Collective Agency and Cooperation in Natural and Artificial Systems: Explanation, Implementation and Simulation

314 Pages·2015·4.02 MB·English

by  Catrin Misselhorn (eds.)

Philosophical Studies Series
Series Editor: Luciano Floridi
Catrin Misselhorn E ditor
Collective Agency
and Cooperation
in Natural and
Artifi cial Systems
Explanation, Implementation and
Simulation
Philosophical Studies Series
Volume 122
Editor-in-Chief
Luciano Floridi , University of Oxford , Oxford Internet Institute , United Kingdom
Executive Editorial Board
Patrick Allo, Vrije Universiteit Brussel, Belgium
Massimo Durante, Università degli Studi di Torino, Italy
Phyllis Illari, University College London, United Kingdom
Shannon Vallor, Santa Clara University
Board of Consulting Editors
Lynne Rudder Baker, University of Massachusetts at Amherst
Stewart Cohen, Arizona State University, Tempe
Radu Bogdan, Tulane University
Marian David, University of Notre Dame
John M. Fischer, University of California at Riverside
Keith Lehrer, University of Arizona, Tucson
Denise Meyerson, Macquarie University
François Recanati, Institut Jean-Nicod, EHESS, Paris
Mark Sainsbury, University of Texas at Austin
Barry Smith, State University of New York at Buffalo
Nicholas D. Smith, Lewis & Clark College
Linda Zagzebski, University of Oklahoma
More information about this series at h ttp://www.springer.com/series/6459
Catrin Misselhorn
Editor
Collective Agency and
Cooperation in Natural
and Artifi cial Systems
Explanation, Implementation and Simulation
Editor
Catrin Misselhorn
Institute of Philosophy
University of Stuttgart
Stuttgart , Germany
Philosophical Studies Series
ISBN 978-3-319-15514-2 ISBN 978-3-319-15515-9 (eBook)
DOI 10.1007/978-3-319-15515-9
Library of Congress Control Number: 2015948369
Springer Cham Heidelberg New York Dordrecht London
© Springer International Publishing Switzerland 2015
T his work is subject to copyright. All rights are reserved by the Publisher, whether the whole or part of
the material is concerned, specifi cally the rights of translation, reprinting, reuse of illustrations, recitation,
broadcasting, reproduction on microfi lms or in any other physical way, and transmission or information
storage and retrieval, electronic adaptation, computer software, or by similar or dissimilar methodology
now known or hereafter developed.
T he use of general descriptive names, registered names, trademarks, service marks, etc. in this publication
does not imply, even in the absence of a specifi c statement, that such names are exempt from the relevant
protective laws and regulations and therefore free for general use.
T he publisher, the authors and the editors are safe to assume that the advice and information in this book
are believed to be true and accurate at the date of publication. Neither the publisher nor the authors or the
editors give a warranty, express or implied, with respect to the material contained herein or for any errors
or omissions that may have been made.
Printed on acid-free paper
Springer International Publishing AG Switzerland is part of Springer Science+Business Media (www.
springer.com)
Pref ace
H uman-machine interaction (HMI) and human-computer interaction (HCI) became
popular in the 1980s parallel to the rise of personal computers. At that point, the
approaches that were used for simple technical tools seemed no longer appropriate
for complex technical devices like computers. In contrast to simple tools, these
complex technical devices confronted their users with a huge amount of information
and a lot more affordances for interacting with them. Moreover, it became evident
that poorly designed human-machine interfaces can lead to serious trouble. A well-
known example in the fi eld is the Three Mile Island accident where, among other
factors, failures in human-machine interaction led to an accident in a nuclear power
plant. This suggested that a more interactive paradigm of human-machine interac-
tion was required. Properties like usability, ergonomy, and transparency were
approached from a multidisciplinary point of view involving computer sciences,
psychology, and other fi elds of study.
Since then, even more complex systems involving humans and machines have
arisen. These not only comprise a one-on-one interaction between humans and arti-
fi cial agents as is the case with computers or robots but also—as the term “artifi cial
systems” indicates—complex dynamical structures such as industrial processes,
fi nancial transactions, or transportation networks. These systems create new chal-
lenges, because they are more autonomous, intelligent, and dynamical, i.e., human
beings are not in full control of them, and they are the subject of uncertainty to a
much higher degree. To address these issues, a new research paradigm proved nec-
essary: human-machine cooperation (HMC). The premise that was unique to this
new paradigm was that humans have to c ooperate with machines in order to cope
with these complex and dynamical situations.
