Ethics for Robots How to Design a Moral

Algorithm Derek Leben pdf download

https://textbookfull.com/product/ethics-for-robots-how-to-design-
a-moral-algorithm-derek-leben/

Download more ebook from https://textbookfull.com

We believe these products will be a great fit for you. Click
the link to download now, or visit textbookfull.com
to discover even more!

Moral choices an introduction to ethics Rae

https://textbookfull.com/product/moral-choices-an-introduction-
to-ethics-rae/

Caring A Relational Approach to Ethics and Moral

Education 2nd Edition Nel Noddings

https://textbookfull.com/product/caring-a-relational-approach-to-
ethics-and-moral-education-2nd-edition-nel-noddings/

The Ethical Algorithm The Science of Socially Aware

Algorithm Design 6th Edition Kearns

https://textbookfull.com/product/the-ethical-algorithm-the-
science-of-socially-aware-algorithm-design-6th-edition-kearns/

The Moral of the Story: An Introduction to Ethics Nina

Rosenstand

https://textbookfull.com/product/the-moral-of-the-story-an-
introduction-to-ethics-nina-rosenstand/
Algorithm Design with Haskell Richard S. Bird

https://textbookfull.com/product/algorithm-design-with-haskell-
richard-s-bird/

Humans And Robots Ethics Agency And Anthropomorphism

Sven Nyholm

https://textbookfull.com/product/humans-and-robots-ethics-agency-
and-anthropomorphism-sven-nyholm/

Algorithm Design Practice for Collegiate Programming

Contests and Education 1st Edition Yonghui Wu

https://textbookfull.com/product/algorithm-design-practice-for-
collegiate-programming-contests-and-education-1st-edition-
yonghui-wu/

Beyond the Algorithm AI Security Privacy and Ethics 1st

Edition Santos

https://textbookfull.com/product/beyond-the-algorithm-ai-
security-privacy-and-ethics-1st-edition-santos/

Soft Robots for Healthcare Applications Design Modeling

and Control 1st Edition Shane Xie

https://textbookfull.com/product/soft-robots-for-healthcare-
applications-design-modeling-and-control-1st-edition-shane-xie/
ETHICS FOR ROBOTS

Ethics for Robots describes and defends a method for designing and evaluating
ethics algorithms for autonomous machines, such as self-driving cars and search
and rescue drones. Derek Leben argues that such algorithms should be evaluated by
how effectively they accomplish the problem of cooperation among self-interested
organisms, and therefore, rather than simulating the psychological systems that have
evolved to solve this problem, engineers should be tackling the problem itself, tak-
ing relevant lessons from our moral psychology.
Leben draws on the moral theory of John Rawls, arguing that normative moral
theories are attempts to develop optimal solutions to the problem of cooperation. He
claims that Rawlsian Contractarianism leads to the ‘Maximin’ principle – the action
that maximizes the minimum value – and that the Maximin principle is the most
effective solution to the problem of cooperation. He contrasts the Maximin principle
with other principles and shows how they can often produce non-cooperative results.
Using real-world examples – such as an autonomous vehicle facing a situation
where every action results in harm, home care machines, and autonomous weap-
ons systems – Leben contrasts Rawlsian algorithms with alternatives derived from
utilitarianism and natural rights libertarianism.
Including chapter summaries and a glossary of technical terms, Ethics for Robots
is essential reading for philosophers, engineers, computer scientists, and cognitive
scientists working on the problem of ethics for autonomous systems.

Derek Leben is Associate Professor of Philosophy, University of Pittsburgh at

Johnstown, USA.
ETHICS FOR ROBOTS
How to Design a Moral Algorithm

Derek Leben
First published 2019
by Routledge
2 Park Square, Milton Park, Abingdon, Oxon OX14 4RN
and by Routledge
711 Third Avenue, New York, NY 10017
Routledge is an imprint of the Taylor & Francis Group, an informa business
© 2019 Derek Leben
The right of Derek Leben to be identified as author of this work has been
asserted by him in accordance with sections 77 and 78 of the Copyright,
Designs and Patents Act 1988.
All rights reserved. No part of this book may be reprinted or reproduced
or utilised in any form or by any electronic, mechanical, or other means,
now known or hereafter invented, including photocopying and recording,
or in any information storage or retrieval system, without permission in
writing from the publishers.
Trademark notice: Product or corporate names may be trademarks or
registered trademarks, and are used only for identification and explanation
without intent to infringe.
British Library Cataloguing-in-Publication Data
A catalogue record for this book is available from the British Library
Library of Congress Cataloging-in-Publication Data
Names: Leben, Derek, author.
Title: Ethics for robots: how to design a moral algorithm / Derek Leben.
Description: Abingdon, Oxon; New York, NY: Routledge, 2018. |
Includes bibliographical references and index.
Identifiers: LCCN 2018007376 | ISBN 9781138716155 (hbk: alk. paper) |
ISBN 9781138716179 (pbk: alk. paper) | ISBN 9781315197128 (ebk)
Subjects: LCSH: Robotics–Moral and ethical aspects. |
Robotics–Safety measures.
Classification: LCC TJ211.28 .L43 2018 | DDC 174/.9629892–dc23
LC record available at https://lccn.loc.gov/2018007376
ISBN: 978-1-138-71615-5 (hbk)
ISBN: 978-1-138-71617-9 (pbk)
ISBN: 978-1-315-19712-8 (ebk)
Typeset in Bembo
by Deanta Global Publishing Services, Chennai, India
Dedicated to Sean and Ethan
CONTENTS

Acknowledgments viii

Introduction 1

1 Moral psychology 7

2 Cooperation problems 25

3 Theories 42

4 Contractarianism 59

5 Ethics engines 76

6 Avoiding collisions 97

7 Saving lives 116

8 Keeping the peace 130

Conclusions 146

Glossary 151
Index 155
ACKNOWLEDGMENTS

Thanks to David Danks, Jeff Maynes, and Michael Cox for helpful feedback.

And thanks to the 61B café for being a good place to write a book.
INTRODUCTION

The word robot was coined by the Czech playwright and journalist Karel Capek
in his 1920 play, Rossum’s Universal Robots. The play is about a company that cre-
ates biological human workers that don’t have a soul, so they can take over all the
jobs that humans usually do. If this idea seemed far-fetched in the 1920s, it doesn’t
seem that way today. Machines are rapidly taking over tasks that were previously
performed by humans in every domain of our society, including agriculture,
factory production, transportation, medicine, retail sales, finance, education, and
even warfare. Most of these machines have simple programs that allow them to
automatically perform a single repetitive task like welding car doors, scanning
products, or vacuuming the f loor. Gradually, machines are beginning to also
take over tasks which involve weighing several different options to arrive at a
decision, like driving, diagnosing diseases, and responding to threats. Let’s call
these kinds of tasks complex. For our purposes, a robot is any physically embodied
machine that can perform complex tasks without any direct human intervention.
According to this definition, a robot doesn’t necessarily have to look like a
human. For instance, I’ll consider driverless cars and certain kinds of missile
systems to be robots. The key feature of a robot is that it is autonomous, which
means making decisions based on principles or reasoning without direct human
intervention. Autonomy has a special meaning in moral philosophy; it’s not just
being able to act in response to input (this might be called automatic), but instead,
being able to think and make decisions in a responsible way. This requires a
minimal kind of artificial intelligence, but I will consider AI to be a broader
class of systems that are not necessarily embodied and more general in the scope
of their abilities. The robots that we’re most interested in are machines that
operate dangerous vehicles or equipment, perform medical services, and provide
security. These machines will be making decisions that could result in harm to
others, which is why we need a framework for designing ethical robots.
2 Introduction

One early proposal for a robot ethics originated in the science-fiction stories
of Isaac Asimov. Asimov’s three laws of robotics are:

1. A robot may not harm a human being or, through inaction, allow a human
being to come to harm.
2. A robot must obey orders given it by human beings except where such
orders would conf lict with the First Law.
3. A robot must protect its own existence, as long as such protection does not
conf lict with the First or Second Law.

These principles seem appealing at first, and they work well enough as a rough
guideline for most normal situations. It’s no coincidence that the first rule looks
like the “golden rule” that’s been repeated throughout cultural and religious
traditions from Confucianism to Christianity.
Even though “don’t cause or allow harm” is good as a rough guideline for
behavior, it won’t work as a detailed rule for decision-making. One major prob-
lem is that this law is useless until we define what counts as harm. Is it harming
someone to insult them, lie to them, or trespass on their property? What about
actions that violate a person’s consent or dignity? Some actions are only likely to
cause harm, but what’s the threshold for likely harm? Every action could possibly
lead to harm, so failing to specify this threshold will leave robots paralyzed with
fear, unable to perform even the most basic of tasks.
Another problem with a rule like “don’t cause or allow harm” is that it immedi-
ately breaks down once we venture to situations where every action leads to harm
for a human being, even the action of doing nothing at all. These cases are called
moral dilemmas. Although moral dilemmas aren’t common, they do happen, and
they can be disturbingly frequent in fields like medicine and warfare. Doctors
must sometimes decide between respecting a patient’s wishes and doing what’s
best for her health. Soldiers must sometimes decide between killing defenseless
enemies and allowing civilians to die. Moral dilemmas are cases where harms to
one person or group are inevitable, so an agent must decide which harm is worse.
What we need is a more specific way to define and weigh harms.
Asimov’s laws might be called a “top-down” approach to programming ethi-
cal robots, where we use a set of rigid rules to constrain and guide behavior.
An alternative approach, which is sometimes called “bottom-up,” uses machine
learning techniques to change a robot’s behavior in response to positive or nega-
tive feedback. It’s true that some of the most impressive examples of machines
performing complex human tasks in the past decade have used this kind of rein-
forcement learning within a multilayered network of f lexible artificial neurons.
For instance, in 2010, the Stanford computer scientist Fei-Fei Li began an annual
competition where competitors would each submit their object recognition
algorithms to be tested against the most massive database of images in the world,
called ImageNet. The algorithms were compared based on performance on tasks
like recognition and classification. The human error rate on these tasks is about
 Introduction  3

5–6 percent. In 2010, the winning algorithm had an error rate of 28 percent. In
2011, the error rate was 25 percent. Then, in 2012, the introduction of “deep
learning” techniques produced a winning algorithm with an error rate of just 15
percent. Since then, the winner has improved its accuracy every year, with the
2015 algorithm having an error rate of 3.5 percent, which is better than human
judgment. If machine learning can simulate human performance at recognizing
images, using language, and driving vehicles, why not moral judgments?
The challenge for a machine learning approach to designing robot ethics is
that choices must be made about what kind of information is used as positive or
negative feedback. If we are using human judgments to model machine judg-
ments, then robots will inevitably incorporate the biases and inconsistencies
in our own psychology: preference for people who are familiar or genetically
related, ignoring the effects of our actions on people who are very distant, and
relying on false beliefs about what kinds of actions are harmful. A considerable
number of human beings over the course of history have been raised to approve
of horrible things like genocide, rape, slavery, torture, and child abuse, to name
just a few. Even if we take historical exemplars like Aristotle as our training set,
a well-respected model citizen in Aristotle’s homeland of fourth-century (bc)
Macedonia would probably be a slave-holding pedophile. My point is that we
can’t simply point a machine learning network at human behavior and shout:
“Learn!” Instead, machine learning approaches will need to make important
theoretical assumptions about what kinds of data are morally important.
The aim of this book is to provide a general theoretical framework for design-
ing moral algorithms, whether they be “top-down” or “bottom-up.” Many of
the engineers and scientists working on this problem don’t have training in eth-
ics, and don’t seem to think that they need any. However, it’s impossible to
design an ethics procedure for machines without making substantial theoretical
assumptions about how we solve moral problems. Perhaps the only way to objec-
tively solve moral problems is by understanding the function of morality. Armed
with a proper understanding of the practical function of morality, we can turn
to the engineering task of designing an artificial system that performs this same
function just as well as humans, if not better.
How can a machine be better at making moral decisions than a human being?
In the twenty-first century, most of us have no problem acknowledging that
computers can make better decisions than an average human when it comes
to games or calculations, but how could a machine ever surpass us in some-
thing so fundamentally human as moral decisions? This response reveals an impor-
tant assumption: that morality is essentially a product of our own minds, and
somehow limited to human beings. In philosophy, the term for this position is
“anti-realism,” as opposed to “realism.” The choice between realism or anti-
realism turns out to be the most important initial assumption in any discussion
about ethics.
If moral realism is false, and there are no objective mind-independent answers
to moral questions, then ethics is about a set of psychological responses and
4 Introduction

