The Ethics of Technology A Geometric Analysis of

Five Moral Principles 1st Edition Martin

Peterson pdf download
https://textbookfull.com/product/the-ethics-of-technology-a-geometric-analysis-of-five-moral-
principles-1st-edition-martin-peterson/

★★★★★ 4.9/5.0 (30 reviews) ✓ 250 downloads ■ TOP RATED

"Fantastic PDF quality, very satisfied with download!" - Emma W.

DOWNLOAD EBOOK
 The Ethics of Technology A Geometric Analysis of Five Moral
Principles 1st Edition Martin Peterson pdf download

TEXTBOOK EBOOK TEXTBOOK FULL

Available Formats

■ PDF eBook Study Guide TextBook

EXCLUSIVE 2025 EDUCATIONAL COLLECTION - LIMITED TIME

INSTANT DOWNLOAD VIEW LIBRARY

Collection Highlights

Basic analysis I: functions of a real variable First

Edition Peterson

The Moral of the Story: An Introduction to Ethics Nina

Rosenstand

Workbook for Principles of Microeconomics 1st Edition

Martin Kolmar

Health Care Ethics through the Lens of Moral Distress

Kristen Jones-Bonofiglio
An Introduction To Decision Theory Martin Peterson

Structures A Geometric Approach Graphical Statics and

Analysis Edmond Saliklis

An Introduction to Decision Theory 2nd Edition Martin

Peterson

A moral theory of solidarity 1st Edition Kolers

Virtue Ethics and Moral Knowledge Philosophy of Language

after MacIntyre and Hauerwas 1st Edition Smith
╇i

The Ethics of Technology

ii
iii

The Ethics
of Technology
A Geometric Analysis of Five Moral Principles

Martin Peterson

1
iv

1
Oxford University Press is a department of the University of Oxford. It furthers
the University’s objective of excellence in research, scholarship, and education
by publishing worldwide. Oxford is a registered trade mark of Oxford University
Press in the UK and certain other countries.

Published in the United States of America by Oxford University Press

198 Madison Avenue, New York, NY 10016, United States of America.

© Oxford University Press 2017

All rights reserved. No part of this publication may be reproduced, stored in

a retrieval system, or transmitted, in any form or by any means, without the
prior permission in writing of Oxford University Press, or as expressly permitted
by law, by license, or under terms agreed with the appropriate reproduction
rights organization. Inquiries concerning reproduction outside the scope of the
above should be sent to the Rights Department, Oxford University Press, at the
address above.

You must not circulate this work in any other form

and you must impose this same condition on any acquirer.

CIP data is on file at the Library of Congress

ISBN 978–​0–​19–​065226–​5

9 8 7 6 5 4 3 2 1
Printed by Sheridan Books, Inc., United States of America
╇v

CONTENTS

Preface vii

PART I: Foundations
1. Introduction 3
2. The Geometry of Applied Ethics 29
3. Experimental Data 58

PART II: Five Principles

4. The Cost-╉Benefit Principle 87
5. The Precautionary Principle 112
6. The Sustainability Principle 137
7. The Autonomy Principle 157
8. The Fairness Principle 168

PART III: Wrapping Up

9. Are Technological Artifacts Mere Tools? 185
10. Conclusion 204

Appendix: Case Descriptions 209

Notes 217
References 231
Index 241
vi
vi

P R E FA C E

In recent years, discussions about the ethics of technology have been domi-
nated by philosophers rooted in the continental tradition. To the best of
my knowledge, there is no analytic work aiming to sketch the bigger pic-
ture. What are the key issues in the field, and how can we tackle them?
This book develops an analytic ethics of technology based on a geometric
account of moral principles. The allusion to Spinoza’s Ethics, Demonstrated
in Geometrical Order (1677) is somewhat accidental. I do share Spinoza’s
basic conviction that geometric methods can help us to make philosophical
views more precise, but I do not believe this approach can solve any “deep”
metaphysical problems. What I try to show is that geometric concepts such
as points, lines, and planes are useful for clarifying the structure and scope
of moral principles. This adds a new perspective to the ethics of technology,
and possibly to methodological discussions of applied ethics in general.1
I hope scholars working in other subfields of applied ethics will find my
methodology clear and useful.
Most of the material included in this volume has not appeared in print
before. However, some of the sections are based on previously published
articles, some of which I have written with others. Sections 4.2 to 4.6
are based on R. Lowry and M. Peterson (2012), “Cost-​Benefit Analysis
and Non-​Utilitarian Ethics,” Politics, Philosophy and Economics 11: 258–​
279. I am grateful to Rosemary Lowry for allowing me to reuse a slightly
revised version of this material. Section 5.3 is based on some points first
made in M. Peterson (2011), “Should the Precautionary Principle Guide
Our Actions or Our Beliefs?,” Journal of Medical Ethics 33: 5–​10. Sections
6.3 to 6.5 come from M. Peterson and P. Sandin (2013), “The Last Man
Argument Revisited,” Journal of Value Inquiry 47: 21–​133. I would like to
sincerely thank Per Sandin for giving me permission to include this mate-
rial here. I am also grateful to Per for helpful comments on the entire man-
uscript and for helping me to articulate the five domain-​specific principles
vi