Although the new paradigm has led to progress in this research area, it has a
blind spot. It did not suffi ciently take into account the fact that collective agency
involving human and artifi cial agents differs in some important respects from indi-
vidual agency. The behavior of these systems cannot be reduced to the behavior and
intentions of the individuals that constitute them. This is a central insight of the
philosophical debate about the collective agency of humans which has been the
subject of intense study over the last few decades. Philosophical analyses range
v
vi Preface
from simple forms of “uniform behavior” or “emergent coordination” to complex
forms of collective agency which involve shared intentional attitudes like beliefs,
intentions, or emotions. However, a shortcoming of these analyses is that they usu-
ally do not take into account hybrid systems involving humans and artifi cial agents.
The aim of this volume is to bring these two separate strands of research together
in order to understand the nature of collective agency in natural, artifi cial, and
hybrid systems better and to improve the design and performance of hybrid systems
involving human and artifi cial agents. Problems that arise in the construction of
these systems are addressed from a theoretical and practical as well as from an
ethical point of view.
The book consists of four sections. The fi rst section is dedicated to the different
conceptions of agency that can be found in the discussion. It is concerned with the
question under which conditions an object can be considered as an agent and at
which point cooperation and collective agency arise. The fi rst chapter serves as an
introduction to the topic of the volume. It provides a conceptual framework of dif-
ferent forms of individual as well as collective agency and cooperation in natural
and artifi cial systems in order to help the reader to better understand and situate the
goals and arguments of the other contributions. In the second chapter, Fiebich,
Schwarzkopf, and Nguyen analyze the notion of cooperation at the intersection of
philosophy, psychology, and informatics with the aim of describing different forms
of cooperation and collective agency involving humans and robots in a two-
dimensional framework. Sabine Thürmel provides a gradual and multidimensional
concept of agency that applies to the interaction of human with non-human agents.
Her approach combines elements from Bruno Latour’s Actor-Network Theory
(ANT), the socio-technical approach to distributed agency of Rammert and
colleagues, and Luciano Floridi’s view of artifi cial agents.
T he second section of the book turns to human-machine cooperation. Ipke
Wachsmuth argues for the view that embodied cooperative systems should not just
be seen as tools but have to be treated as partners. He then shows how different
levels of intentionality (including emotions) – which are necessary for endowing
technical systems with collaborative functionality – have been implemented in the
humanoid artifi cial agent Max by his Bielefeld research group.
The example of Max might lead to the assumption that artifi cial agents should be
as human-like as possible in order to make human-machine cooperation natural and
intuitive. Yet, as the so-called hypothesis of the Uncanny Valley suggests, this is not
the case. Very human-like artifi cial agents actually produce negative feelings of
eeriness in humans. Valentin Schwind tracks in his contribution the cognitive and
subconscious mechanisms which are responsible for this phenomenon. He argues
that cultural, social, and aesthetic aspects of the phenomenon have so far not
received the attention that they deserve.
Hauke Behrendt, Markus Funk, and Oliver Korn provide examples of human-
machine cooperation in practice—assistive technologies at workplaces—and dis-
cuss their ethical implications. How do they change the individual workplaces and
the working environment? May handicapped people profi t from these technologies
in terms of better inclusion in the world of employment? How do assistive
Preface vii
t echnologies have to be evaluated regarding justice and the meaning of work for the
individual good life? Christian Neuhäuser takes up another ethical issue that is
becoming more pressing the more autonomous and cognitively powerful artifi cial
systems get. He addresses the question of whether robots can be morally responsi-
ble for their doings. His answer to this question is in the negative, but he neverthe-
less suggests that artifi cial systems can form part of responsibility networks together
with humans.
The focus of the third section is the transition from individual to collective
agency. Most of the contributions in this part challenge some of the assumptions
that more standard views of collective agency have made so far. One of these
assumptions is that collective agency involves collective intentionality or the so-
called we-intentions. Olle Blomberg, in contrast, describes forms of joint action that
do not require genuine collective intentionality or we-intentions, but only agent-
neutral goals.