cultural traditions. None of these responses and traditions would be inherently

better or worse than the other. There would be no “correct” answers to moral
questions. In designing robot ethics, programmers could be accused of incorpo-
rating their own biases involving gender, sexual orientation, race, religion, or
cultural tradition. Ethics would become a kind of identity politics, where each
group lobbies against the other to try to have their own intuitions and favorite
principles represented. This approach is just as disastrous in ethics as it is in poli-
tics. In Plato’s Republic, he describes the “democratic man,” who is pushed and
pulled about by competing internal interests, never having a general rule that
manages and guides these interests. This person is in desperate need of a unify-
ing framework for tying these interests together, which we call a constitution. If
programmers design robots to be a sampler platter of various moral intuitions and
principles, they risk creating a real-life version of Plato’s democratic man, lacking
any constitution to its decisions.
On the other hand, if morality has adapted in response to a practical problem,
then there might exist an objective and mind-independent set of solutions to
this problem. In this kind of moral realism, we can survey human psychology
and culture in a context of discovery, as we explore the possible landscape of
attempted solutions. Ultimately, though, some solutions are objectively better
than others. In principle, solutions to the problem of cooperation could be gener-
ated without reference to any historical traditions, texts, or intuitions.
My training is in “Western” philosophy, so in this book, I’ll be mostly dis-
cussing figures like Plato, Hume, Kant, and Rawls. I’ll suggest that one moral
theory, Contractarianism, provides the best solution to the problem of coopera-
tion. I wouldn’t be surprised if this solution has been independently discovered
in different cultures, as is often the case in the history of science and math-
ematics. The reason that I won’t be discussing these non-Western traditions is
not cultural bias or insensitivity; instead, it’s a commitment to moral realism,
and the idea that solutions to moral problems are engineering problems, like
constructing a bridge. We can certainly gain inspiration from bridge-building
traditions in ancient China, India, and the Americas, but this isn’t essential to
solving the problem. Ultimately, the solutions to moral problems are transcend-
ent of culture. This also entails the conclusion that robots might not only surpass
us in their abilities to solve moral problems, but also help us understand our
own limitations.
It’s not just philosophers who can help engineers; the relationship is a two-
way street. Designing algorithms for machine behavior can help us gain focus
in guiding human behavior. Philosophers sometimes forget about the need for
concrete rules in decision-making. There’s a famous story about a student who
approached the philosopher Jean-Paul Sartre asking for advice. The student’s
brother had gone off to fight in the French Resistance, leaving him to stay
behind and care for their ailing mother. Despite this obligation to take care of his
mother, the student also felt a strong obligation to go off and fight for his brother
and country. Sartre’s response was: “You are free, therefore choose – that is to say
 Introduction  5

invent. No rule of general morality can show you what you ought to do: no signs
are vouchsafed in this world.” I imagine that the student’s response was a sar-
castic: “Great, thanks.” There are many moral theories that make vague claims
about one being a virtuous person and expressing care for other people, but these
theories don’t provide any practical guidelines for actions. Thinking about how
we would program a machine to be “virtuous” or “caring” illustrates how useless
these moral theories can be. It forces us to be specific in ways that we’ve never
been forced to be, and to do the hard work needed to produce a real decision-
procedure, not just for machines but for ourselves.
The first half of this book will survey approaches to modeling ethics algo-
rithms based on (1) universal features of human moral psychology, (2) successful
strategies in cooperation games, and (3) historically inf luential moral theories.
All of these approaches are promising, but what is needed is an overarching
theoretical framework tying them together. If we view them as linked together
by functionality, it will enable us to take the parts of each approach that work
and ignore the parts that are irrelevant. Specifically, I’ll argue that our moral
intuitions are the product of a psychological network that adapted in response to
the problem of enforcing cooperative behavior among self-interested organisms.
Moral theories are (sometimes unconscious) attempts to clarify and generalize
these intuitive judgments. Once we view both moral intuitions and theories as
goal-directed, it’s possible to objectively evaluate which features are most effec-
tive at accomplishing this goal. As the most effective solution to cooperation
problems, Contractarianism contains the universal features of our moral psy-
chology and extends them consistently to non-cooperative contexts like moral
dilemmas. It provides a detailed way of determining which objects are valuable,
what kinds of actions are harmful, and the importance of concepts like rights and
consent, as well as providing a general rule called the Maximin principle that can
produce a unique decision in even the most difficult of moral dilemmas. If you
aren’t convinced by my arguments that ethics is connected to cooperation prob-
lems, the rest of the book is still valuable: you can just think of it as “cooperation
for robots” instead.
The second half of the book will examine how Contractarianism can be
turned into a program for autonomous machines. Using chess engines as a
model, I describe how an ethics engine based on Contractarianism would
operate, what its algorithms might look like, and how this procedure applies to
decisions made by robots in various domains. The domains include transpor-
tation, saving lives, and keeping the peace. There are many important ques-
tions about robot ethics that aren’t discussed here. These questions include:
“Under what conditions is a robot morally responsible for its actions?” “Who
do we hold responsible when a robot misbehaves?” “When, if ever, does a
robot become a person with rights?” “What are the social, economic, and legal
implications of allowing robots to take over more of our decision-making?”
Instead, the focus of this book is: “If we are going to build autonomous robots
in domains like transportation, medicine, and war, here are the moral principles
6 Introduction

that should constrain their decisions.” This leaves open whether autonomous
robots can ever be genuinely responsible, worthy of rights, or even a good idea
to build in the first place. My suspicion is that full automation is a good idea
in domains like transportation, but a bad idea in domains like warfare. I am
also generally optimistic that increasing the presence of autonomous machines
into our society and economy will produce beneficial results, providing greater
wealth, leisure, and security for even the worst-off members of the population.
However, I have been wrong before.
Designing ethical machines is an interdisciplinary project, and this book will
draw on work in math, biology, economics, philosophy, computer science, and
cognitive science. I will start by apologizing to each of these fields, because I’m
going to oversimplify topics that require volumes to address in detail. This even
includes my own field of philosophy; I’ll often be squeezing important debates
that span centuries of time and volumes of texts into a single page or even a sin-
gle paragraph. I’m painfully aware of the extent to which these topics are being
condensed and often oversimplified. However, given that I am trying to present
a broad theoretical framework here in an accessible and compact format, I hope
that the philosophers will forgive me as well.
The goal of interdisciplinary research is to tie together work in different fields
in a new way, leading to unexpected results. I’ve tried to write the book at a
level that makes it accessible to any motivated reader. Mathematical details and
intimidating symbols have been left out or pushed to endnotes. Any graphs or
formal expressions that I’ve left in the main text should be accessible to anyone
with even the most basic math background, and I encourage you to spend a
minute or two working through them. Many people don’t have the time or the
attention span to read an entire book. Don’t feel bad about it. I’ve tried to make
each chapter somewhat independent, so feel free to skip around. For better or
worse, every citizen of the industrialized world must now take an interest in
robot decision-making, and the people working on them should be trying to
bring the conversation to the public as much as possible.
1
MORAL PSYCHOLOGY

Harry Truman had only been the vice-president for 82 days when, on April
12, 1945, Franklin Roosevelt died. Truman became the 33rd president at the
end of World War II, when the Japanese Empire had been pushed back to its
mainland islands, and preparations were being made for the Allied invasion of
Japan. The planned invasion, called Operation Downfall, was estimated to last
for years and cost perhaps half a million American lives, along with millions
more Japanese casualties.
Just 13 days after taking office, the U.S. Secretary of War, Henry Stimson,
sent Truman a letter describing a “secret matter” that “I think you ought to
know about … without delay.” This secret was the atomic bomb, which had
been the product of secret research conducted at the Los Alamos research site.
Within just a few months of learning about what he called “the most terrible
bomb in the history of the world,” Truman now faced a monumental decision:
would he allow the planned invasion of Japan to continue its deadly course,
costing millions of lives, or deliberately drop atomic bombs on Japanese cities,
at a cost of maybe only a few hundred thousand lives, to force the enemy into
unconditional surrender?
Truman chose to drop the atomic bomb, first on Hiroshima (August 6) and
then on Nagasaki (August 9). The death tolls are estimated at 90,000–120,000
for Hiroshima, and 60,000–80,000 for Nagasaki, the majority of both being
civilian deaths. Thousands of innocent men, women, and children were imme-
diately killed by the blast, and thousands more died over the following weeks
from radiation exposure. On August 12, Japan declared unconditional surrender.
There has been intense debate since the bombing as to whether the decision was
morally acceptable. Public opinion has shifted dramatically over time, with Pew
polls suggesting that 85 percent of Americans approving of the bombing in 1945,
but only 57 percent of Americans approving in 2016. Critics argue that Japan
8 Moral psychology

was already looking for a way to end the war, and that there were other options
that could have forced them to surrender without dropping additional bombs.
For the sake of this discussion, let’s assume that these really were the only two
options: either invade Japan at a cost of millions of lives, or drop bombs on two
mostly civilian populations at a cost of hundreds of thousands of lives. Which is
the right decision?
To those of you familiar with moral philosophy, you’ll recognize a similarity
here to one of the most famous thought-experiments in ethics, called the trolley
problem, originally developed by philosophers Philippa Foot (1967) and Judith
Thomson (1976). In the standard version of the trolley problem, a runaway train
is heading towards five people, but you can save them by diverting the train to
a side-track. Unfortunately, there is a single person on the side-track who will
certainly die as a result of pulling the switch (Figure 1.1).
There’s a reason why the trolley problem has been a useful tool for philoso-
phers and psychologists; dilemmas like these can reveal the inner workings of
the way that we make moral judgments. According to Mikhail’s (2011) extensive
cross-cultural surveys, most people surveyed on this question think the right
decision is to pull the switch. You might think that the reason for pulling the
switch is obvious: pick the action that saves the most lives. However, when pre-
sented with an alternate scenario, where it’s necessary to push a large man in
front of the train to save five people (Figure 1.2), almost everybody rejects it as
morally wrong. This is strange, since the “save the most lives” principle predicts
that pushing a man in front of a train is no different from diverting the train to
kill a pedestrian; both actions are sacrificing one to save five. It’s obvious from
this example that actual moral judgments are more complicated than we thought.
Trolley Problem (Bystander)

States P1 P2 P3 P4 P5 P6
Do Nothing 0 –99 –99 –99 –99 –99
Pull Switch –99 0 0 0 0 0

FIGURE 1.1 
Payoffs in the Trolley Problem (bystander version), measured in terms
of changes from each player’s current state. For instance, 0 means you
experience no change from your current state, while –100 is a maximal
loss from your current state. For now, these numbers don’t matter, we can
call them a percentage of loss. Let’s say that getting hit by a train will lead
to a 99 percent chance of losing everything that’s important. P1 is the
person on the side-track, and P2–P6 are the five people on the main track.
 Moral psychology 9

Trolley Problem (Footbridge)

States P1 P2 P3 P4 P5 P6
Do Nothing 0 –99 –99 –99 –99 –99
Pull Man –99 0 0 0 0 0

FIGURE 1.2 Payoffs in the Trolley Problem (footbridge version). P1 is the man on the
bridge, and P2–P6 are the five people on the main track.

This chapter will give a broad survey of the universal categories and rules that
people use when making moral judgments. We’ll see that moral judgments are a
unique system in the human mind, distinct from social conventions, emotions,
and religious beliefs. I’ll be calling this a functional network to emphasize that our
system for making moral judgments is probably a collection of older psychologi-
cal capacities that have been co-opted over time for a single practical purpose.
Many researchers who are developing ethics algorithms for machines view moral
judgment as a cognitive trait like language, and the obvious way to implement
this trait in a machine is to simulate our moral grammar. I’ll try to show how
our moral psychology carries along with it many useful features, but also many
unfortunate ones that we don’t want to program into robots. The following
chapters will develop a framework for determining exactly which of these fea-
tures are useful or unfortunate.

Moral grammar
Any survey of human universals, like that compiled by anthropologist Donald
Brown (1991), will include the fact that all human beings judge some actions as
wrong and others as permissible. It’s not hard to think of actions that most of us
intuitively categorize as morally wrong: homicide, battery, rape, abuse, discrim-
ination, cheating, and massive deception. These judgments are made quickly,
automatically, and without effort. If I were to ask you why you think that cheat-
ing is wrong, you might be baff led and find it hard to answer. You might even
say: “It just is!” This doesn’t mean that there are no rules for evaluating right and
wrong, but it suggests that these rules are unconscious, like asking someone why
they perceive a table as rectangular or why they accept a sentence as grammatical.
The job of moral psychology is to discover how people understand the categories
of wrong and acceptable, and how they sort actions into these categories.
10 Moral psychology

Skeptics about morality may question whether there is a unique kind of thing
called a moral judgment. Here are some ways of thinking that morality is really
something else entirely:

•• “Morality is nothing more than a set of social conventions like etiquette

and laws!”
•• “Morality is nothing more than a set of religious commands!”
•• “Morality is nothing more than an expression of people’s emotions!”