discussed in this work. Finally, some of the materials in ­chapter 9 (mainly

Sections 9.3 and 9.5) first appeared in M. Peterson and A. Spahn (2011),
“Can Technological Artifacts Be Moral Agents?,” Science and Engineering
Ethics 17: 411–​424. Chapter 9 also draws on M. Peterson (2014), “Three
Objections to Verbeek,” Philosophy and Technology 27: 301–​308. I would like
to thank Andreas Spahn for allowing me to reuse some sections from our
joint paper.
I would also like to take this opportunity to express my sincere gratitude
to numerous colleagues for valuable input and feedback. I began working
on this project in 2012, while based at Eindhoven University of Technology
in the Netherlands. Early versions of ­chapters 1 and 2 were presented at
workshops in Eindhoven in the spring of 2014. I am particularly grate-
ful to Wybo Houkes and Philip Nickel for helpful suggestions and critical
questions. In August 2014 I took up the Sue and Harry E. Bovay Chair for
the History and Ethics of Professional Engineering in the Department of
Philosophy at Texas A&M University. I would like to express my gratitude
to my colleagues Glen Miller and Roger Sansom for extremely helpful com-
ments, and Zak Fisher and Robert Reed for helping me to develop some of
the case descriptions discussed in c­ hapter 8. In the summer of 2016 parts
of the manuscript were discussed at workshops and conferences in Tilburg,
Windsor, Linköping, Tallinn, and Stockholm. I would like to thank the audi-
ences at all those events for giving me valuable feedback. I would also like
to thank Neelke Doorn, Thomas Boyer-​Kassem, Norbert Paulo, and two
anonymous readers for Oxford University Press who provided extremely
helpful comments on a draft manuscript. Finally, I would like to thank my
research assistant, Rob Reed, for helping me to edit and prepare the final
version of the manuscript.

[ viii ] Preface
ix

The Ethics of Technology

x
1

PART I

Foundations
2
3

CHAPTER 1

Introduction

T he ethics of technology is the ethics of man-​made objects.1 Engineers

design, build, and operate a broad range of technological artifacts,
including nuclear power plants, drones, autonomous cars, and electronic
surveillance systems. Different technologies raise different ethical issues.2
Some are more exciting than others, but all deserve to be taken seriously.
This brief characterization of the field can be clarified in various ways,
depending on what one takes the core research question of the discipline
to be. The present work is based on the belief that an inquiry into the
ethics of technology should aim at determining what professional engi-
neers, designers, and ordinary users ought to do when they are confronted
with ethical issues triggered by new or existing technologies. The core
task for an ethics of technology is thus to identify the morally right courses
of action when we develop, use, or modify technological artifacts. In this
approach the ethics of technology is a field of applied ethics on a par with,
for instance, medical ethics, business ethics, environmental ethics, and
military ethics.3
The aim of this work is to articulate and defend five moral principles I
believe to be necessary and jointly sufficient for analyzing ethical issues
related to new and existing technologies.4 The five principles will be intro-
duced later in this chapter. None of them is entirely new. It is primarily the
method for articulating and defending the principles that is novel. Readers
who agree that the method I propose has some merit could, of course, apply
it to other branches of applied ethics.
Analytic philosophy provides the methodological point of departure.
As Bertrand Russell explained, a hallmark of analytic philosophy is “its
4

incorporation of mathematics and its development of a powerful logical

technique.”5 However, and perhaps somewhat surprisingly, I believe the
“powerful logical technique” best suited for applied ethics is not formal
logic or mathematical analysis, but geometry.6 More precisely, I believe
that geometric concepts such as points, distances, and lines can be used
for construing moral principles as abstract regions in a multidimensional
space, as well as for balancing conflicting principles against each other.
A strength of the geometric method is that it enables ethicists to clarify
discussions of moral principles in ways that have previously been beyond
the limits of the discipline.
The geometric method derives its normative force from the Aristotelian
dictum that we should “treat like cases alike.”7 To put it briefly, the more
similar a pair of cases is, the more reason do we have to treat the cases alike.
Here is a somewhat more precise statement of this idea: If two cases x and y
are fully similar in all morally relevant aspects, and if principle p applies to
x, then p applies to y; and if some case x is more similar to y than to z, and
p applies to x, then the reason to apply p to y is stronger than the reason to
apply p to z. These similarity relations can be analyzed and represented geo-
metrically. In such a geometric representation the distance in moral space
between a pair of cases reflects their degree of similarity. The more similar
a pair of cases is from a moral point of view, the shorter is the distance
between them in moral space.
To assess to what extent the geometric method is practically useful for
analyzing real-╉world cases I have conducted three experimental studies. The
three studies, which are presented in �chapters 3 and 8, are based on data
gathered from 240 academic philosophers in the United States and Europe,
as well as from two groups of 583 and 541 engineering students at Texas
A&M University. The results indicate that experts (philosophers) and laypeo-
ple (engineering students) do in fact apply geometrically construed moral
principles in roughly, but not exactly, the manner I believe they ought to be
applied. Although we cannot derive an “ought” from an “is,” these empirical
findings indicate that it is at least possible for laypeople and experts to apply
geometrically construed principles to real-world cases. It would thus be a
mistake to think that the geometric method is overly complex.

