel and Mrs. Garrod ... Mr. Hugh Whitbread ... Mr. Bowley ...
Mrs. Hilbery ... Lady Mary Maddox ... Mr. Quin ... intoned Wilkin. She
had six or seven words with each, and they went on, they went into
the rooms; into something now, not nothing, since Ralph Lyon had
beat back the curtain.
And yet for her own part, it was too much of an effort. She was not
enjoying it. It was too much like being—just anybody, standing there;
anybody could do it; yet this anybody she did a little admire, couldn’t
help feeling that she had, anyhow, made this happen, that it marked
a stage, this post that she felt herself to have become, for oddly
enough she had quite forgotten what she looked like, but felt herself
a stake driven in at the top of her stairs. Every time she gave a party
she had this feeling of being something not herself, and that every
one was unreal in one way; much more real in another. It was, she
thought, partly their clothes, partly being taken out of their ordinary
ways, partly the background, it was possible to say things you
couldn’t say anyhow else, things that needed an effort; possible to
go much deeper. But not for her; not yet anyhow.
“How delightful to see you!” she said. Dear old Sir Harry! He would
know every one.
And what was so odd about it was the sense one had as they came
up the stairs one after another, Mrs. Mount and Celia, Herbert Ainsty,
Mrs. Dakers—oh and Lady Bruton!
“How awfully good of you to come!” she said, and she meant it—it
was odd how standing there one felt them going on, going on, some
quite old, some....
What name? Lady Rosseter? But who on earth was Lady Rosseter?
“Clarissa!” That voice! It was Sally Seton! Sally Seton! after all these
years! She loomed through a mist. For she hadn’t looked like that,
Sally Seton, when Clarissa grasped the hot water can, to think of her
under this roof, under this roof! Not like that!
All on top of each other, embarrassed, laughing, words tumbled out
—passing through London; heard from Clara Haydon; what a chance
of seeing you! So I thrust myself in—without an invitation....
One might put down the hot water can quite composedly. The lustre
had gone out of her. Yet it was extraordinary to see her again, older,
happier, less lovely. They kissed each other, first this cheek then
that, by the drawing-room door, and Clarissa turned, with Sally’s
hand in hers, and saw her rooms full, heard the roar of voices, saw
the candlesticks, the blowing curtains, and the roses which Richard
had given her.
“I have five enormous boys,” said Sally.
She had the simplest egotism, the most open desire to be thought
first always, and Clarissa loved her for being still like that. “I can’t
believe it!” she cried, kindling all over with pleasure at the thought of
the past.
But alas, Wilkins; Wilkins wanted her; Wilkins was emitting in a voice
of commanding authority as if the whole company must be
admonished and the hostess reclaimed from frivolity, one name:
“The Prime Minister,” said Peter Walsh.
The Prime Minister? Was it really? Ellie Henderson marvelled. What
a thing to tell Edith!
One couldn’t laugh at him. He looked so ordinary. You might have
stood him behind a counter and bought biscuits—poor chap, all
rigged up in gold lace. And to be fair, as he went his rounds, first with
Clarissa then with Richard escorting him, he did it very well. He tried
to look somebody. It was amusing to watch. Nobody looked at him.
They just went on talking, yet it was perfectly plain that they all knew,
felt to the marrow of their bones, this majesty passing; this symbol of
what they all stood for, English society. Old Lady Bruton, and she
looked very fine too, very stalwart in her lace, swam up, and they
withdrew into a little room which at once became spied upon,
guarded, and a sort of stir and rustle rippled through every one,
openly: the Prime Minister!
Lord, lord, the snobbery of the English! thought Peter Walsh,
standing in the corner. How they loved dressing up in gold lace and
doing homage! There! That must be, by Jove it was, Hugh
Whitbread, snuffing round the precincts of the great, grown rather
fatter, rather whiter, the admirable Hugh!