It’s true that social conventions, religion, and emotions are strongly associated
with moral judgments. But, as social scientists say, correlation isn’t causation. In
order to talk about morality, we’ll start by distinguishing it from these closely
related phenomena.
Social conventions like “men have short hair and women wear dresses” are
used to guide people’s behavior, but the rules themselves are arbitrary and vary
widely from culture to culture. A famous (and probably apocryphal) story of
the Persian Emperor Darius describes how the ancient Indians in his court
judged eating their dead parents to be a sign of great respect, while the ancient
Greeks found it to be a sign of great offense. The typical conclusion from sto-
ries like this is that morality is an arbitrary set of rules, and there’s no common
ground we can use to settle disagreements, so we should just be tolerant of all
moral beliefs.
In response to this story, the philosopher James Rachels (2009) points out
that there is a lot more agreement about moral beliefs than social conventions.
In the story of Darius, both the Indians and Greeks agreed on the more general
rule: honor your parents. They just disagreed about which practice is the best
way of honoring them, eating or not eating them. It’s important to keep in
mind how a person’s empirical beliefs about how the world works can dramati-
cally change her moral judgments. For example, torturing an innocent person
may look like a barbaric practice to us, but if the people doing it genuinely
believe that this action is saving the person’s immortal soul from an eternity
of torture, or saving an entire society from the wrath of the gods, then the
judgment begins to look less crazy. Ask yourself how your moral beliefs would
change if you were to genuinely believe that a cow could contain the soul of
your dead parents, or that some group intentionally caused the diseases your
family is experiencing. Barbaric actions often turn out to be based on the same
moral principles that we endorse, but tragically incorrect information about
how the world works.
Even if a distant society approves of slavery and child abuse, hopefully you
still think these actions are morally wrong. Tolerance is a good attitude to have
about social conventions, but it’s a disastrous policy to have about moral beliefs.
Several decades ago, the developmental psychologist Elliot Turiel (1983) and
his colleagues conducted experiments where some groups of children were told
stories about conventional rule violations (“Is it okay to wear pajamas to school if
 Moral psychology 11

teacher says so?”) and other groups of children were told stories about moral rule
violations (“Is it okay to hit another student if teacher says so?”). Five-year-old
children tend to accept that social norm violations are acceptable when a teacher
approves of it, but actions like hitting another child are typically judged to be
wrong, even if an authority or other community approves.
This distinction between ethics and authority is something advocated by phi-
losophers from Plato to Kant. It’s true that there are laws against attacking stran-
gers, but it would be ridiculous to say that the reason you don’t punch strangers
in the face is that you’re worried about getting arrested! Similarly, there are reli-
gious commands against murder, lying, and stealing, but it’s strange to say that
Christians or Muslims would suddenly become more violent if not for religious
laws. The rate of violent behavior among atheists is no different than the rate
of violence among religious followers, and there is no difference between athe-
ists and religious people in their responses to scenarios like the trolley problem.
In 2011, U.S. Senator Trent Franks accidentally made this point clear when he
defended the motto of the United States, “In God We Trust,” by arguing that
without a trust in God, the country would collapse into violent chaos:

An atheist state is as brutal as the thesis that it rests upon, and there is no
reason for us to gather in this place [the U.S. Chamber of Commerce], we
should just let anarchy prevail, because after all, we are just worm food.

This speech was mocked by The Daily Show host Jon Stewart, who continued to
fill in the logical consequences of the senator’s comment:

I guess what I’m saying here, Mister Speaker, is that this four-word motto
is right now the only thing standing between me and a nihilistic killing
spree of epic proportions. Seriously, I just want to state for the congres-
sional record: I do not know right from wrong.

The audience laughed at this joke because they recognized that basing your
moral beliefs entirely on what someone else tells you is a childish attitude which
leads to absurd conclusions. In a dramatic example, Immanuel Kant (1798) wrote
of the story of Abraham and Isaac:

[In] the myth of the sacrifice that Abraham was going to make by butcher-
ing and burning his only son at God’s command … Abraham should have
replied to this supposedly divine voice: “That I ought not kill my good son
is quite certain. But that you, this apparition, are God – of that I am not
certain, and never can be, not even if this voice rings down to me from the
(visible) heaven.”

Kant was an extremely devoted religious believer, and he’s saying to basically
ignore the commands of God when it comes to ethics. Millions of religious
12 Moral psychology

believers also regularly ignore the teachings of authority figures over issues
they believe are right or wrong. Despite the Catholic Church insisting that
contraception is morally wrong, a 2014 Pew poll found that 79 percent of
Catholics believe it to be permissible. Drastic changes in moral beliefs seem to
have no obvious connection to religious beliefs; between 2007 and 2014, Pew
research shows that all religious and non-religious groups in the United States
changed their views about the permissibility of homosexuality at roughly the
same rate. Even Mormons and Evangelical Christians show a change from
24–26 percent to 36 percent acceptance of homosexuality, despite no clear
difference in the commands of their churches. Religion certainly has an inf lu-
ence on what sorts of information people are exposed to, but there’s no reason
to think this has any greater inf luence on moral beliefs than other sources of
information. There are similar religious differences in beliefs about evolution
and global warming, but it’s silly to think that religion is the cause of beliefs
about global warming, or that beliefs about global warming are nothing more
than religious attitudes.
What about emotions? If I judge an action to be morally wrong, I will almost
certainly feel upset about it, angry at people who do it, and motivated to avoid
doing that action. You would think it’s crazy if someone believes an action is
morally wrong and feels happy about people doing it. Philosophers like David
Hume (1738) have used this connection to argue that emotions are the ultimate
cause of moral judgments. His argument is that morality always involves motiva-
tion, and the only source of motivation is the emotions, so morality must origi-
nate (at least in part) from human emotions. As Hume summarizes:

[I]t is impossible that the distinction betwixt moral good and evil can be
made by reason; since that distinction has an inf luence on our actions, of
which reason alone is incapable.

I agree with Hume that the emotions play an important role in moral judgments,
but there’s a big difference between playing a role in morality and being all there is
to morality. An example of this radical view is found in A.J. Ayer’s theory (1936)
that words like wrong are nothing more than expressions of emotions:

[I]f I say to someone, “you acted wrongly in stealing that money,” … It

is as if I had said, “you stole that money,” in a peculiar tone of horror, or
written if with the addition of some special exclamation marks. The tone,
or the exclamation marks, adds nothing to the literal meaning of the sen-
tence. It merely serves to show that the expression is attended by certain
feelings of that speaker.

Ayer’s speculation about the meaning of words like ought and wrong has turned out
to be incorrect. Our best linguistic theories about the meaning of words like ought
show that they are operators acting like the words all and some. There are patterns
 Moral psychology 13

in the way that people use moral terms that are identical to patterns in the way
people use all and some. For example, everyone accepts the inferences:

1. Every baby gets sick =

It’s false that some babies don’t get sick
2. Not every baby is cute =
Some babies aren’t cute
3. Every baby is loveable Þ
Therefore, (at least) some babies are loveable.

It turns out that if we make the connection: [all = obligatory] and [some = per-
missible], then the same patterns show up in inferences with moral terms:

1. Jonathon must be in his office =

It’s not acceptable for Jonathon not to be in his office
2. Jonathon doesn’t have to be in his office =
It’s acceptable for Jonathon to not be in his office
3. Jonathon must be in his office Þ
Therefore, it’s (at least) acceptable that Jonathon is in his office.1

Why would it be that patterns in moral terms act like all and some? The typical
answer that linguists like Angelika Kratzer (1977) suggest is that these words
involve the same basic operations, called universal and existential quantifica-
tion. Call these operators P for “permissible” and O for “obligated,” so that
O[ Jonathon is in his office] means “Jonathon is obligated to be in his office.”
There is an entire field called deontic logic devoted entirely to the study of these
operators. We won’t get into the details here, but the point is that any informa-
tion under the scope of these kinds of operators has to have a very specific kind of
structure that involves predicates and connectives. This shows that moral judg-
ments must be more than merely emotional responses, although emotions may
play a very important role in determining how certain actions are classified as
permissible or required (this is similar to an argument initially developed by the
philosophers Gottlob Frege and Peter Geach).
Instead of thinking about moral judgments as social conventions, religious
beliefs, or emotional responses, a better analogy is to think of them as a functional
network like human language. In his book, Elements of Moral Cognition (2011),
John Mikhail describes this analogy in detail. Just like our quick and automatic
responses about grammar, judgments about the permissibility of actions are the
product of a set of rules about which speakers are largely unconscious. Mikhail
calls this set of rules moral grammar. The study of moral grammar investigates
the categories and rules used by speakers to move from perceptions of actions to
judgments about permissibility or impermissibility.
One way that the linguistic analogy is helpful is that it shows how we can
appeal to our own intuitions as evidence about the unconscious psychological
14 Moral psychology

processes that produced them. This is one of the many changes that Noam
Chomsky brought to linguistics during the 1950s and 1960s. The analogy also
shows how speakers can be making use of unconscious rules that they themselves
can’t explicitly describe. For example, as competent English speakers, we can
easily transform the sentence John is running into Is John running? but most of us
are incapable of explaining the rule for this: “Move the first aux verb after the
subject to the front; if there is no aux verb, insert a do/does.” This is fascinating:
it’s a rule that we all use but can’t explicitly articulate. Moral rules seem to have
a similar ineffable quality to them; they’re easy to use but hard to explain. As
U.S. Supreme Court Justice Potter Stewart famously remarked about obscenity:

I shall not today attempt further to define the kinds of material I under-
stand to be expressed within that shorthand description [hardcore pornog-
raphy], and perhaps I could never succeed in intelligibly doing so. But I
know it when I see it…

As much as this remark has been mocked over the years, it’s the same quality
that linguistic rules have: I can’t articulate what makes a sentence of my native
language grammatical or ungrammatical, but I know a well-constructed one
when I see it. This isn’t because the rules are part of some magical realm that we
can only detect with extrasensory perception. Instead, it’s because these rules
are structures in our minds and brains; we have conscious access only to their
outputs. With enough careful study, we can hypothesize about and reconstruct
what rules our brains are using to form grammatical sentences and sort actions
into categories of wrong and acceptable.
In human languages, it is truly incredible how a massive amount of variation
can be generated from toggling parameters on a few simple underlying rules.
This insight can be framed as a way of answering the “nature/nurture” question
for a cognitive trait. As Steven Pinker (2002) points out, the boring (and almost
trivially true) answer to this question is “a little bit of both,” but the interesting
answer is describing in detail exactly how innate features of the human mind
enable parts of language to be acquired through experience. According to the
linguistic analogy, some components of our moral judgments, like what objects
are valuable and which effects are harmful, may be acquired through emotional
responses and cultural norms. However, the way that valuable objects and harm-
ful effects are framed within a system of rules may be constrained by only a lim-
ited set of configurations determined by the structure of our moral psychology.
As the cognitive scientist David Marr (1981) argued, there are at least three
distinct levels of explanation for a cognitive trait like language or moral judg-
ment: (1) the way a system is implemented, (2) the categories and rules it uses,
and (3) the goal of the system. For example, a cash register is implemented in
the hardware of the machine, it uses a few basic computational rules, and the
function of the machine is to report prices and exchange money. If morality is a
functional network of the human mind, we want to know how it’s implemented
in the brain, the rules that it’s using to evaluate actions, and what its historical
 Moral psychology 15

function is. I’ll have nothing to say about how moral judgments are implemented
in the human brain, but the rest of this chapter will discuss some of the abstract
entities and rules that the moral network uses, and the next chapter will look at
its evolutionary and historical function.

Elements of moral grammar

Psychologists Fiery Cushman and Joshua Greene (2012) have argued that moral
dilemmas are useful because they smash our normal moral intuitions together,
allowing us to see what pieces they’re made of, like high-energy particle colli-
sions. Researchers like Mikhail predict that manipulating dilemmas will reveal a
small set of elements, like the following 2:

AGENT
PATIENT
INTEND
CAUSE
STATE
HARM

Like any good theory, we want to get the smallest number of elements necessary
to build up all the bigger structures from them. A good theory of chemistry will
have elements like hydrogen and oxygen, then build larger structures like water
molecules. Similarly, the hope of a computational theory of moral grammar is
that all moral beliefs and even more complex concepts like innocence and consent
can be built as molecules from these basic elements.
Let’s start with AGENT and PATIENT. Just like it’s hard to build a sentence
without a noun and a direct object, it’s equally hard to make a moral judgment
without an agent and a patient. Roughly, agents are the ones who perform the
action and patients are the ones who experience its effects. In their book, The
Mind Club (2016), psychologists Daniel Wegner and Kurt Gray describe some
of the features that humans use to detect what objects are agents and patients.
While people reliably identify normal adult humans as both agents and patients,
some objects are only identified as agents (gods and robots), while others are only
identified as patients (cute animals and babies). Wegner and Gray hypothesize
that features like perceived power and control are essential for identifying an
agent, while patients are picked out by movement at a human-like speed, having
human-like facial features, and reacting to pain in a human-like way. A patient
might also be identified because she is similar, familiar, or genetically related.
In a startling example of how genetic relatedness inf luences our moral judg-
ments, the biologist April Bleske-Rechek and her colleagues (2010) used the trol-
ley problem to modulate the relationship between the agent and people harmed,
varying the victims by sex (M/F), age (2, 20, 45, 70), and relatedness (stran-
ger, cousin, uncle/aunt, grandfather/grandmother, son/daughter, brother/sister,
mother/father). Participants were then asked: “Would you f lip the switch in this
16 Moral psychology

situation?” The authors found that “participants were increasingly unwilling to

f lip the switch on targets of increasing levels of genetic relatedness.” It won’t sur-
prise you that people are less willing to switch the track when it’s their own mother
on the side-track. What’s surprising is the way that judgments appear to move so
closely with genetic relatedness. While 77.4 percent of participants answered that
they would f lip the switch on a stranger, only 51.8 percent answered that they
would f lip the switch on a cousin (0.125 genetic relatedness), 47.4 percent answered
that they would f lip the switch on an aunt/uncle, nephew/niece, or grandparent
(0.25 genetic relatedness), and 34.3 percent answered that they would f lip the
switch on a parent or sibling (0.5 genetic relatedness). Relying on genetic related-
ness to identify moral importance can often lead to disastrous consequences; need-
less to say, if we are designing a machine based on our own moral judgments, this
is not the kind of feature that will be helpful to incorporate.
Agents and patients are connected by the element CAUSE. In our original
trolley scenario, most people judge switching the train onto a side-track to be
morally permissible, but pushing a large man in front of the train as a means of
stopping it to be morally wrong, even though they produce the same effects (one
dead to save five). When my students are asked to explain what’s different about
these two scenarios, they often respond that pushing a man in front of a train is
“actually killing him,” while switching the train to a side-track is really just the
train killing him. This sounds a bit strange at first. After all, murderers don’t
usually defend their action by insisting: “I didn’t kill him, the bullet did!” One
exception to this might be Charles Guiteau, who shot President James Garfield,
inf licting what would have normally been a non-fatal wound, but one that the
inept doctors treated with methods that led to Garfield’s death. At trial, Guiteau
famously insisted that he only shot Garfield, but the doctors had killed him. Cases
like these bring in weird features of our perception of causation. It’s true that
Garfield wouldn’t have died if not for Guiteau shooting him. But it’s also true
that other human agents (the doctors) played an important causal role in his death.
As psychologists who study causal reasoning have gradually discovered, there
are many ways that people can think about causation. Sometimes this involves spa-
tial processes like physical contact, but other times it involves thinking about what
happens in alternate possibilities. In the physical-force kind, causation is a physical
contact between entities where a quantity of the cause is transferred to a quantity in
the effect. This is the difference between causing a man to die by pushing him ver-
sus directing a trolley in his direction. In the counterfactual kind of causal reason-
ing, causation is based on whether some outcome happens in alternate possibilities
where the agent didn’t intervene. When I say that Don Corleone caused the death
of someone by ordering a hit, this doesn’t necessarily require physical contact. All
it requires is thinking about possibilities where Don Corleone didn’t order the hit,
and finding that the victim would have otherwise survived. Under the counter-
factual interpretation, people are not only sensitive to whether harmful intentions
actually cause harmful effects, but how close they come to doing so. Cases of negli-
gence, victimless crimes, and attempted homicide are good examples.
 Moral psychology 17

Actions are more than just: AGENT CAUSE PATIENT. Instead, we care
about a part of the agent’s mind doing the work, and we can label that part of
her mind with the category INTEND. Imagine the difference between seeing
someone stumble and fall on a stranger, compared with the same person inten-
tionally knocking the stranger down. Intention is the difference between murder
and manslaughter, between harms done accidentally and purposely. Hurricanes and
other natural disasters create more destruction than any serial killer, yet we don’t
view them as responsible for their actions because they don’t have intentions. It’s
easy enough to recognize how important intentions are, but any lawyer will tell
you that it’s hard to establish when a person has one! We typically use cues like
behavior, statements, and character traits to establish intent, but these aren’t always
reliable. Intention seems to be more than just a desire or motivation, and somehow
connected to actual plans in a concrete way.
Intentions are not only important for agents, but also for determining which
states are bad for patients. The concept of consent is produced when we consider
the intentions of a patient. Specifically, when a patient intends to be in the state
that the agent causes him to be in, that’s probably what most speakers mean by an
action being consensual. Actions like euthanasia, employment, and sex between
willing partners are often different from murder, slavery, and rape entirely on the
basis of consent. Just like verbal cues, behavior, and character are used to establish
the intentions of the agent, the same kinds of evidence are often used to establish
the consent of the patient. It should be noted that establishing a patient’s consent
is just as tricky as establishing an agent’s intention, and many debates in ethics and
law surround the conditions under which consent is or isn’t present.
Finally, in addition to agents, patients, causes, and intentions, people are sensitive
to an element that can be called HARM. The trolley problem is limited in that the
same type of harm is always involved (physical harm). You might suspect that moral
judgments often involve different types of harm, and you’d be correct. Some actions
result in psychological and emotional damage, or destruction to people’s reputation,
social standing, or relationships. Lying, cheating, and stealing are often judged to be
morally wrong, even if they don’t necessarily involve direct physical harm. To give
a physics analogy, the element HARM is more like a vector than a scalar: it doesn’t
just have a magnitude but also a direction in any number of possible dimensions of
harm. If you thought that it was difficult to give precise conditions for what counts
as an agent and a patient, or when intentions are present, you’ll find it just as difficult
to define and measure dimensions and magnitudes of harm.
The psychologist Jonathan Haidt (2012) denies that harm is an essential element
in moral thinking, insisting that Western academics have narrowed the definition
of morality to only those judgments involving harm but neglecting other important
features of moral judgments like purity, authority, and sanctity. In my view, what
we’re calling HARM is not in conf lict with disgust, since the two exist at different
levels of explanation. HARM is an element in moral grammar used to measure
the damage done to a patient; it can be instantiated by evaluations of physical suf-
fering from a wide range of sources, including a projection of suffering from the
18 Moral psychology

evaluator’s own disgust and purity judgments. This is where Hume was correct;
people typically use their own emotional responses to evaluate harm. However, it’s
important that these emotions are also projected onto the agent, which is the dif-
ference between “that’s gross” and “that’s wrong.” Experiments by Kurt Gray and
colleagues (2014) suggest that even moral evaluations based on purity are always
projected onto an implicit victim who suffers some damage. These victims might
be people in alternate imagined realities, like the potential victims of drunk driv-
ers or the potential children that might have existed from continuing a pregnancy.
The patient is viewed as having a part of their identity damaged or destroyed,
even if this entity is entirely a projection of the speaker’s own emotional responses.
Unfortunately, this is another inconsistent and inaccurate way humans apply moral
rules that often leads to disastrous consequences.

Rules in moral grammar

If there’s one rule that’s consistent across the varieties of moral grammars, it’s a
rule against harmful battery. This is a point emphasized by John Mikhail (2014),
who notes the consistencies across legal systems that define battery as: an agent
with a bad intention coming into contact with a non-consenting patient and producing bad
consequences for that patient. Representing this definition with our elements of
moral grammar looks like the following:

Harmful battery
An agent’s intention causes a patient to be in a state by physical contact,
the state is physically damaging to the patient, and the patient doesn’t
intend to be in that state.3

There are three clauses in this definition (separated by and): one involving causa-
tion, the other involving harm, and the third involving consent. For instance, a
criminal stabbing a random innocent person in the leg is an obvious case of battery,
but a surgeon creating an incision in a patient’s leg is not battery (even though it’s
causing the same state), because the patient has consented to the contact and it’s
not perceived as harm. The bystander who switches the train to a side-track causes
the person on the side-track to be in a state of harm by pulling the switch, but this
doesn’t count as harmful battery, because it wasn’t caused by direct physical force.
You can already start to imagine all the possible variations of battery that can be
generated by toggling the settings on these elements. Many of these are also estab-
lished legal entities like “offensive battery,” which toggles settings on the kinds of
harm (physical vs. psychological):

Offensive battery
An agent’s intention causes a patient to be in a state by physical con-
tact, the state is psychologically distressing to the patient, and the patient
doesn’t intend to be in that state.4
 Moral psychology 19

We could also toggle the settings on CAUSE between physical contact and
something more indirect, where an agent causes physical harm to the patient
through not performing an action, which is called an omission. The agent per-
forms an omission whenever she could have intervened in nearby counterfactual
situations but doesn’t, like allowing someone to drown when she could easily
save them:

Harmful negligence
An agent’s intention causes a patient to be in a state by not performing
an action, the state is physically damaging to the patient, and the patient
doesn’t intend to be in that state.5

We won’t get into the details about all these variations. I’ll just note that they are
much like the variations in natural languages: minor changes on a shared deep
structure. We also won’t worry too much about the exact structure of the concepts
and rules. I agree with Mikhail that battery is a paradigm case of actions judged to
be wrong, and that other wrong actions like homicide and rape can be generated by
small variations in the actions defined here. It’s also useful to show how this can gen-
erate the idea of an innocent person with respect to battery: a person is innocent of
battery when they haven’t intentionally caused it, even if they might still intend the
patient to be in a bad state. Importantly, just wanting someone to experience harm
might be permissible, but causing them to experience harm is wrong. As fictional
detectives emphasize, there’s nothing illegal about just wanting someone dead.
Under what conditions are actions like battery and homicide judged to be
wrong? We could define a simple rule: “If [harmful battery], then wrong,” but
that’s not enough to get at the complexities of people’s beliefs. Almost everyone
allows for battery in cases of punishment and self-defense. A more sophisticated
rule that includes these cases would be a conditional that restricts the prohibition
on battery to innocent people. Call this the Intentional Harm Rule:

Intentional Harm Rule

If an agent has not committed battery on you, then it’s not permissible for
you to commit battery on that agent.6

This rule essentially says: don’t intentionally cause harm to innocent people
without their consent. But our definition doesn’t specify anything about guilty
people, and we wanted to explain things like punishment and self-defense.
Thus, we need to construct another rule within our system that accounts for the
permissibility of retribution; call it the Retribution Rule:

Retribution Rule
If an agent has caused battery to you, then it is permissible to cause battery
to that agent.7
20 Moral psychology

The Retribution Rule can be toggled to form many varieties, like self-defense,
revenge, and punishment. Despite the fact that the biblical Jesus claims that
both punishment and revenge are impermissible, most Christians have histori-
cally ignored this teaching in their own moral judgments. This is yet another
example of religious believers ignoring the commands of authorities about
moral judgments.
Let’s assume that these two rules, along with all the associated variations
produced by toggling parameters, adequately explain human moral judgments.
Following the linguistic analogy, we can think of these variations as different
moral languages that are each derived from a set of universal elements.

Moral grammar machines

It’s not hard to imagine using a combination of top-down and bottom-up
approaches to turn our universal moral grammar into a computer program; each
element has independently been successfully modeled. For decades, logicians and
computer scientists have been developing deontic logic programs that can put into
practice our intuitive reasoning about permissions and obligations. Wieringa and
Meyer (2012) present a detailed overview of the ways in which deontic logic pro-
grams have been applied to domains like parking permits, tax law, and citizen-
ship policies. The economist Susan Athey (2015) describes how machine learning
techniques are currently being used to make extremely accurate causal predictions
about the likely effects of administering a drug, merging two companies, or raising
the price of a good. Categories like who counts as an AGENT or PATIENT can
be easily modeled using a sufficiently large data set of human responses. However,
the larger question that must be addressed when bringing these elements together
into an artificial moral grammar is: should we be taking this approach to designing
an ethical robot? Do we want robot ethics to be just another incarnation of our
own moral judgments?
Consider a recent article from the journal Cognition titled “Learning a
Commonsense Moral Theory” (2017), co-authored by Max Kleiman-Weiner,
Rebecca Saxe, and Joshua Tenenbaum. In this article, the authors attempt to train a
learning algorithm to develop something like the moral grammar described here.
One component of their model is the evaluation of what I’ve been calling the
PATIENT category. How will this system evaluate which people are valuable, and
how much value is assigned to each person? The authors write:

We start by supposing that through the course of one’s life, one will acquire
attachments for various people or even groups of people. These attachments
and feelings can be represented through the vector introduced in the previ-
ous section. As mentioned in the introduction, these values could come from
empathy and emotional responses, imagination and stories, morally charged
analogical deliberation, love, contact, exposure etc.
Other documents randomly have
different content
“Why?” and the worthy Morgeson laughed sweetly—“I see, my dear
Mr Tempest, you are like most men of genius—you do not
understand business. The reason why we give the first two hundred
and fifty copies away is in order to be able to announce at once in all
the papers that ‘The First Large Edition of the New Novel by
Geoffrey Tempest being exhausted on the day of publication, a
Second is in Rapid Preparation.’ You see we thus hoodwink the
public, who of course are not in our secrets, and are not to know
whether an edition is two hundred or two thousand. The Second
Edition will of course be ready behind the scenes, and will consist of
another two hundred and fifty.”
“Do you call that course of procedure honest?” I asked quietly.
“Honest? My dear sir! Honest?” And his countenance wore a
virtuously injured expression—“Of course it is honest! Look at the
daily papers! Such announcements appear every day—in fact they
are getting rather too common. I freely admit that there are a few
publishers here and there who stick up for exactitude and go to the
trouble of not only giving the number of copies in an Edition, but
also publishing the date of each one as it was issued,—this may be
principle if they like to call it so, but it involves a great deal of
precise calculation and worry! If the public like to be deceived, what
is the use of being exact! Now, to resume,—your second edition will
be sent off ‘on sale or return’ to provincial booksellers, and then we
shall announce—“In consequence of the Enormous Demand for the
new novel by Geoffrey Tempest, the Large Second Edition is out of
print. A Third will be issued in the course of next week.” And so on,
and so on, till we get to the sixth or seventh edition (always
numbering two hundred and fifty each) in three volumes; perhaps
we
99 can by skilful management work it to a tenth. It is only a

question of diplomacy and a little dexterous humbugging of the

trade. Then we shall arrive at the one-volume issue, which will
require different handling. But there’s time enough for that. The
frequent advertisements will add to the expense a bit, but if you
don’t mind—”
“I don’t mind anything,” I said—“so long as I have my fun.”
“Your fun?” he queried surprisedly—“I thought it was fame you
wanted, more than fun!”
I laughed aloud.
“I’m not such a fool as to suppose that fame is secured by
advertisement,” I said—“For instance I am one of those who think
the fame of Millais as an artist was marred when he degraded
himself to the level of painting the little green boy blowing bubbles
of Pears’s Soap. That was an advertisement. And that very incident
in his career, trifling though it seems, will prevent his ever standing
on the same dignified height of distinction with such masters in art
as Romney, Sir Peter Lely, Gainsborough or Reynolds.”
“I believe there is a great deal of justice in what you say;” and
Morgeson shook his head wisely—“Viewed from a purely artistic and
sentimental standpoint you are right.” And he became suddenly
downcast and dubious. “Yes,—it is a most extraordinary thing how
fame does escape people sometimes just when they seem on the
point of grasping it. They are ‘boomed’ in every imaginable way, and
yet after a time nothing will keep them up. And there are others
again who get kicked and buffeted and mocked and derided——”
“Like Christ?” I interposed with a half smile. He looked shocked,—he
was a Non-conformist,—but remembering in time how rich I was, he
bowed with a meek patience.
“Yes”—and he sighed—“as you suggest, Mr Tempest, like Christ.
Mocked and derided and opposed at every turn,—and yet by the
queerest caprice of destiny, they succeed in winning a world-wide
fame and power——”
“Like
100 Christ again!” I said mischievously, for I loved to jar his non-
conformist conscience.
“Exactly!” He paused, looking piously down. Then with a return of
secular animation he added—“But I was not thinking of the Great
Example just then, Mr Tempest—I was thinking of a woman.”
“Indeed!” I said indifferently.
“Yes—a woman, who despite continued abuse and opposition is
rapidly becoming celebrated. You are sure to hear of her in literary
and social circles”—and he gave me a furtive glance of doubtful
inquiry—“but she is not rich, you know,—only famous. However,—we
have nothing to do with her just now—so let us return to business.
The one uncertain point in the matter of your book’s success is the
attitude of the critics. There are only six leading men who do the
reviews, and between them they cover all the English magazines and
some of the American too, as well as the London papers. Here are
their names”—and he handed me a pencilled memorandum,—“and
their addresses, as far as I can ascertain them, or the addresses of
the papers for which they most frequently write. The man at the
head of the list, David McWhing, is the most formidable of the lot.
He writes everywhere about everything,—being a Scotchman he’s
bound to have his finger in every pie. If you can secure McWhing,
you need not trouble so much about the others, as he generally
gives the ‘lead,’ and has his own way with the editors. He is one of
the ‘personal friends’ of the editor of the Nineteenth Century for
example, and you would be sure to get a notice there, which would
otherwise be impossible. No reviewer can review anything for that
magazine unless he is one of the editor’s friends.2 You must manage
McWhing, or he might, just for the sake of ‘showing off,’ cut you up
rather roughly.”
“That would not matter,” I said, diverted at the idea of ‘managing
McWhing,’—“A little slating always helps a book to sell.”
“In
101 some cases it does,”—and Morgeson stroked his thin beard

perplexedly—“But in others it most emphatically does not. Where

there is any very decided or daring originality, adverse criticism is
always the most effective. But a work like yours requires fostering
with favour,—wants ‘booming’ in short——”
“I see!” and I felt distinctly annoyed—“You don’t think my book
original enough to stand alone?”
“My dear sir!—you are really—really—! what shall I say?” and he
smiled apologetically—“a little brusque? I think your book shows
admirable scholarship and delicacy of thought,—if I find fault with it
at all, it is perhaps because I am dense. The only thing it lacks in my
opinion is what I should call tenaciousness, for want of a better
expression,—the quality of holding the reader’s fancy fixed like a
nail. But after all this is a common failing of modern literature; few
authors feel sufficiently themselves to make others feel.”
I made no reply for a moment. I was thinking of Lucio’s remarks on
this very same subject.
“Well!” I said at last—“If I had no feeling when I wrote the book, I
certainly have none now. Why man, I felt every line of it!—painfully
and intensely!”
“Ay, ay indeed!” said Morgeson soothingly—“Or perhaps you thought
you felt, which is another very curious phase of the literary
temperament. You see, to convince people at all, you must first
yourself be convinced. The result of this is generally a singular
magnetic attraction between author and public. However I am a bad
hand at argument,—and it is possible that in hasty reading I may
have gathered a wrong impression of your intentions. Anyhow the
book shall be a success if we can make it so. All I venture to ask of
you is that you should personally endeavour to manage McWhing!”
I promised to do my best, and on this understanding we parted. I
realised that Morgeson was capable of greater discernment than I
had imagined, and his observations had given me material for
thought
102 which was not altogether agreeable. For if my book, as he
said, lacked tenacity, why then it would not take root in the public
mind,—it would be merely the ephemeral success of a season,—one
of those brief ‘booms’ in literary wares for which I had such
unmitigated contempt,—and Fame would be as far off as ever,
except that spurious imitation of it which the fact of my millions had
secured. I was in no good humour that afternoon, and Lucio saw it.
He soon elicited the sum and substance of my interview with
Morgeson, and laughed long and somewhat uproariously over the
proposed ‘managing’ of the redoubtable McWhing. He glanced at the
five names of the other leading critics and shrugged his shoulders.
“Morgeson is quite right,”—he said—“McWhing is intimate with the
rest of these fellows—they meet at the same clubs, dine at the same
cheap restaurants and make love to the same painted ballet-girls. All
in a comfortable little fraternal union together, and one obliges the
other on their several journals when occasion offers. Oh yes! I
should make up to McWhing if I were you.”
“But how?” I demanded, for though I knew McWhing’s name well
enough having seen it signed ad nauseam to literary articles in
almost every paper extant, I had never met the man; “I cannot ask
any favour of a press critic.”
“Of course not!” and Lucio laughed heartily again—“If you were to
do such an idiotic thing what a slating you’d get for your pains!
There’s no sport a critic loves so much as the flaying of an author
who has made the mistake of lowering himself to the level of asking
favours of his intellectual inferiors! No, no, my dear fellow!—we shall
manage McWhing quite differently,—I know him, though you do not.”
“Come, that’s good news!” I exclaimed—“Upon my word Lucio, you
seem to know everybody.”
“I think I know most people worth knowing—“ responded Lucio
quietly—“Though I by no means include Mr McWhing in the category
of worthiness. I happened to make his personal acquaintance in a
somewhat singular and exciting manner. It was in Switzerland, on
that
103 awkward ledge of rock known as the Mauvais Pas. I had been

some weeks in the neighbourhood on business of my own, and

being surefooted and fearless, was frequently allowed by the guides
to volunteer my services with theirs. In this capacity of amateur
guide, capricious destiny gave me the pleasure of escorting the timid
and bilious McWhing across the chasms of the Mer de Glace, and I
conversed with him in the choicest French all the while, a language
of which, despite his boasted erudition, he was deplorably ignorant.
I knew who he was I must tell you, as I know most of his craft, and
had long been aware of him as one of the authorised murderers of
aspiring genius. When I got him on the Mauvais Pas, I saw that he
was seized with vertigo; I held him firmly by the arm and addressed
him in sound strong English thus—‘Mr McWhing, you wrote a
damnable and scurrilous article against the work of a certain poet’
and I named the man—‘an article that was a tissue of lies from
beginning to end, and which by its cruelty and venom embittered a
life of brilliant promise, and crushed a noble spirit. Now, unless you
promise to write and publish in a leading magazine a total
recantation of this your crime when you get back to England,—if you
get back!—giving that wronged man the ‘honourable mention’ he
rightly deserves,—down you go! I have but to loosen my hold!’
Geoffrey, you should have seen McWhing then! He whined, he
wriggled, he clung! Never was an oracle of the press in such an
unoracular condition. ‘Murder!—murder!’ he gasped, but his voice
failed him. Above him towered the snow peaks like the summits of
that Fame he could not reach and therefore grudged to others,—
below him the glittering ice-waves yawned in deep transparent
hollows of opaline blue and green,—and afar off the tinkling cowbells
echoed through the still air, suggestive of safe green pastures and
happy homes. ‘Murder!’ he whispered gurglingly. ‘Nay!’ said I, ‘’tis I
should cry Murder!—for if ever an arresting hand held a murderer,
mine holds one now! Your system of slaying is worse than that of
104 midnight assassin, for the assassin can but kill the body,—you
the
strive to kill the soul. You cannot succeed, ’tis true, but the mere
attempt is devilish. No shouts, no struggles will serve you here,—we
are alone with Eternal Nature,—give the man you have slandered his
tardy recognition, or else, as I said before—down you go!’ Well, to
make my story short, he yielded, and swore to do as I bade him,—
whereupon placing my arm round him as though he were my tender
twin-brother, I led him safely off the Mauvais Pas and down the
kindlier hill, where, what with the fright and the remains of vertigo,
he fell a’weeping grievously. Would you believe it, that before we
reached Chamounix we had become the best friends in the world?
He explained himself and his rascally modes of action, and I nobly
exonerated him,—we exchanged cards,—and when we parted, this
same author’s bug-bear McWhing, overcome with sentiment and
whisky toddy (he is a Scotchman you know) swore that I was the
grandest fellow in the world, and that if ever he could serve me he
would. He knew my princely title by this time, but he would have
given me a still higher name. ‘You are not—hic—a poet yourself?’ he
murmured, leaning on me fondly as he rolled to bed. I told him no.
‘I am sorry—very!’ he declared, the tears of whisky rising to his eyes,
‘If you had been I would have done a great thing for you,—I would
have boomed you,—for nothing!’ I left him snoring nobly, and saw
him no more. But I think he’ll recognize me, Geoffrey;—I’ll go and
look him up personally. By all the gods!—if he had only known Who
held him between life and death upon the Mauvais Pas!”
I stared, puzzled.
“But he did know”—I said—“Did you not say you exchanged cards?”
“True, but that was afterwards!” and Lucio laughed; “I assure you,
my dear fellow, we can ‘manage’ McWhing!”
I was intensely interested in the story as he told it,—he had such a
dramatic
105 way of speaking and looking, while his very gestures
brought the whole scene vividly before me like a picture. I spoke out
my thought impulsively.
“You would certainly have made a superb actor, Lucio!”
“How do you know I am not one?” he asked with a flashing glance,
then he added quickly—“No,—there is no occasion to paint the face
and prance over the boards before a row of tawdry footlights like the
paid mimes, in order to be histrionically great. The finest actor is he
who can play the comedy of life perfectly, as I aspire to do. To walk
well, talk well, smile well, weep well, groan well, laugh well—and die
well!—it is all pure acting,—because in every man there is the dumb
dreadful immortal Spirit who is real,—who cannot act,—who Is,—and
who steadily maintains an infinite though speechless protest against
the body’s Lie!”
I said nothing in answer to this outburst,—I was beginning to be
used to his shifting humours and strange utterances,—they
increased the mysterious attraction I had for him, and made his
character a perpetual riddle to me which was not without its subtle
charm. Every now and then I realized, with a faintly startled sense of
self-abasement, that I was completely under his dominance,—that
my life was being entirely guided by his control and suggestion,—but
I argued with myself that surely it was well it should be so, seeing
he had so much more experience and influence than I. We dined
together that night as we often did, and our conversation was
entirely taken up with monetary and business concerns. Under
Lucio’s advice I was making several important investments, and
these matters gave us ample subject for discussion. At about eleven
o’clock, it being a fine frosty evening and fit for brisk walking, we
went out, our destination being the private gambling club to which
my companion had volunteered to introduce me as a guest. It was
situated at the end of a mysterious little back street, not far from the
respectable precincts of Pall-Mall, and was an unpretentious looking
house enough outside, but within, it was sumptuously though
tastelessly
106 furnished. Apparently, the premises were presided over
by a woman,—a woman with painted eyes and dyed hair who
received us first of all within the lamp-lighted splendours of an
Anglo-Japanese drawing-room. Her looks and manner undisguisedly
proclaimed her as a demi-mondaine of the most pronounced type,—
one of those ‘pure’ ladies with a ‘past’ who are represented as such
martyrs to the vices of men. Lucio said something to her apart,—
whereupon she glanced at me deferentially and smiled,—then rang
the bell. A discreet looking man-servant in sober black made his
appearance, and at a slight sign from his mistress who bowed to me
as I passed her, proceeded to show us upstairs. We trod on a carpet
of the softest felt,—in fact I noticed that everything was rendered as
noiseless as possible in this establishment, the very doors being
covered with thick baize and swinging on silent hinges. On the upper
landing, the servant knocked very cautiously at a side-door,—a key
turned in the lock, and we were admitted into a long double room,
very brilliantly lit with electric lamps, which at a first glance seemed
crowded with men playing at rouge et noir and baccarat. Some
looked up as Lucio entered and nodded smilingly,—others glanced
inquisitively at me, but our entrance was otherwise scarcely noticed.
Lucio drawing me along by the arm, sat down to watch the play,—I
followed his example and presently found myself infected by the
intense excitement which permeated the room like the silent tension
of the air before a thunderstorm. I recognised the faces of many
well known public men,—men eminent in politics and society whom
one would never have imagined capable of supporting a gambling
club by their presence and authority. But I took care to betray no
sign of surprise, and quietly observed the games and the gamesters
with almost as impassive a demeanour as that of my companion. I
was prepared to play and to lose,—I was not prepared however for
the strange scene which was soon to occur and in which I, by force
of circumstances was compelled to take a leading part.