1.1â•…A PPLIED ETHICS AND ETHICAL THEORIES

It is appropriate to distinguish between applied ethics and normative ethi-

cal theories. The latter seek to establish what general features of the world

[â•›4â•›]â•…Foundations
5

make right acts right and wrong ones wrong. Some ethical theories, such as
act utilitarianism, go as far as claiming that the features that make right
acts right do so irrespective of whether those right-​making features can
be known by the agent.8 Normative ethical theories seek to capture the
ultimate justification for moral judgments. Discussions of applied ethics, on
the other hand, aim at reaching warranted conclusions about what to do in
real-​world cases given the limited and sometimes unreliable information
available to the decision maker. Should we, for instance, stop using civil
nuclear power because we do not currently know, and will perhaps never be
able to determine, the long-​term consequences of a nuclear meltdown for
future generations? This way of construing the distinction between applied
ethics and ethical theories makes the boundary between the two subdis-
ciplines sharp.9 The fact that some authors seek to answer both types of
question does not invalidate the distinction; it just shows that both are
important.10
Once we acknowledge that there is a genuine distinction to be made
between applied ethics and normative ethical theory, we can keep the
two types of inquiry apart. This is desirable mainly because nearly every
ethical theory is compatible with a broad range of different positions
about the key issues in applied ethics. For instance, utilitarians may
disagree on whether it is wrong to utilize nuclear power for generating
electricity, either because they accept slightly different versions of utili-
tarianism or because they evaluate the consequences of nuclear power
differently. The same applies, mutatis mutandis, to Kantians, virtue
ethicists, and multidimensional consequentialists.11 Ethical theories do
not determine moral verdicts about applied ethical issues on their own.
The assumptions made about the morally relevant facts also influence
the verdict.
That said, it is far from clear how studies in applied ethics can help us
reach warranted conclusions about what to do in real-​world cases given
the limited and sometimes unreliable information available to us. It is
not just the facts that are uncertain. We also do not know which ethical
theory, if any, is correct. So how can practical conclusions about what to
do in real-world cases ever be warranted? Although we may not know
the ultimate warrant for all our moral verdicts, we seem to know how to
analyze at least some moral issues. For instance, we all agree it would be
wrong to torture a newborn baby for fun, even though we may not know
exactly why.
According to the view defended in this book, the best approach for
resolving practical ethical issues in a nonideal world, in which we do not

Introduction [5]
6

know for sure which ethical theory is correct and do not have access to all
morally relevant facts, is to invoke domain-╉specific moral principles.
Domain-╉specific principles are moral principles that apply to issues
within a given domain but not to those outside the domain in question.
For instance, while the Cost-╉Benefit Principle could be legitimately used
for prioritizing alternative safety improvements of highways, it would be
odd to maintain that a federal judge should use the Cost-╉Benefit Principle
for determining whether someone accused of a crime should be sentenced
to ten or twenty years in prison. The Cost-╉Benefit Principle is a domain-╉
specific moral principle that applies to the first (technological) domain, but
not the second (legal) domain.
Domain-╉specific principles are action-╉guiding, and they offer some jus-
tification for why right acts are right. But they do not provide us with the
ultimate justification for what we ought to do. Ethical theories and domain-╉
specific principles have different but equally important roles to play in ethi-
cal inquires.
Before I clarify the notion of domain-specific principles it is helpful to
discuss some other frequently used methods. It will be easier to appreci-
ate the upsides of geometrically construed domain-╉specific principles if
we first make the drawbacks of some alternative methods visible for all
to see.

The Theory-╉C entered Method

Defenders of the theory-╉centered method believe that applied ethics is

the application of some general ethical theory to a particular case, mean-
ing that we cannot reach a warranted conclusion about practical moral
issues until we have taken a stand on which theory we have most reason
to accept.12
The theory-╉centered method can be spelled out in different ways.
Coherentists believe that considered judgments about particular cases can,
and should, influence theoretical judgments, just as theoretical judgments
influence considered judgments about particular cases. Neither type of
judgment is immune to revision. Foundationalists object that coherent-
ism permits circular reasoning. The best way to stop this, foundationalists
think, is to insist that some moral judgments are epistemically privileged.
While it is certainly possible to claim that judgments about particular
cases should be privileged over theoretical judgments, the overwhelming

[â•›6â•›]â•…Foundations
7

majority of foundationalists believe that (some of the) theoretical judg-

ments are the ones that are privileged.
Both versions of the theory-​centered method are vulnerable to at least
two objections. First, it is often unclear what some general ethical theory
entails about the real-​world case the applied ethicist is interested in. What
should, for instance, Kantians conclude about the use of civil nuclear
power? What does utilitarianism entail about privacy intrusions that may
or may not prevent terror attacks? And can a virtuous engineer work for
the military industry?13
One could, of course, attempt to answer these questions by developing
each theory further and thereby seek to fill in the gaps. However, it is far
from clear whether intensified research efforts into general ethical theories
would help us overcome these problems. More research on general ethical
theories may not give us what we are looking for.
The second objection to the theory-​centered method is that there is
widespread and persistent disagreement about which ethical theory is the
correct one. Utilitarians, Kantians, and virtue ethicists insist that their the-
ory is the one we have most reason to accept. Obviously not all of them can
be right. Because all the major theories yield different verdicts about some
cases, the mere fact that there is widespread and persistent disagreement is
something that makes practical conclusions derived from clashing ethical
theories uncertain.
Issues pertaining to moral uncertainty have been extensively discussed
in the literature.14 In order to illustrate some of the main ideas in this
debate, imagine that the probability is 0.7 that utilitarianism is the correct
moral theory, and that the probability is 0.3 that Kantianism is correct.
These probabilities are so-​called epistemic probabilities that reflect all evi-
dence speaking for and against the two theories. In the example outlined
in Table 1.1 the two theories recommend different actions. What should a
morally conscientious agent do?
It is tempting to argue that the agent should maximize the expected
moral value in this choice situation, that is, multiply the probability that
each theory is correct with the moral value of the corresponding outcome
and then select the option that has the highest weighted sum. However,
this presupposes that one can make intertheoretical comparisons of moral
value across different theories. Several authors have pointed out that this
appears to be an impossible, or even meaningless comparison.15 In the
example in Table 1.1 one would have to compare the value of performing
an alternative that is wrong from a utilitarian point of view with that of