He looked always as if he were on duty, thought Peter, a privileged,
but secretive being, hoarding secrets which he would die to defend,
though it was only some little piece of tittle-tattle dropped by a court
footman, which would be in all the papers to-morrow. Such were his
rattles, his baubles, in playing with which he had grown white, come
to the verge of old age, enjoying the respect and affection of all who
had the privilege of knowing this type of the English public school
man. Inevitably one made up things like that about Hugh; that was
his style; the style of those admirable letters which Peter had read
thousands of miles across the sea in the Times, and had thanked
God he was out of that pernicious hubble-bubble if it were only to
hear baboons chatter and coolies beat their wives. An olive-skinned
youth from one of the Universities stood obsequiously by. Him he
would patronise, initiate, teach how to get on. For he liked nothing
better than doing kindnesses, making the hearts of old ladies
palpitate with the joy of being thought of in their age, their affliction,
thinking themselves quite forgotten, yet here was dear Hugh driving
up and spending an hour talking of the past, remembering trifles,
praising the home-made cake, though Hugh might eat cake with a
Duchess any day of his life, and, to look at him, probably did spend a
good deal of time in that agreeable occupation. The All-judging, the
All-merciful, might excuse. Peter Walsh had no mercy. Villains there
must be, and God knows the rascals who get hanged for battering
the brains of a girl out in a train do less harm on the whole than
Hugh Whitbread and his kindness. Look at him now, on tiptoe,
dancing forward, bowing and scraping, as the Prime Minister and
Lady Bruton emerged, intimating for all the world to see that he was
privileged to say something, something private, to Lady Bruton as
she passed. She stopped. She wagged her fine old head. She was
thanking him presumably for some piece of servility. She had her
toadies, minor officials in Government offices who ran about putting
through little jobs on her behalf, in return for which she gave them
luncheon. But she derived from the eighteenth century. She was all
right.
And now Clarissa escorted her Prime Minister down the room,
prancing, sparkling, with the stateliness of her grey hair. She wore
ear-rings, and a silver-green mermaid’s dress. Lolloping on the
waves and braiding her tresses she seemed, having that gift still; to
be; to exist; to sum it all up in the moment as she passed; turned,
caught her scarf in some other woman’s dress, unhitched it, laughed,
all with the most perfect ease and air of a creature floating in its
element. But age had brushed her; even as a mermaid might behold
in her glass the setting sun on some very clear evening over the
waves. There was a breath of tenderness; her severity, her prudery,
her woodenness were all warmed through now, and she had about
her as she said good-bye to the thick gold-laced man who was doing
his best, and good luck to him, to look important, an inexpressible
dignity; an exquisite cordiality; as if she wished the whole world well,
and must now, being on the very verge and rim of things, take her
leave. So she made him think. (But he was not in love.)
Indeed, Clarissa felt, the Prime Minister had been good to come.
And, walking down the room with him, with Sally there and Peter
there and Richard very pleased, with all those people rather inclined,
perhaps, to envy, she had felt that intoxication of the moment, that
dilatation of the nerves of the heart itself till it seemed to quiver,
steeped, upright;—yes, but after all it was what other people felt,
that; for, though she loved it and felt it tingle and sting, still these
semblances, these triumphs (dear old Peter, for example, thinking
her so brilliant), had a hollowness; at arm’s length they were, not in
the heart; and it might be that she was growing old but they satisfied
her no longer as they used; and suddenly, as she saw the Prime
Minister go down the stairs, the gilt rim of the Sir Joshua picture of
the little girl with a muff brought back Kilman with a rush; Kilman her
enemy. That was satisfying; that was real. Ah, how she hated her—
hot, hypocritical, corrupt; with all that power; Elizabeth’s seducer; the
woman who had crept in to steal and defile (Richard would say,
What nonsense!). She hated her: she loved her. It was enemies one
wanted, not friends—not Mrs. Durrant and Clara, Sir William and
Lady Bradshaw, Miss Truelock and Eleanor Gibson (whom she saw
coming upstairs). They must find her if they wanted her. She was for
the party!
There was her old friend Sir Harry.
“Dear Sir Harry!” she said, going up to the fine old fellow who had
produced more bad pictures than any other two Academicians in the
whole of St. John’s Wood (they were always of cattle, standing in
sunset pools absorbing moisture, or signifying, for he had a certain
range of gesture, by the raising of one foreleg and the toss of the
antlers, “the Approach of the Stranger”—all his activities, dining out,
racing, were founded on cattle standing absorbing moisture in sunset
pools).