2 The author has Mr Knowles’s own written authority for this fact. Back
107
X

A S soon as the immediate game we were watching was finished,

the players rose, and greeted Lucio with a good deal of
eagerness and effusion. I instinctively guessed from their manner
that they looked upon him as an influential member of the club, a
person likely to lend them money to gamble with, and otherwise to
oblige them in various ways, financially speaking. He introduced me
to them all, and I was not slow to perceive the effect my name had
upon most of them. I was asked if I would join in a game of
baccarat, and I readily consented. The stakes were ruinously high,
but I had no need to falter for that. One of the players near me was
a fair-haired young man, handsome in face and of aristocratic
bearing,—he had been introduced to me as Viscount Lynton. I
noticed him particularly on account of the reckless way he had of
doubling his stakes suddenly and apparently out of mere bravado,
and when he lost, as he mostly did, he laughed uproariously as
though he were drunk or delirious. On first beginning to play I was
entirely indifferent as to the results of the game, caring nothing at
all as to whether I had losses or gains. Lucio did not join us, but sat
apart, quietly observant, and watching me, so I fancied, more than
anyone. And as chance would have it, all the luck came my way, and
I won steadily. The more I won the more excited I became, till
presently my humour changed and I was seized by a whimsical
desire to lose. I suppose it was the touch of some better impulse in
my
108 nature that made me wish this for young Lynton’s sake. For he

seemed literally maddened by my constant winnings, and continued

his foolhardy and desperate play,—his young face grew drawn and
sharply thin, and his eyes glittered with a hungry feverishness. The
other gamesters, though sharing in his run of ill-luck, seemed better
able to stand it, or perhaps they concealed their feelings more
cleverly,—anyhow I know I caught myself very earnestly wishing that
this devil’s luck of mine would desert me and set in the young
Viscount’s direction. But my wishes were no use,—again and again I
gathered up the stakes, till at last the players rose, Viscount Lynton
among them.
“Well, I’m cleaned out!” he said, with a loud forced laugh. “You must
give me my chance of a revanche to-morrow, Mr Tempest!”
I bowed.
“With pleasure!”
He called a waiter at the end of the room to bring him a brandy and
soda, and meanwhile I was surrounded by the rest of the men, all of
them repeating the Viscount’s suggestion of a ‘revanche,’ and
strenuously urging upon me the necessity of returning to the club
the next night in order to give them an opportunity of winning back
what they had lost. I readily agreed, and while we were in the midst
of talk, Lucio suddenly addressed young Lynton.
“Will you make up another game with me?” he inquired. “I’ll start
the bank with this,”—and he placed two crisp notes of five hundred
pounds each on the table.
There was a moment’s silence. The Viscount was thirstily drinking
his brandy-and-soda, and glanced over the rim of his tall tumbler at
the notes with covetous bloodshot eyes,—then he shrugged his
shoulders indifferently. “I can’t stake anything,” he said; “I’ve already
told you I’m cleaned out,—‘stony-broke,’ as the slang goes. It’s no
use my joining.”
“Sit down, sit down, Lynton!” urged one man near him. “I’ll lend you
enough to go on with.”
“Thanks,
109 I’d rather not!” he returned, flushing a little. “I’m too much
in your debt already. Awfully good of you all the same. You go on,
you fellows, and I’ll watch the play.”
“Let me persuade you Viscount Lynton,” said Lucio, looking at him
with his dazzling inscrutable smile—“just for the fun of the thing! If
you do not feel justified in staking money, stake something trifling
and merely nominal, for the sake of seeing whether the luck will
turn”—and here he took up a counter—“This frequently represents
fifty pounds,—let it represent for once something that is not valuable
like money,—your soul, for example!” A burst of laughter broke from
all the men. Lucio laughed softly with them.
“We all have, I hope, enough instruction in modern science to be
aware that there is no such thing as a soul in existence”—he
continued. “Therefore, in proposing it as a stake for this game at
baccarat, I really propose less than one hair of your head, because
the hair is a something, and the soul is a nothing! Come! Will you
risk that non-existent quantity for the chance of winning a thousand
pounds?”
The Viscount drained off the last drop of brandy, and turned upon
us, his eyes flushing mingled derision and defiance.
“Done!” he exclaimed; whereupon the party sat down.
The game was brief,—and in its rapid excitement, almost breathless.
Six or seven minutes sufficed, and Lucio rose, the winner. He smiled
as he pointed to the counter which had represented Viscount
Lynton’s last stake.
“I have won!” he said quietly. “But you owe me nothing, my dear
Viscount, inasmuch as you risked—Nothing! We played this game
simply for fun. If souls had any existence of course I should claim
yours;—I wonder what I should do with it by the way!” He laughed
good-humouredly. “What nonsense, isn’t it!—and how thankful we
ought to be that we live in advanced days like the present, when
such silly superstitions are being swept aside by the march of
progress and pure Reason! Good-night! Tempest and I will give you,
your
110 full revenge to-morrow,—the luck is sure to change by then,
and you will probably have the victory. Again—good-night!”
He held out his hand,—there was a peculiar melting tenderness in
his brilliant dark eyes,—an impressive kindness in his manner.
Something—I could not tell what—held us all for the moment
spellbound as if by enchantment, and several of the players at other
tables, hearing of the eccentric stake that had been wagered and
lost, looked over at us curiously from a distance. Viscount Lynton,
however, professed himself immensely diverted, and shook Lucio’s
proffered hand heartily.
“You are an awfully good fellow!” he said, speaking a little thickly
and hurriedly—“And I assure you seriously if I had a soul I should be
very glad to part with it for a thousand pounds at the present
moment. The soul wouldn’t be an atom of use to me and the
thousand pounds would. But I feel convinced I shall win to-morrow!”
“I am equally sure you will!” returned Lucio affably, “In the
meantime, you will not find my friend here, Geoffrey Tempest, a
hard creditor,—he can afford to wait. But in the case of the lost
soul,”—here he paused, looking straight into the young man’s eyes,
—“of course I cannot afford to wait!”
The Viscount smiled vaguely at this pleasantry, and almost
immediately afterwards left the club. As soon as the door had closed
behind him, several of the gamesters exchanged sententious nods
and glances.
“Ruined!” said one of them in a sotto-voce.
“His gambling debts are more than he can ever pay”—added another
—“And I hear he has lost a clear fifty thousand on the turf.”
These remarks were made indifferently, as though one should talk of
the weather,—no sympathy was expressed,—no pity wasted. Every
gambler there was selfish to the core, and as I studied their
hardened faces, a thrill of honest indignation moved me,—
indignation mingled with shame. I was not yet altogether callous or
cruel-hearted,
111 though as I look back upon those days which now
resemble a wild vision rather than a reality, I know that I was
becoming more and more of a brutal egoist with every hour I lived.
Still I was so far then from being utterly vile, that I inwardly resolved
to write to Viscount Lynton that very evening, and tell him to
consider his debt to me cancelled, as I should refuse to claim it.
While this thought was passing through my mind, I met Lucio’s gaze
fixed steadily upon me. He smiled,—and presently signed to me to
accompany him. In a few minutes we had left the club, and were out
in the cold night air under a heaven of frostily sparkling stars.
Standing still for a moment, my companion laid his hand on my
shoulder.
“Tempest, if you are going to be kind-hearted or sympathetic to
undeserving rascals, I shall have to part company with you!” he said,
with a curious mixture of satire and seriousness in his voice—“I see
by the expression of your face that you are meditating some silly
disinterested action of pure generosity. Now you might just as well
flop down on these paving stones and begin saying prayers in public.
You want to let Lynton off his debt,—you are a fool for your pains.
He is a born scoundrel,—and has never seen his way to being
anything else,—why should you compassionate him? From the time
he first went to college till now, he has been doing nothing but live a
life of degraded sensuality,—he is a worthless rake, less to be
respected than an honest dog!”
“Yet some one loves him I daresay!” I said.
“Some one loves him!” echoed Lucio with inimitable disdain—“Bah!
Three ballet girls live on him if that is what you mean. His mother
loved him,—but she is dead,—he broke her heart. He is no good I
tell you,—let him pay his debt in full, even to the soul he staked so
lightly. If I were the devil now, and had just won the strange game
we played to-night, I suppose according to priestly tradition, I should
be piling up the fire for Lynton in high glee,—but being what I am, I
say let the man alone to make his own destiny,—let things take their
course,—and as he chose to risk everything, so let him pay
everything.”
We
112 were by this time walking slowly into Pall Mall,—I was on the

point of making some reply, when catching sight of a man’s figure on

the opposite side of the way, not far from the Marlborough Club, I
uttered an involuntary exclamation.
“Why there he is!” I said—“there is Viscount Lynton!”
Lucio’s hand closed tightly on my arm.
“You don’t want to speak to him now, surely!”
“No. But I wonder where he’s going? He walks rather unsteadily.”
“Drunk, most probably!”
And Lucio’s face presented the same relentless expression of scorn I
had so often seen and marvelled at.
We paused a moment, watching the Viscount strolling aimlessly up
and down in front of the clubs,—till all at once he seemed to come
to a sudden resolution, and stopping short, he shouted,
“Hansom!”
A silent-wheeled smart vehicle came bowling up immediately. Giving
some order to the driver, he jumped in. The cab approached swiftly
in our direction,—just as it passed us the loud report of a pistol
crashed on the silence.
“Good God!” I cried reeling back a step or two—“He has shot
himself!”
The hansom stopped,—the driver sprang down,—club-porters,
waiters, policemen and no end of people starting up from Heaven
knows where, were on the scene on an instant,—I rushed forward to
join the rapidly gathering throng, but before I could do so, Lucio’s
strong arm was thrown round me, and he dragged me by main force
away.
“Keep cool, Geoffrey!” he said—“Do you want to be called up to
identify? And betray the club and all its members? Not while I am
here to prevent you! Check your mad impulses, my good fellow,—
they will lead you into no end of difficulties. If the man’s dead he’s
dead, and there’s an end of it.”
“Lucio! You have no heart!” I exclaimed, struggling violently to
escape
113 from his hold—“How can you stop to reason in such a case!
Think of it! I am the cause of all the mischief!—it is my cursed luck
at baccarat this evening that has been the final blow to the wretched
young fellow’s fortunes,—I am convinced of it!—I shall never forgive
myself—”
“Upon my word, Geoffrey, your conscience is very tender!” he
answered, holding my arm still more closely and hurrying me away
despite myself—“You must try and toughen it a little if you want to
be successful in life. Your ‘cursed luck’ you think, has caused
Lynton’s death? Surely it is a contradiction in terms to call luck
‘cursed,’—and as for the Viscount, he did not need that last game at
baccarat to emphasise his ruin. You are not to blame. And for the
sake of the club, if for nothing else, I do not intend either you or
myself to be mixed up in a case of suicide. The coroner’s verdict
always disposes of these incidents comfortably in two words
—‘Temporary insanity.’”
I shuddered. My soul sickened as I thought that within a few yards
of us was the bleeding corpse of the man I had so lately seen alive
and spoken with,—and notwithstanding Lucio’s words I felt as if I
had murdered him.
“‘Temporary insanity’”—repeated Lucio again, as if speaking to
himself—“all remorse, despair, outraged honour, wasted love,
together with the scientific modern theory of Reasonable
Nothingness—Life a Nothing, God a Nothing,—when these drive the
distracted human unit to make of himself also a nothing, ‘temporary
insanity’ covers up his plunge into the infinite with an untruthful
pleasantness. However, after all, it is as Shakespeare says, a mad
world!”
I made no answer. I was too overcome by my own miserable
sensations. I walked along almost unconscious of movement, and as
I stared bewilderedly up at the stars they danced before my sight
like fireflies whirling in a mist of miasma. Presently a faint hope
occurred to me.
“Perhaps,” I said, “he has not really killed himself? It may be only an
attempt?”
“He
114 was a capital shot”—returned Lucio composedly,—“That was his