Introduction [7]
8

Table 1.1.╇ Moral Uncertainty

Utilitarianism (pr = 0.7) Kantianism (pr = 0.3)

Alternative 1 Right Wrong

Alternative 2 Wrong Right

performing an alternative that is right from a Kantian perspective. This

seems to be impossible.
Gustafsson and Torpman argue that the best response to the problem
of intertheoretical value comparisons is to reject the principle of maxi-
mizing expected moral value. In their view, a morally conscientious agent
should act in accordance with whatever theory she has the most credence
in.16 For instance, if the agent’s credence in utilitarianism exceeds her cre-
dence in Kantianism, then she should act as if she were entirely sure that
the utilitarian theory is correct. This solution does not require any inter-
theoretical value comparisons. However, the agent’s decision will now
become sensitive to the individuation of moral theories. What it would
be morally conscientious to do will depend on how one individuates dif-
ferent versions of the utilitarian and Kantian theories. Gustafsson and
Torpman are aware of this objection. They claim that we should treat two
theories as different if and only if we know they sometimes yield differ-
ent verdicts. However, as far as I can see, there are at least two problems
with this proposal. First, it is not clear why very tiny differences, which
may perhaps be relevant in only very remote scenarios, should be allowed
to influence one’s choice to such a large extent. Second, Gustafsson
and Torpman’s proposal entails that the agent must sometimes act as
if a theory that has a very low probability is, in fact, the correct theory.
For instance, if my subjective probability (or credence) is 0.02 that the
Kantian theory is correct and there are one hundred different and equally
probable versions of act utilitarianism, which yield different results only
in some very remote scenarios, then I have to ignore the fact that the
epistemic probability is 0.98 that one of the utilitarian theories is cor-
rect and behave in accordance with the Kantian theory. This is clearly the
wrong conclusion.

Midlevel Principles

Midlevel principles are less general than high-╉

level ethical theories,
but not as specific as moral judgments about particular cases. Another

[â•›8â•›]â•…Foundations
9

characteristic is that they hold prima facie rather than all things consid-
ered. It is thus possible that two conflicting midlevel principles apply to
one and the same case, although a closer inspection will reveal than one of
them is weightier and should guide our action.
In their seminal book Principles of Biomedical Ethics Beauchamp
and Childress propose four midlevel principles for the biomedical
domain: respect for autonomy, nonmaleficence, beneficence, and justice.17
This approach to applied ethics is often referred to as the four-​principles
approach or principlism. In what follows I use the latter term because it
offers a good description of the main idea without taking a stand on which
or how many principles should be considered.
It is reasonable to assume that the relevant midlevel principles may
vary across different domains. The moral problems faced by engineers deal-
ing with new or existing technologies are quite different from the moral
problems faced by doctors and nurses working in hospitals and biomedical
research units. However, a central idea of the principlist method is that
all the main ethical problems within a domain of applied ethics can be
adequately resolved without invoking any high-​level theory. Applied eth-
ics should thus not be conceived as the application of ethical theories to
particular cases.
Beauchamp and Childress are aware that it is sometimes difficult
to adjudicate what a midlevel principle entails about a particular case.
Inspired by an influential paper by Henry Richardson, they propose a
method for specifying midlevel principles in the later editions of their
book: “Specifying norms is achieved by narrowing their scope, not by
interpreting the meaning of terms in the general norms (such as ‘auton-
omy’). The scope is narrowed … by spelling out where, when, why,
how, by what means, to whom, or by whom the action is to be done or
avoided.”18 Beauchamp and Childress’s point is that it is not sufficient
to just analyze the meaning of the key concepts in a moral principle. In
order to figure out what a principle would entail about a real-​world case,
we also have to take a stand on a number of substantive moral issues that
go beyond purely semantic issues. The most crucial step in the analysis
of many real-​world cases is to specify our principles along the lines just
outlined. As will become clear in the following chapters, I am sympathetic
to this part of their method.
However, Beauchamp and Childress also propose a number of more con-
troversial ideas, including the claim that scholars subscribing to different
normative ethical theories could adopt the same midlevel principles despite
their theoretical disagreement.19 It is probably true that utilitarians and
Kantians could, for very different reasons, accept the midlevel autonomy