Another infl uential view in the collective agency debate is that collective inten-
tionality involves the interconnected planning of several agents. Stephen Butterfi ll
questions this view in his contribution and tries to show that parallel planning can,
in certain cases, already be suffi cient for genuine collective agency. In contrast to
interconnected planning, in parallel planning each agent individually forms a plan
of his or her own and the other’s actions.
Tom Poljansek argues in his chapter that small-scale models (like going for a
walk together) which are mostly used as the paradigm of collective agency are mis-
leading and that one has to include cases of large-scale collective agency in order to
understand the phenomenon properly. Among the aspects that have an impact on
large-scale collective agency, the “feeling of belonging to a group” fi gures promi-
nently for him.
A nna Strasser then turns to the question of which conditions are necessary for
artifi cial systems to be part of collective action. She holds that the standard philo-
sophical views of collective agency are cognitively much too demanding and that
there are simpler forms of collective agency involving humans and artifi cial systems.
As Strasser argues by drawing parallels to non-human animals and small children,
these forms of collective agency amount to more than mere tool use.
Finally, Mog Stapleton and Tom Froese challenge the theoretical framework that
underlies many theories of collective agency with the help of the enactivist para-
digm in cognitive science. According to them, the fault of the received views is that
they abstract away from the biological body as the foundation of cognition and
agency. Stapleton and Froese develop a gradual concept of collective agency for
biological agents, but remain skeptical when it comes to the possibility of collective
agency of computational agents.
The fourth and fi nal section concerns social simulations and their relevance for
collective agency. The aim of these simulations is to explain and predict the behav-
ior of collectively acting systems. Moreover, they may also help to improve our
understanding of collective agency in general. Meike Tilebein and Maximilian
Happach show how simulations are used to model social interaction processes in
management science. Eckhart Arnold uses the example of Axelrod’s famous
viii Preface
simulations of the evolution of cooperation to argue from a philosophical point of
view, that one should not trust social simulations naively. Finally, Joanna Bryson
discusses social simulations in the broader framework of the cultural evolution of
intelligence, particularly the emergence of altruistic behavior. As these examples
show, the project of this book is also highly relevant for research and practice in the
area of social simulations.
The idea for this book goes back to the conference “Collective Agency and
Cooperation in Natural and Artifi cial Systems” which we organized in July 2013 at
the University of Stuttgart. The talks given at the conference provide the basis for
the current collection. I would like to thank the University of Stuttgart and the
Cluster of Excellence Simulation Technology (SimTech) for the generous support
and funding of this conference. I am, moreover, grateful for the extensive help in
organizing the conference and preparing the manuscript of this book provided by
Hauke Behrendt, Anja Berninger, Jörg Fingerhut, Wulf Loh, Martin Maga, Christoph
Michel, Tom Poljansek, Mog Stapleton, and Tobias Störzinger. Last but not least I
would like to thank the contributors to this volume who have each written important
and innovative contributions to this evolving fi eld of research.
Stuttgart, Germany Catrin Misselhorn
Contents
Part I Concepts of Agency
1 Collective Agency and Cooperation in Natural
and Artificial Systems ............................................................................. 3
Catrin Misselhorn
2 Cooperation with Robots? A Two-Dimensional Approach ................. 25
Anika Fiebich , Nhung Nguyen , and Sarah Schwarzkopf
3 The Participatory Turn: A Multidimensional Gradual
Agency Concept for Human and Non-human Actors .......................... 45
Sabine Thürmel
Part II Human-Machine Cooperation
4 Embodied Cooperative Systems: From Tool to Partnership .............. 63
Ipke Wachsmuth
5 Historical, Cultural, and Aesthetic Aspects
of the Uncanny Valley ............................................................................. 81
Valentin Schwind
6 Ethical Implications Regarding Assistive Technology
at Workplaces .......................................................................................... 109
Hauke Behrendt , Markus Funk , and Oliver Korn
7 Some Sceptical Remarks Regarding Robot
Responsibility and a Way Forward ....................................................... 131
Christian Neuhäuser
ix