one quality. He has no principles,—but he was a good marksman. I

cannot imagine his missing aim.”
“It is horrible! An hour ago alive, ... and now ... I tell you, Lucio, it is
horrible!”
“What is? Death? It is not half so horrible as Life lived wrongly,”—he
responded, with a gravity that impressed me in spite of my emotion
and excitement—“Believe me, the mental sickness and confusion of
a wilfully degraded existence are worse tortures than are contained
in the priestly notions of Hell. Come come, Geoffrey, you take this
matter too much to heart,—you are not to blame. If Lynton has
given himself the ‘happy dispatch’ it is really the best thing he could
do,—he was of no use to anybody, and he is well out of it. It is
positively weak of you to attach importance to such a trifle. You are
only at the beginning of your career——”
“Well, I hope that career will not lead me into any more such
tragedies as the one enacted to-night,”—I said passionately—“If it
does, it will be entirely against my will!”
Lucio looked at me curiously.
“Nothing can happen to you against your will”—he replied; “I
suppose you wish to imply that I am to blame for introducing you to
the club? My good fellow, you need not have gone there unless you
had chosen to do so! I did not bind and drag you there! You are
upset and unnerved,—come into my room and take a glass of wine,
—you will feel more of a man afterwards.”
We had by this time reached the hotel, and I went with him
passively. With equal passiveness I drank what he gave me, and
stood, glass in hand, watching him with a kind of morbid fascination
as he threw off his fur-lined overcoat and confronted me, his pale
handsome face strangely set and stern, and his dark eyes glittering
like cold steel.
“That last stake of Lynton’s, ... to you—” I said falteringly—“His
soul——”
“Which
115 he did not believe in, and which you do not believe in!”
returned Lucio, regarding me fixedly. “Why do you now seem to
tremble at a mere sentimental idea? If fantastic notions such as God,
the Soul, and the Devil were real facts, there would perhaps be
cause for trembling, but being only the brainsick imaginations of
superstitious mankind, there is nothing in them to awaken the
slightest anxiety or fear.”
“But you”—I began—“you say you believe in the soul?”
“I? I am brainsick!” and he laughed bitterly—“Have you not found
that out yet? Much learning hath driven me mad, my friend! Science
has led me into such deep wells of dark discovery, that it is no
wonder if my senses sometimes reel,—and I believe—at such insane
moments—in the Soul!”
I sighed heavily.
“I think I will go to bed,” I answered. “I am tired out,—and
absolutely miserable!”
“Alas, poor millionaire!” said Lucio gently,—“I am sorry, I assure you,
that the evening has ended so disastrously.”
“So am I!” I returned despondently.
“Imagine it!” he went on, dreamily regarding me—“If my beliefs,—
my crack-brained theories,—were worth anything,—which they are
not—I could claim the only positive existing part of our late
acquaintance Viscount Lynton! But,—where and how to send in my
account with him? If I were Satan now....”
I forced a faint smile.
“You would have cause to rejoice!” I said.
He moved two paces towards me, and laid his hands gently on my
shoulders.
“No, Geoffrey”—and his rich voice had a strange soft music in it
—“No, my friend! If I were Satan I should probably lament!—for
every lost soul would of necessity remind me of my own fall, my own
despair,—and
116 set another bar between myself and heaven!
Remember,—the very Devil was an Angel once!”
His eyes smiled, and yet I could have sworn there were tears in
them. I wrung his hand hard,—I felt that notwithstanding his
assumed coldness and cynicism, the fate of young Lynton had
affected him profoundly. My liking for him gained new fervour from
this impression, and I went to bed more at ease with myself and
things in general. During the few minutes I spent in undressing I
became even able to contemplate the tragedy of the evening with
less regret and greater calmness,—for it was certainly no use
worrying over the irrevocable,—and, after all, what interest had the
Viscount’s life for me? None. I began to ridicule myself for my own
weakness and disinterested emotion,—and presently, being
thoroughly fatigued, fell sound asleep. Towards morning however,
perhaps about four or five o’clock, I woke suddenly as though
touched by an invisible hand. I was shivering violently, and my body
was bathed in a cold perspiration. In the otherwise dark room there
was something strangely luminous, like a cloud of white smoke or
fire. I started up, rubbing my eyes,—and stared before me for a
moment, doubting the evidence of my own senses. For, plainly
visible and substantially distinct, at a distance of perhaps five paces
from my bed, stood three Figures, muffled in dark garments and
closely hooded. So solemnly inert they were,—so heavily did their
sable draperies fall about them that it was impossible to tell whether
they were men or women,—but what paralysed me with amazement
and terror was the strange light that played around and above them,
—the spectral, wandering, chill radiance that illumined them like the
rays of a faint wintry moon. I strove to cry out,—but my tongue
refused to obey me—and my voice was strangled in my throat. The
Three remained absolutely motionless,—and again I rubbed my
eyes, wondering if this were a dream or some hideous optical
delusion. Trembling in every limb, I stretched my hand towards the
bell intending to ring violently for assistance,—when—a Voice, low
and
117 thrilling with intense anguish, caused me to shrink back
appalled, and my arm fell nerveless at my side. “Misery!”
The word struck the air with a harsh reproachful clang, and I nearly
swooned with the horror of it. For now one of the Figures moved,
and a face gleamed out from beneath its hooded wrappings—a face
white as whitest marble and fixed into such an expression of
dreadful despair as froze my blood. Then came a deep sigh that was
more like a death-groan, and again the word, “Misery!” shuddered
upon the silence.
Mad with fear, and scarcely knowing what I did, I sprang from the
bed, and began desperately to advance upon these fantastic
masqueraders, determined to seize them and demand the meaning
of this practical and untimely jest,—when suddenly all Three lifted
their heads and turned their faces on me,—such faces!—
indescribably awful in their pallid agony,—and a whisper more
ghastly than a shriek, penetrated the very fibres of my
consciousness—“Misery!”
With a furious bound I flung myself upon them,—my hands struck
empty space. Yet there—distinct as ever—they stood, glowering
down upon me, while my clenched fists beat impotently through and
beyond their seemingly corporeal shapes! And then—all at once—I
became aware of their eyes,—eyes that watched me pitilessly,
stedfastly, and disdainfully,—eyes that like witch-fires, seemed to
slowly burn terrific meanings into my very flesh and spirit. Convulsed
and almost frantic with the strain on my nerves, I abandoned myself
to despair,—this awful sight meant death I thought,—my last hour
had surely come! Then—I saw the lips of one of those dreadful faces
move ... some superhuman instinct in me leaped to life, ... in some
strange way I thought I knew, or guessed the horror of what that
next utterance would be, ... and with all my remaining force I cried
out—
“No! No! Not that eternal Doom! ... Not yet!”
Fighting
118 the vacant air, I strove to beat back those intangible weird
Shapes that loomed above me, withering up my soul with the fixed
stare of their angry eyes, and with a choking call for help, I fell, as it
were, into a pit of darkness, where I lay mercifully unconscious.
119
XI

H OW the ensuing hours between this horrible episode and full

morning elapsed I do not know. I was dead to all impressions. I
woke at last, or rather recovered my senses to see the sunlight
pouring pleasantly through the half-drawn curtains at my window,
and to find myself in bed in as restful a position as though I had
never left it. Was it then merely a vision I had seen?—a ghastly sort
of nightmare? If so, it was surely the most abhorrent illusion ever
evolved from dreamland! It could not be a question of health, for I
had never felt better in my life. I lay for some time quiescent,
thinking over the matter, with my eyes fixed on that part of the room
where those Three Shapes had seemingly stood; but I had lately got
into such a habit of cool self-analysis, that by the time my valet
brought my early cup of coffee, I had decided that the whole thing
was a dreadful fantasy, born of my own imagination, which had no
doubt been unduly excited by the affair of Viscount Lynton’s suicide.
I soon learned that there was no room left for doubt as to that
unhappy young nobleman’s actual death. A brief account of it was in
the morning papers, though as the tragedy had occurred so late at
night there were no details. A vague hint of ‘money difficulties’ was
thrown out in one journal,—but beyond that, and the statement that
the body had been conveyed to the mortuary there to await an
inquest, there was nothing said, either personal or particular. I found
Lucio
120 in the smoking-room, and it was he who first silently pointed
out to me the short paragraph headed ‘Suicide of a Viscount.’
“I told you he was a good shot!” he commented.
I nodded. Somehow I had ceased to feel much interest in the
subject. My emotion of the previous evening had apparently
exhausted all my stock of sympathy and left me coldly indifferent.
Absorbed in myself and my own concerns, I sat down to talk and
was not long before I had given a full and circumstantial account of
the spectral illusion which had so unpleasantly troubled me during
the night. Lucio listened, smiling oddly.
“That old Tokay was evidently too strong for you!” he said, when I
had concluded my story.
“Did you give me old Tokay?” I responded laughing—“Then the
mystery is explained! I was already overwrought, and needed no
stimulant. But what tricks the imagination plays us to be sure! You
have no idea of the distinct manner in which those three phantoms
asserted themselves! The impression was extraordinarily vivid.”
“No doubt!” And his dark eyes studied me curiously. “Impressions
often are very vivid. See what a marvellously real impression this
world makes upon us, for example!”
“Ah! But then the world is real!” I answered.
“Is it? You accept it as such, I daresay, and things are as they
appear to each separate individual. No two human beings think
alike; hence there may be conflicting opinions as to the reality or
non-reality of this present world. But we will not take unnecessary
plunges into the infinite question of what is, as contrasted with what
appears to be. I have some letters here for your consideration. You
have lately spoken of buying a country estate—what say you to
Willowsmere Court in Warwickshire? I have had my eye on that
place for you,—it seems to me just the very thing. It is a magnificent
old pile; part of it dates from Elizabeth’s time. It is in excellent
repair; the grounds are most picturesque, the classic river Avon
winds with rather a broad sweep through the park,—and the whole
thing,
121 with a great part of the furniture included, is to be sold for a
mere song;—fifty thousand pounds cash. I think you had better go
in for it; it would just suit your literary and poetic tastes.”
Was it my fancy, or had his musical voice the faintest touch of a
sneer as he uttered the last words? I would not allow myself to think
this possible, and answered quickly,—
“Anything you recommend must be worth looking at, and I’ll
certainly go and see it. The description sounds well, and
Shakespeare’s country always appeals to me. But wouldn’t you like
to secure it for yourself?”
He laughed.
“Not I! I live nowhere for long. I am of a roving disposition, and am
never happy tied down to one corner of the earth. But I suggest
Willowsmere to you for two reasons,—first that it is charming and
perfectly appointed; secondly, that it will impress Lord Elton
considerably if he knows you are going to buy it.”
“How so?”
“Why, because it used to be his property”—returned Lucio quietly
—“till he got into the hands of the Jews. He gave them Willowsmere
as security for loans, and latterly they have stepped in as owners.
They’ve sold most of the pictures, china, bric-a-brac and other
valuables. By the way, have you noticed how the legended God still
appears to protect the house of Israel? Particularly the ‘base usurer’
who is allowed to get the unhappy Christian into his clutches nine
times out of ten? And no remedy drops from heaven! The Jew
always triumphs. Rather inconsistent isn’t it, on the part of an
equitable Deity!” His eyes flashed strange scorn. Anon he resumed
—“As a result of Lord Elton’s unfortunate speculations, and the Jews’
admirable shrewdness, Willowsmere, as I tell you is in the market,
and fifty thousand pounds will make you the envied owner of a place
worth a hundred thousand.”
“We dine at the Eltons’ to-night, do we not?” I asked musingly.
“We
122 do. You cannot have forgotten that engagement and Lady Sibyl

so soon surely!” he answered laughing.