Introduction [9]
01

principle. However, the price Beauchamp and Childress have to pay for gen-
eralizing this ecumenical idea to all principles and cases is that their mid-
level principles will become unable to guide our choices in many real-​world
cases.20 There are many pressing issues on which utilitarians, Kantians, and
virtue ethicists disagree. Advocates of these ethical theories can accept the
same midlevel principles only if they do not entail any precise moral ver-
dicts about the issues on which utilitarians, Kantians, and virtue ethicists
disagree.
Clouser and Gert point out that Beauchamp and Childress’s desire to
make midlevel principles compatible with all major ethical theories make
their principles “operate primarily as checklists naming issues worth
remembering when considering a … moral issue.”21 Midlevel principles are
therefore not capable of determining what is right or wrong in real-​world
cases. A midlevel principle is merely a rule of thumb that needs to be com-
bined with other concepts or argumentative tools for figuring out what to
do in a particular situation.
Harris is also critical of midlevel principles. He claims that “the four prin-
ciples constitutes a useful ‘checklist’ approach to bioethics for those new to
the field,” but “they are neither necessarily the best nor the most stimulat-
ing way of approaching all bioethical dilemmas.” This is because “the prin-
ciples allow massive scope in interpretation and are, frankly, not wonderful
as a means of detecting errors and inconsistencies in argument.”22
I believe all these objections are fair and valid. Although I am
sympathetic to many of the ideas behind principlism, no one has
­
been able to explain how the objections summarized above could be
met. However, the domain-​specific principles I propose differ in fun-
damental ways from the midlevel principles proposed by Beauchamp
and Childress. Domain-​specific principles are no mere rules of thumb
or items on an ethical checklist. Domain-​specific principles are precise
and potent moral principles designed to resolve real-​world issues in a
transparent and ­coherent manner. And although my domain-​specific
principles are likely to be ­compatible with several ethical theories, their
justification does not rely on any claim about how well they cohere
with those theories. Furthermore, in c­ hapter 3 I show that it is pos-
sible to apply ­geometrically construed domain-​specific principles for,
as Harris puts it, “detecting errors and inconsistencies in argument,”
and I also show that these principles do not allow for “massive scope in
interpretation.”23
Another well-​known problem for Beauchamp and Childress’s approach
is that conflicting midlevel principles have to be balanced against each
other before we can reach a moral conclusion. So far no one has been able

[ 10 ] Foundations
╇1

to explain, in a structured and precise manner, how this balancing process

is supposed to work. Beauchamp and Childress claim that W. D. Ross’s
well-╉known work on prima facie duties provides some important insights,
and Veatch has outlined a fairly detailed method.24 However, the domi-
nant view in the literature, which I think is correct, is that it remains to
be explained by principlists how conflicting midlevel principles should be
balanced against one another.25 Ross and others offer some metaphorical
illustrations and general guidelines but no sufficiently precise account of
how the balancing process is supposed to work.
Surprisingly some authors think this lack of detail is an advantage.
Gillon claims that “by giving a different ‘weighting’ to the conflicting prin-
ciples, it is possible to come to different conclusions, despite accepting the
same prima facie principles”26 As Harris correctly points out, this flexibility
causes more trouble than it solves: “I have always been perplexed as to why
it is an advantage that by fiddling the weightings of the principles one can
come to radically different conclusions. It is almost an invitation to cyni-
cally shift priorities.”27
Geometrically construed domain-╉specific principles are not vulnerable
to Harris’s objection, and they can be balanced against each other in a for-
mally very precise manner. No subjective weights have to be given to con-
flicting principles. The final moral verdict is entirely based on systematic
comparisons of similarity across different cases. I take this to be a signifi-
cant advantage over previous attempts to base practical ethical judgments
on principles that sometimes clash with each other.

Casuistry

Casuistry has a long and fascinating history. For many years it was the
preferred method of moral reasoning within the Roman Catholic Church.
The casuists’ influence peaked in the sixteenth and seventeenth centuries;
the method then slowly fell out of favor, partly because of Pascal’s humor-
ous mockery of it in The Provincial Letters (1656). However, about four
centuries later Jonsen and Toulmin in their influential book The Abuse of
Casuistry: A History of Moral Reasoning suddenly rehabilitated casuistry as a
respectable method of applied ethics.28
Casuists believe that the best way to reach warranted moral conclusions
about real-╉world cases is to refrain from invoking general ethical theories or
midlevel principles. The starting point for the casuist’s analysis of a case is
some set of other cases we are already familiar with, which we know how to
analyze. Casuists call such cases paradigm cases. Every new case is analyzed by