“What are you laughing at?” she asked him. For Willie Titcomb and
Sir Harry and Herbert Ainsty were all laughing. But no. Sir Harry
could not tell Clarissa Dalloway (much though he liked her; of her
type he thought her perfect, and threatened to paint her) his stories
of the music hall stage. He chaffed her about her party. He missed
his brandy. These circles, he said, were above him. But he liked her;
respected her, in spite of her damnable, difficult upper-class
refinement, which made it impossible to ask Clarissa Dalloway to sit
on his knee. And up came that wandering will-o’-the-wisp, that
vagulous phosphorescence, old Mrs. Hilbery, stretching her hands to
the blaze of his laughter (about the Duke and the Lady), which, as
she heard it across the room, seemed to reassure her on a point
which sometimes bothered her if she woke early in the morning and
did not like to call her maid for a cup of tea; how it is certain we must
die.
“They won’t tell us their stories,” said Clarissa.
“Dear Clarissa!” exclaimed Mrs. Hilbery. She looked to-night, she
said, so like her mother as she first saw her walking in a garden in a
grey hat.
And really Clarissa’s eyes filled with tears. Her mother, walking in a
garden! But alas, she must go.
For there was Professor Brierly, who lectured on Milton, talking to
little Jim Hutton (who was unable even for a party like this to
compass both tie and waistcoat or make his hair lie flat), and even at
this distance they were quarrelling, she could see. For Professor
Brierly was a very queer fish. With all those degrees, honours,
lectureships between him and the scribblers he suspected instantly
an atmosphere not favourable to his queer compound; his prodigious
learning and timidity; his wintry charm without cordiality; his
innocence blent with snobbery; he quivered if made conscious by a
lady’s unkempt hair, a youth’s boots, of an underworld, very
creditable doubtless, of rebels, of ardent young people; of would-be
geniuses, and intimated with a little toss of the head, with a sniff—
Humph!—the value of moderation; of some slight training in the
classics in order to appreciate Milton. Professor Brierly (Clarissa
could see) wasn’t hitting it off with little Jim Hutton (who wore red
socks, his black being at the laundry) about Milton. She interrupted.
She said she loved Bach. So did Hutton. That was the bond between
them, and Hutton (a very bad poet) always felt that Mrs. Dalloway
was far the best of the great ladies who took an interest in art. It was
odd how strict she was. About music she was purely impersonal.
She was rather a prig. But how charming to look at! She made her
house so nice if it weren’t for her Professors. Clarissa had half a
mind to snatch him off and set him down at the piano in the back
room. For he played divinely.
“But the noise!” she said. “The noise!”
“The sign of a successful party.” Nodding urbanely, the Professor
stepped delicately off.
“He knows everything in the whole world about Milton,” said Clarissa.
“Does he indeed?” said Hutton, who would imitate the Professor
throughout Hampstead; the Professor on Milton; the Professor on
moderation; the Professor stepping delicately off.
But she must speak to that couple, said Clarissa, Lord Gayton and
Nancy Blow.
Not that they added perceptibly to the noise of the party. They were
not talking (perceptibly) as they stood side by side by the yellow
curtains. They would soon be off elsewhere, together; and never had
very much to say in any circumstances. They looked; that was all.
That was enough. They looked so clean, so sound, she with an
apricot bloom of powder and paint, but he scrubbed, rinsed, with the
eyes of a bird, so that no ball could pass him or stroke surprise him.
He struck, he leapt, accurately, on the spot. Ponies’ mouths quivered
at the end of his reins. He had his honours, ancestral monuments,
banners hanging in the church at home. He had his duties; his
tenants; a mother and sisters; had been all day at Lords, and that
was what they were talking about—cricket, cousins, the movies—
when Mrs. Dalloway came up. Lord Gayton liked her most awfully.
So did Miss Blow. She had such charming manners.
“It is angelic—it is delicious of you to have come!” she said. She
loved Lords; she loved youth, and Nancy, dressed at enormous
expense by the greatest artists in Paris, stood there looking as if her
body had merely put forth, of its own accord, a green frill.
“I had meant to have dancing,” said Clarissa.
For the young people could not talk. And why should they? Shout,
embrace, swing, be up at dawn; carry sugar to ponies; kiss and
caress the snouts of adorable chows; and then all tingling and
streaming, plunge and swim. But the enormous resources of the
English language, the power it bestows, after all, of communicating
feelings (at their age, she and Peter would have been arguing all the
evening), was not for them. They would solidify young. They would
be good beyond measure to the people on the estate, but alone,
perhaps, rather dull.