“No, I have not forgotten”—I said at last, after a little silence. “And I
will buy this Willowsmere. I will telegraph instructions to my lawyers
at once. Will you give me the name and address of the agents?”
“With pleasure, my dear boy!” And Lucio handed me a letter
containing the particulars concerning the sale of the estate and
other items. “But are you not making up your mind rather suddenly?
Hadn’t you better inspect the property first? There may be things
you object to——”
“If it were a rat-infested barrack,” I said resolutely—“I would still buy
it! I shall settle the matter at once. I wish to let Lord Elton know this
very night that I am the future owner of Willowsmere!”
“Good!”—and my companion thrust his arm through mine as we left
the smoking-room together—“I like your swiftness of action
Geoffrey. It is admirable! I always respect determination. Even if a
man makes up his mind to go to hell, I honour him for keeping to his
word, and going there straight as a die!”
I laughed, and we parted in high good-humour,—he to fulfil a club
engagement, I to telegraph precise instructions to my legal friends
Messrs Bentham and Ellis, for the immediate purchase in my name
at all costs, risks or inconveniences, of the estate known as
Willowsmere Court in the county of Warwick.
That evening I dressed with more than common care, giving my
man Morris almost as much trouble as if I had been a fidgetty
woman. He waited upon me however with exemplary patience, and
only when I was quite ready did he venture to utter what had
evidently been on his mind for some time.
“Excuse me sir,”—he then observed—“but I daresay you’ve noticed
that there’s something unpleasant-like about the prince’s valet,
Amiel?”
“Well,
123 he’s rather a down-looking fellow if that’s what you mean,”—I
replied—“But I suppose there’s no harm in him.”
“I don’t know about that sir,”—answered Morris severely; “He does a
great many strange things I do assure you. Downstairs with the
servants he goes on something surprising. Sings and acts and
dances too, as if he were a whole music-hall.”
“Really!” I exclaimed in surprise—“I should never have thought it.”
“Nor should I sir, but it’s a fact.”
“He must be rather an amusing fellow then,”—I continued,
wondering that my man should take the accomplishments of Amiel
in such an injured manner.
“Oh, I don’t say anything against his amusingness,”—and Morris
rubbed his nose with a doubtful air—“It’s all very well for him to cut
capers and make himself agreeable if he likes,—but it’s the deceit of
him that surprises me sir. You’d think to look at him that he was a
decent sort of dull chap with no ideas beyond his duty, but really sir,
it’s quite the contrary, if you’ll believe me. The language he uses
when he’s up to his games downstairs is something frightful! And he
actually swears he learnt it from the gentlemen of the turf, sir! Last
night he was play acting, and taking off all the fashionable folks,—
then he took to hypnotising—and upon my word it made my blood
run cold.”
“Why, what did he do?” I asked with some curiosity.
“Well, sir, he took one of the scullery-maids and sat her in a chair
and just pointed at her. Pointed at her and grinned, for all the world
like a devil out of a pantomime. And though she is generally a
respectable sober young woman, if she didn’t get up with a screech
and commence dancing round and round like a lunatic, while he kept
on pointing. And presently she got to jumping and lifting her skirts
that high that it was positively scandalous! Some of us tried to stop
her and couldn’t; she was like mad, till all at once number twenty-
two
124 bell rang—that’s the prince’s room,—and he just caught hold of

her, set her down in her chair again and clapped his hands. She
came to directly, and didn’t know a bit what she’d been doing. Then
twenty-two bell rang again, and the fellow rolled up his eyes like a
clergyman and said, ‘Let us pray!’ and off he went.”
I laughed.
“He seems to have a share of humour at anyrate,”—I said; “I should
not have thought it of him. But do you think these antics of his are
mischievous?”
“Well that scullery girl is very ill to-day,”—replied Morris; “I expect
she’ll have to leave. She has what she calls the ‘jumps’ and none of
us dare tell her how she got them. No sir, believe me or not as you
like, there’s something very queer about that Amiel. And another
thing I want to know is this—what does he do with the other
servants?”
“What does he do with the other servants?” I repeated bewilderedly
—“What on earth do you mean?”
“Well sir, the prince has a chef of his own hasn’t he?” said Morris
enumerating on his fingers—“And two personal attendants besides
Amiel,—quiet fellows enough who help in the waiting. Then he has a
coachman and groom. That makes six servants altogether. Now none
of these except Amiel are ever seen in the hotel kitchens. The chef
sends all the meals in from somewhere, in a heated receptacle—and
the two other fellows are never seen except when waiting at table,
and they don’t live in their own rooms all day, though they may
sleep there,—and nobody knows where the carriage and horses are
put up, or where the coachman and groom lodge. Certain it is that
both they and the chef board out. It seems to me very mysterious.”
I began to feel quite unreasonably irritated.
“Look here, Morris,” I said—“There’s nothing more useless or more
harmful than the habit of inquiring into other people’s affairs. The
prince has a right to live as he likes, and do as he pleases with his
servants—I am sure he pays royally for his privileges. And whether
his
125 cook lives in or out, up in the skies or down in a cellar is no

matter of mine. He has been a great traveller and no doubt has his
peculiarities; and probably his notions concerning food are very
particular and fastidious. But I don’t want to know anything about
his ménage. If you dislike Amiel, it’s easy to avoid him, but for
goodness sake don’t go making mysteries where none exist.”
Morris looked up, then down, and folded one of my coats with
special care. I saw I had effectually checked his flow of confidence.
“Very well, sir,”—he observed, and said no more.
I was rather diverted than otherwise at my servant’s solemn account
of Amiel’s peculiarities as exhibited among his own class,—and when
we were driving to Lord Elton’s that evening I told something of the
story to Lucio. He laughed.
“Amiel’s spirits are often too much for him,”—he said—“He is a
perfect imp of mischief and cannot always control himself.”
“Why, what a wrong estimate I have formed of him!” I said—“I
thought he had a peculiarly grave and somewhat sullen disposition.”
“You know the trite saying—appearances are deceptive?” went on
my companion lightly—“It’s extremely true. The professed humourist
is nearly always a disagreeable and heavy man personally. As for
Amiel, he is like me in the respect of not being at all what he seems.
His only fault is a tendency to break the bounds of discipline, but
otherwise he serves me well, and I do not inquire further. Is Morris
disgusted or alarmed?”
“Neither I think,” I responded laughing—“He merely presents himself
to me as an example of outraged respectability.”
“Ah then, you may be sure that when the scullery-maid was dancing,
he observed her steps with the closest nicety;” said Lucio—“Very
respectable men are always particular of inspection into these
matters! Soothe his ruffled feelings, my dear Geoffrey, and tell him
that
126 Amiel is the very soul of virtue! I have had him in my service for

a long time, and can urge nothing against his character as a man.
He does not pretend to be an angel. His tricks of speech and
behaviour are the result of a too constant repression of his natural
hilarity, but he is really an excellent fellow. He dabbled in hypnotic
science when he was with me in India; I have often warned him of
the danger there is in practising this force on the uninitiated. But—a
scullery-maid!—heavens!—there are so many scullery-maids! One
more or less with the ‘jumps’ will not matter. This is Lord Elton’s.”
The carriage stopped before a handsome house situated a little back
from Park Lane. We were admitted by a man-servant gorgeous in
red plush, white silk hose and powdered wig, who passed us on
majestically to his twin-brother in height and appearance, though
perhaps a trifle more disdainful in bearing, and he in his turn
ushered us upstairs with the air of one who should say “See to what
ignominious degradation a cruel fate reduces so great a man!” In the
drawing-room we found Lord Elton, standing on the hearth-rug with
his back to the fire, and directly opposite him in a low arm chair,
reclined an elegantly attired young lady with very small feet. I
mention the feet, because as I entered they were the most
prominent part of her person, being well stretched out from beneath
the would-be concealment of sundry flounced petticoats towards the
warmth of the fire which the Earl rather inconsiderately screened
from view. There was another lady in the room sitting bolt upright
with hands neatly folded on her lap, and to her we were first of all
introduced when Lord Elton’s own effusive greetings were over.
“Charlotte, allow me,—my friends, Prince Lucio Rimânez—Mr
Geoffrey Tempest; gentlemen, my sister-in-law, Miss Charlotte
Fitzroy.”
We bowed; the lady gave us a dignified bend of the head. She was
an imposing looking spinster, with a curious expression on her
features which was difficult to construe. It was pious and prim, but it
also
127 suggested the idea that she must have seen something
excessively improper once in her life and had never been able to
forget it. The pursed-up mouth, the round pale-coloured eyes and
the chronic air of insulted virtue which seemed to pervade her from
head to foot all helped to deepen this impression. One could not
look at Miss Charlotte long without beginning to wonder irreverently
what it was that had in her long past youth so outraged the cleanly
proprieties of her nature as to leave such indelible traces on her
countenance. But I have since seen many English women look so,
especially among the particularly ‘high bred,’ old and plain-featured
of the “upper ten.” Very different was the saucy and bright
physiognomy of the younger lady to whom we were next presented,
and who, raising herself languidly from her reclining position, smiled
at us with encouraging familiarity as we made our salutations.
“Miss Diana Chesney,”—said the Earl glibly—“You perhaps know her
father, prince,—you must have heard of him at any rate—the famous
Nicodemus Chesney, one of the great railway-kings.”
“Of course I know him”—responded Lucio warmly—“Who does not! I
have met him often. A charming man, gifted with most remarkable
humour and vitality—I remember him perfectly. We saw a good deal
of each other in Washington.”
“Did you though?” said Miss Chesney with a somewhat indifferent
interest,—“He’s a queer sort of man to my thinking; rather a cross
between the ticket-collector and custom-house officer combined, you
know! I never see him but what I feel I must start on a journey
directly—railways seem to be written all over him. I tell him so. I say
‘Pa, if you didn’t carry railway-tracks in your face you’d be better
looking.’ And you found him humorous, did you?”
Laughing at the novel and free way in which this young person
criticised her parent, Lucio protested that he did.
“Well I don’t,”—confessed Miss Chesney—“But that may be because
I’ve heard all his stories over and over again, and I’ve read most of
them in books besides,—so they’re not much account to me. He tells
some
128 of them to the Prince of Wales whenever he can get a chance,
—but he don’t try them off on me any more. He’s a real clever man
too; he’s made his pile quicker than most. And you’re quite right
about his vitality,—my!—his laugh takes you into the middle of next
week!”
Her bright eyes flashed merrily as she took a comprehensive survey
of our amused faces.
“Think I’m irreverent, don’t you?” she went on—“But you know Pa’s
not a ‘stage parent’ all dressed out in lovely white hair and
benedictions,—he’s just an accommodating railway-track, and he
wouldn’t like to be reverenced. Do sit down, won’t you?”—then
turning her pretty head coquettishly towards her host—“Make them
sit down, Lord Elton,—I hate to see men standing. The superior sex,
you know! Besides you’re so tall,” she added, glancing with
unconcealed admiration at Lucio’s handsome face and figure, “that
it’s like peering up an apple-tree at the moon to look at you!”
Lucio laughed heartily, and seated himself near her—I followed his
example; the old Earl still kept his position, legs a-straddle, on the
hearth-rug, and beamed benevolence upon us all. Certainly Diana
Chesney was a captivating creature; one of those surface-clever
American women who distinctly divert men’s minds without in the
least rousing their passions.
“So you’re the famous Mr Tempest?” she said, surveying me critically
—“Why, it’s simply splendid for you isn’t it? I always say it’s no use
having a heap of money unless you’re young,—if you’re old, you only
want it to fill your doctor’s pockets while he tries to mend your poor
tuckered-out constitution. I once knew an old lady who was left a
legacy of a hundred thousand pounds when she was ninety-five.
Poor old dear, she cried over it. She just had sense enough to
understand what a good time she couldn’t have. She lived in bed,
and her only luxury was a halfpenny bun dipped in milk for her tea.
It was all she cared for.”
“A hundred thousand pounds would go a long way in buns!” I said
smiling.
“Wouldn’t
129 it just!” and the fair Diana laughed—“But I guess you’ll
want something a little more substantial for your cash Mr Tempest! A
fortune in the prime of life is worth having. I suppose you’re one of
the richest men about just now, aren’t you?”
She put the question in a perfectly naïve frank manner and seemed
to be unconscious of any undue inquisitiveness in it.
“I may be one of the richest,”—I replied, and as I spoke the thought
flashed suddenly across me how recently I had been one of the
poorest!—“But my friend here, the prince, is far richer than I.”
“Is that so!” and she stared straight at Lucio, who met her gaze with
an indulgent, half satirical smile—“Well now! I guess Pa’s no better
than a sort of pauper after all! Why, you must have the world at
your feet!”
“Pretty much so,”—replied Lucio composedly—“But then, my dear
Miss Chesney, the world is so very easily brought to one’s feet.
Surely you know that?”
And he emphasized the words by an expressive look of his fine eyes.
“I guess you mean compliments,”—she replied unconcernedly—“I
don’t like them as a rule, but I’ll forgive you this once!”
“Do!” said Lucio, with one of his dazzling smiles that caused her to
stop for a moment in her voluble chatter and observe him with
mingled fascination and wonderment.
“And you too are young, like Mr Tempest,”—she resumed presently.
“Pardon me!” interrupted Lucio—“I am many years older.”
“Really!” exclaimed Lord Elton at this juncture—“You don’t look it,
does he Charlotte?”
Miss Fitzroy thus appealed to, raised her elegant tortoise-shell-
framed glasses to her eyes and peered critically at us both.
“I should imagine the prince to be slightly the senior of Mr
Tempest”—she remarked in precise high-bred accents—“But only
Welcome to our website – the ideal destination for book lovers and
knowledge seekers. With a mission to inspire endlessly, we offer a
vast collection of books, ranging from classic literary works to
specialized publications, self-development books, and children's
literature. Each book is a new journey of discovery, expanding
knowledge and enriching the soul of the reade

Our website is not just a platform for buying books, but a bridge
connecting readers to the timeless values of culture and wisdom. With
an elegant, user-friendly interface and an intelligent search system,
we are committed to providing a quick and convenient shopping
experience. Additionally, our special promotions and home delivery
services ensure that you save time and fully enjoy the joy of reading.

Let us accompany you on the journey of exploring knowledge and

personal growth!

textbookfull.com