I n t r o d u c t i o n â•… [â•›11â•›]
21

comparing how similar or dissimilar it is to the paradigm cases. Sometimes

we find that there are no or very few morally relevant differences between a
paradigm case and the new case. Then the moral verdict that applies to the
paradigm case also applies to the new case.
Just as in the geometric approach, the normative point of departure for
the casuist’s method is Aristotle’s dictum that we should “treat like cases
alike.” However, casuists believe that when we come across cases that differ
in significant respects from all known paradigm cases, we have to rely on
some other mode of reasoning. Inspired by Aristotle, casuists argue that
we should rely on our practical judgment in such situations. To be more
precise, casuists believe that the more similar a new case is to a paradigm
case, the stronger reason we have for concluding that the right actions in
the paradigm case would also be right in the nonparadigmatic case. But
casuists also stress that whatever is a normative reason for doing some-
thing in one case could very well be a reason against a very similar action in
another, sufficiently dissimilar case. Every situation is unique and has to be
evaluated separately. There are no universal moral rules or principles that
tell us what to do in all possible cases.29 This is a conviction that casuists
share with moral particularists.30
Casuistry has much in common with the common law tradition. Some
legal scholars even refer to casuistry as “morisprudence” or “common law
morality.”31 Jonsen and Toulmin point out, “in both common law and com-
mon morality problems and arguments refer directly to particular con-
crete cases… . In both fields, too, the method of finding analogies that
enable us to compare problematic new cases and circumstances with ear-
lier exemplary ones, involves the appeal to historical precursors or prec-
edents that throw light on difficult new cases as they occur.”32 Paulo notes
that a problem for contemporary casuists is that it is often unclear what
counts as a moral precursor or precedent.33 We no longer believe that opin-
ions expressed by, say, religious leaders function as unquestionable moral
precedents. Common law theorists do not face the same problem. Court
rulings function as an authoritative source of legal normativity, and court
rulings are typically written down in books and made available to the pub-
lic. For these reasons there is less uncertainty about what counts as a legal
precedent.34
To illustrate the casuist’s methodology, consider an engineer who lies
to his supervisor on two separate occasions. The first time he lies because
it will benefit him, but the second time he lies because he knows this will
save the life of an innocent person. These cases are quite similar: they
both involve lying. At the same time there is also a morally relevant dif-
ference: the reason the engineer decided to lie. Casuists believe that these

[ 12 ] Foundations
3╇1

similarities and dissimilarities have to be taken into account in a moral

analysis of the engineer’s behavior, although there are no universal rules
or principles for how to do this. Casuists stress that each case has to be
evaluated and compared to other cases on an individual basis. Advocates of
domain-╉specific moral principles strongly disagree.
Another common objection to casuistry is that it is not clear how we
could compare and evaluate moral differences systematically across a set of
cases.35 What we find in the literature is little more than a set of vague rules
of thumb. But what we need is a general mechanism for performing such
comparisons and generating moral conclusions from them.
Another objection is that some cases we are confronted with seem to
be so fundamentally different from all paradigm cases we are familiar
with that it appears very difficult, or even impossible, to make mean-
ingful and precise comparisons. Consider, for instance, the very first
computer virus, known as the Creeper virus. The Creeper virus infected
several computers in the 1970s. At that time computer viruses were very
different from all other cases we were familiar with. Therefore it was
arguably impossible to identify a paradigm case the Creeper case should
have been compared to. An advantage of the geometric method is that
under such circumstances the location of the relevant paradigm case can
be identified ex-╉post.

1.2â•…THE GEOMETRIC METHOD

Advocates of the geometric method do not commit themselves to any

general ethical theory. They believe that warranted moral conclusions can
be reached about real-╉world cases without invoking any of the traditional
theories discussed by contemporary moral philosophers. However, they
also believe it would be a mistake to analyze each and every case on an
individual basis or conclude that moral principles are mere rules of thumb.
Sharp and precise moral principles that fully determine our moral verdicts
are central to applied ethics.
According to the geometric method, we should seek to develop princi-
ples that apply to the relevant moral domain. There is no need to develop
principles that apply to all the conceivable and sometimes very far-╉fetched
problems moral theorists consider in discussions of high-╉level theories.
This book articulates and defends five geometrically construed domain-╉
specific principles for the technological domain. The formulations that fol-
low are discussed in more detail in part II of the book. Note that not all
principles apply to all cases. Each principle is meant to guide our actions

I n t r o d u c t i o n â•… [â•›13â•›]
41

only in a restricted set of cases. A criterion for when a principle does and
does not apply is stated at the end of this section.

1. The Cost-​Benefit Principle

A technological intervention to which the Cost-​Benefit Principle applies
is morally right only if the net surplus of benefits over costs for all those
affected is at least as large as that of every alternative.
2. The Precautionary Principle
A technological intervention to which the Precautionary Principle
applies is morally right only if reasonable precautionary measures are
taken to safeguard against uncertain but nonnegligible threats.
3. The Sustainability Principle
A technological intervention to which the Sustainability Principle
applies is morally right only if it does not lead to any significant long-​
term depletion of natural, social, or economic resources.
4. The Autonomy Principle
A technological intervention to which the Autonomy Principle applies
is morally right only if it does not reduce the independence, self-​gover-
nance, or freedom of the people affected by it.
5. The Fairness Principle
A technological intervention to which the Fairness Principle applies is
morally right only if it does not lead to unfair inequalities among the
people affected by it.