“What a pity!” she said. “I had hoped to have dancing.”
It was so extraordinarily nice of them to have come! But talk of
dancing! The rooms were packed.
There was old Aunt Helena in her shawl. Alas, she must leave them
—Lord Gayton and Nancy Blow. There was old Miss Parry, her aunt.
For Miss Helena Parry was not dead: Miss Parry was alive. She was
past eighty. She ascended staircases slowly with a stick. She was
placed in a chair (Richard had seen to it). People who had known
Burma in the ’seventies were always led up to her. Where had Peter
got to? They used to be such friends. For at the mention of India, or
even Ceylon, her eyes (only one was glass) slowly deepened,
became blue, beheld, not human beings—she had no tender
memories, no proud illusions about Viceroys, Generals, Mutinies—it
was orchids she saw, and mountain passes and herself carried on
the backs of coolies in the ’sixties over solitary peaks; or descending
to uproot orchids (startling blossoms, never beheld before) which
she painted in water-colour; an indomitable Englishwoman, fretful if
disturbed by the War, say, which dropped a bomb at her very door,
from her deep meditation over orchids and her own figure journeying
in the ’sixties in India—but here was Peter.
“Come and talk to Aunt Helena about Burma,” said Clarissa.
And yet he had not had a word with her all the evening!
“We will talk later,” said Clarissa, leading him up to Aunt Helena, in
her white shawl, with her stick.
“Peter Walsh,” said Clarissa.
That meant nothing.
Clarissa had asked her. It was tiring; it was noisy; but Clarissa had
asked her. So she had come. It was a pity that they lived in London
—Richard and Clarissa. If only for Clarissa’s health it would have
been better to live in the country. But Clarissa had always been fond
of society.
“He has been in Burma,” said Clarissa.
Ah. She could not resist recalling what Charles Darwin had said
about her little book on the orchids of Burma.
(Clarissa must speak to Lady Bruton.)
No doubt it was forgotten now, her book on the orchids of Burma, but
it went into three editions before 1870, she told Peter. She
remembered him now. He had been at Bourton (and he had left her,
Peter Walsh remembered, without a word in the drawing-room that
night when Clarissa had asked him to come boating).
“Richard so much enjoyed his lunch party,” said Clarissa to Lady
Bruton.
“Richard was the greatest possible help,” Lady Bruton replied. “He
helped me to write a letter. And how are you?”
“Oh, perfectly well!” said Clarissa. (Lady Bruton detested illness in
the wives of politicians.)
“And there’s Peter Walsh!” said Lady Bruton (for she could never
think of anything to say to Clarissa; though she liked her. She had
lots of fine qualities; but they had nothing in common—she and
Clarissa. It might have been better if Richard had married a woman
with less charm, who would have helped him more in his work. He
had lost his chance of the Cabinet). “There’s Peter Walsh!” she said,
shaking hands with that agreeable sinner, that very able fellow who
should have made a name for himself but hadn’t (always in
difficulties with women), and, of course, old Miss Parry. Wonderful
old lady!
Lady Bruton stood by Miss Parry’s chair, a spectral grenadier,
draped in black, inviting Peter Walsh to lunch; cordial; but without
small talk, remembering nothing whatever about the flora or fauna of
India. She had been there, of course; had stayed with three
Viceroys; thought some of the Indian civilians uncommonly fine
fellows; but what a tragedy it was—the state of India! The Prime
Minister had just been telling her (old Miss Parry huddled up in her
shawl, did not care what the Prime Minister had just been telling
her), and Lady Bruton would like to have Peter Walsh’s opinion, he
being fresh from the centre, and she would get Sir Sampson to meet
him, for really it prevented her from sleeping at night, the folly of it,
the wickedness she might say, being a soldier’s daughter. She was
an old woman now, not good for much. But her house, her servants,
her good friend Milly Brush—did he remember her?—were all there
only asking to be used if—if they could be of help, in short. For she
never spoke of England, but this isle of men, this dear, dear land,
was in her blood (without reading Shakespeare), and if ever a