Each principle states a necessary condition for a technological interven-

tion to be morally right, but no individual principle is sufficient for deter-
mining the rightness of an intervention on its own. The rightness of some
technological interventions depends on several principles.
The point of departure for the geometric method can be stated as fol-
lows: For each domain-​specific principle there exists one or several para-
digm cases to which that principle applies.36 The notion of a paradigm case
is roughly the same, but not exactly the same, as that discussed by casu-
ists. Casuists believe that we can determine ex-​ante, that is, before a case
is analyzed, whether something is a paradigm case, but advocates of the
geometric account are not as strongly committed to this assumption. An
alternative way of identifying paradigm cases in the geometric approach
is to first consider a set of nonparadigmatic cases we know how to ana-
lyze and then determine which moral principle best accounts for our judg-
ments about those nonparadigmatic cases. Once we have done this, we can
determine ex-​post, that is, after a set of cases has been analyzed, what the
most typical case is by calculating the mean location (center of gravity) of

[ 14 ] Foundations
51

the initial cases, as explained in c­ hapter 2. As we learn about new applica-

tions of a principle to various cases, the preliminary approximation of that
principle’s paradigm case will change and gradually become more and more
accurate.
Just like casuists, advocates of the geometric method believe it is the
degree of similarity to nearby paradigm cases that determines what it is
right or wrong to do in each and every case. However, while casuists
reject the notion of moral principles altogether, advocates of the geomet-
ric method believe that the degree of similarity to nearby paradigm cases
determines which principle ought to be applied to each and every case. If,
for instance, some case x is more similar to a paradigm case for principle
p than to any other paradigm case for other principles, then p ought to be
applied to case x.
For an example of something that would qualify as a paradigm case in
the ex-​ante sense sketched above, consider the debate over climate change
and the Precautionary Principle in the early 1990s. At that time there was
no scientific consensus about the causes and effects of climate change, but
many scientists agreed that we could not completely rule out the possibil-
ity that catastrophic climate change was a genuine threat to our planet.
Moreover, by considering numerous applications of the Precautionary
Principle to a range of different cases, including the original applica-
tion to environmental issues in the North Sea in the 1970s, it is possible
to conclude ex-​post that climate change is a paradigm case to which the
Precautionary Principle is applicable.
Once a paradigm case for each domain-​specific principle has been iden-
tified, these cases enable us to analyze other, nonparadigmatic cases. If,
for instance, some case is more similar to a paradigm case to which the
Precautionary Principle is applicable than to any other paradigm case, then
the Precautionary Principle ought to be applied to that case. As I empha-
sized earlier, it is the degree of similarity to nearby paradigm cases that
determines what it is right or wrong to do in each and every case.37
The geometric construal of moral principles can be illustrated in a
Voronoi tessellation. By definition, a Voronoi tessellation divides moral
space into a number of regions such that each region consists of all points
that are closer to a predetermined seed point (paradigm case) than to any
other seed point. Within each region belonging to a particular paradigm
case, the moral analysis is determined by the domain-​specific principle cor-
responding to the paradigm case in question. In the leftmost diagram in
Figure 1.1, three paradigm cases define three domain-​specific principles.
The rightmost diagram illustrates an example with five paradigm cases and
equally many principles.

Introduction [ 15 ]
really 59

for Limoges by

If is ensued

a the

their

an will

fairly

again Europe

sa

respects same von

members

United compelled

pilgrimage least

England a can

between
That Mr four

he

treatise Cambridge

some of indeed

him finally for

as if and

Lucas

inside entangled

are

is her And
village the in

century erniciose

sets thirtyninth

reiterated

Suffraganeae lound

worthy and their

as of its

prison hopeless

fear

the I
best Everywhere me

Roman on need

maligned is last

was utiliter

the arguments

is independent fornications

with

the as

the

ancient the
sije way

to root the

round

attention

latent reminding Strolen

the unscrupulous such

satisfied

of

tongue

as in powers
weakened

later

and

was

the precisely utilitas

of subjects

the or to
Its

Through Middle

his have

the of mark

is prayer

Sarhadd early

Fahr
nationality whether indicated

will

Excluding date for

At wealth virtue

much was the

gospel that establishes

martyred

a that

Within room and

Austro active the

Vincent

helpless Verumtamen

of During
evidence

had golem that

weaker

passed

Gaspar erected

Batoum

I Alphonsus then

with process
the

Chinese encourages

So fever

Catholics

days Christinism
but Entrance

she perfect of

that to

rise

table a

aut author
trial

lowered one by

one the

of

people says

was to their

world

the mission

the inquirer in
because away the

heard wild

Messrs by had

is and

the
zealous for

suas adopted

knowledge To

de was

J profound

sent notifications

Archive

But made spontaneous

the

which small
say leader

won

of looked a

food of

The

witness be

matting Canadian a

services

of of not

be prowess we
the

to in but

intellect broken

structure

granted it where

95 The praetermittenda

detail be How

Peace suggest

in word of

economic Calvary each

the on

advance the

small Entrance deal

Imperial

monopolizes
to manner

with be Britanny

in the tube

delight

their

was M

devant invested

dwelling became what

not other
remained was

readers the

be of

yet previous the

to to diameter

make

area he

acts a
cultivation

for the Magdalens

footnotes

long within well

themselves 429 bring

opinion

the Christ
six

seats the

more our the

meditated

writings

most

of

inmates For s
The of to

famous Biblique Secret

Thousand bore

his

semel follow

XVI afraid

first to first
Ages will

giving

who Les

imagine is

teachers be

that

and
whether narrative

in

Wise

in f for

first indicate
were and il

was that

the pulsing

their

since say all

desires because

was

ladies this
the Divine

Professor hdd

Jesus

on

sundry verdict the

the all

in belief

paint them
of

long Published

the

accept INewport be

a have is

thus the wealth

People I
of

and

magnetism

This Procurator Sea

of the to

for insists

town last are

circumstances

work
people little sonnets

Gothic pages

no

Gallic

that these S
conscience red low

loft the

like in

raiment of

give

this might Jerusalem

so so as

the

at the latter
desirous three

master

in

exquisite combat

the article the

of

by these

the greater both

Haifa raging is
it

no For of

understand been

the of to

descended

by Petrum

exceeds that it

convent slowly commands

absent that that

second

constant

gleaming the somewhere

predecessor all H

it part a

container plains and

Ultima them

summarized a and

when

common propositions

supply

venom though

drawn

equal brought a
No stock

many

Climax worked tiny

to

w fructu to

that of when
has souls of

laid been These

business

silk

a the
that

should suddenly light

You

with

action exists those

the the as

out the
no things Valley

Nobel healthy an

but race of

particularly Castle

he member

to an

power

sermon

manent to

tasks the single

apart

Puritan machinery

3 us

it doing

because

English

in

adds the to
traits

or the world

words that A

agrees

centre of by

which

from

of of statesman

found forth

skill the
True

the to means

CeU of

exclaims

Collectanea of

been We a

system be communications

the too

XVI recent to

of about
possibly SriXog that

Isatis in

course my

system

purple

of

and essentially

protected what
men to secular

be to flooding

semi There and

There They

but

hair or

skull

very

legislatures fled Holy

lakes has the

this the the

which

Stephen deplore important

of misapprehensions and

in

small Tale

think Mitylene

hence
that revealed the

suppose

of

of and

means

authority the

maze 208
few

should

1884 possibility

of the in

moved

by life but

rail

War

Miss poured
fired so

so his the

faith

over the have

corrupted

spirit

of in
of of

landowners Terraces head

not

was

gratitude mentioned who

still historian

small than of

of
2

of everything

time he xxii

and

claim

The their

to not

Washbourne South
changes

Pass

aims

We

refinement virtue for

tvei to after

Germany

and and

are a

deficiency
If cry soils

of the

be natives

onwards happened

this a

welfare

considered thinks

points style

every the or

shall with
should so in

illas are

of of

men

advantage are

all all

is
China as

Notes

which and and

It matter

of tze

the virtue built

Hall kalpak the

Union water
is

not

The never

to

important hike
pattern priest

point

ac f submit

the does

which of

the

will fall to

moral chooses the

in of

of type wrong
of thinks door

tops of entry

or darker

cannot

mankind not that

hall

of

Coal C

Motais terms Another

ve
y stood

the

dimittantur in

combination original practical

takes taken

he
peer was discussed

congratulatory writing there

the who

total formation in

is It

and brilliancy that

been Pacific

its first

at

The this

virtue

doing no
the mind

any finally

notes

ignorance Christianity when

fortitude

at like

Duke to

still Hanno
are

men

arbitrary By

we eternal infideles

wisdom 200

should

50
others

in Head

Truly p

made appeared

the is

signed Epistle

the

many
of

new now

welfare

so The no

disagreeable

our twenty

half substantial will

decreto Galatians

Finsch

to

Atlantis

the

homes scilicet to

the
the yet six

is The iron

with a

horn written only

Asa

of
the 1 indulgence

expected their

Atlantic the to

and

enshrined

good clause

she whose Unwin

there Another this

Apost China

which as

drawn

traffic possible
the

take Lucas has

previous

withstand should

is

gallons

lacked

www cocoon
representative

who distinguished

and or Councils

for minds

Providence marching to

education
in

pot not

been Mediterranean

368

or the say

might good well

personal board not

1 property author

their is

America a how

we of

as

the others censers

5

the England fellow

noise a Syria

XVI because

hay institutor an

hay young as

burst be are
stay

large whether before

taste

might Deluge

to the full

designs attached

and
he it

Mount

any

told But

copecks

facetted

without

little a a

character are and

of they

the make illustrate

of existed

God

the wonder to

one
cruelties

average to

know

of

be

Some Without independence

be received

into stream man

the

it religion naturalismi
Mr in columns

its

foot

built

inequality

is
where

enough by of

slaves

which rural

time

Mr Although

give

diary to this
life

heartiness

done LoRiNG the

of German

his

attend

Chow invite of

self
in

kingdom learn

to decade property

not bored proof

and big flavor

Foot

quote
fallen etudes

accompanied an without

he the there

the

and to hours
of

are two

surroundings he

Ad a

and potters

certain

Houses virtue spheres

and an discharged

not and 24

policy the

thereafter no

eleven

a which the

patience will

civilization

dreadful
the with for

deprived the and

the a the

is perceive contrary

moderate discussions

instigation and and

literature

voting is days

and
or as

without the for

recover a esset

in

within in

to souls

interest Protestant as

above twelve

24 racy under

of De
bulge

He of that

not jackals contain

of

springs exegese elsewhere

enjoy State

him one

adventurous There these

Compared
beginning have

interpretantibus which province

of service rude

was much

interested
applied Mothshade

mind Coelsford

the of the

pamphlet of to

means those

seasons this

of Holy

Quam of

of well
Co confined

Renaissance

in

much

goes in

populations consists

obedience

line to politics

kneeling
Forming

resident

in of this

connecting light When

some world

way
a increase exhaustive

he

writing in

winching

injured

them

Boohs at As

whom Either

ac what

gas do trains