╇ i

Distributed Agency
ii

FOUNDATIONS OF HUMAN INTERACTION

General Editor: N. J. Enfield, The University of Sydney

This series promotes new interdisciplinary research on the elements of human

sociality, in particular as they relate to the activity and experience of commu-
nicative interaction and human relationships. Books in this series explore the
foundations of human interaction from a wide range of perspectives, using mul-
tiple theoretical and methodological tools. A premise of the series is that a proper
understanding of human sociality is only possible if we take a truly interdisci-
plinary approach.

Series Editorial Board:
Michael Tomasello (Max Planck Institute Leipzig)
Dan Sperber (Jean Nicod Institute)
Elizabeth Couper-╉Kuhlen (University of Helsinki)
Paul Kockelman (Yale University)
Sotaro Kita (University of Warwick)
Tanya Stivers (University of California, Los Angeles)
Jack Sidnell (University of Toronto)

Recently published in the series:

Agent, Person, Subject, Self The Instruction of Imagination
Paul Kockelman Daniel Dor
Exploring the Interactional How Traditions Live and Die
Instinct Olivier Morin
Edited by Anna Dina L. Joaquin
The Origins of Fairness
and John H. Schumann
Nicolas Baumard
Relationship Thinking
Requesting Responsibility
N. J. Enfield
Jörg Zinken
Talking About Troubles in
Accountability in Social Interaction
Conversation
Jeffrey Robinson
Gail Jefferson
╇ iii

Distributed Agency

Edited by N. J. Enfield
and
Paul Kockelman

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.

Library of Congress Cataloging-in-Publication Data

Names: Enfield, N. J., 1966– editor. | Kockelman, Paul.
Title: Distributed Agency / [edited by] N. J. Enfield and Paul Kockelman.
Description: Oxford ; New York : Oxford University Press, [2017] |
Series: Foundations of human interaction | Includes bibliographical references and index.
Identifiers: LCCN 2016022303 (print) | LCCN 2016038875 (ebook) |
ISBN 9780190457211 (paperback) | ISBN 9780190457204 (cloth) |
ISBN 9780190457228 (pdf) | ISBN 9780190457235 (Online Content)
Subjects: LCSH: Semiotics—Psychological aspects. | Communication—Psychological aspects. |
Human behavior. | Cognition. | Psycholinguistics. | BISAC: LANGUAGE
ARTS & DISCIPLINES / Linguistics / General. | SOCIAL SCIENCE / Anthropology / Cultural. |
SOCIAL SCIENCE / General.
Classification: LCC P99.4.P78 D47 2017 (print) | LCC P99.4.P78 (ebook) | DDC 128/.4—dc23
LC record available at https://lccn.loc.gov/2016022303

9 8 7 6 5 4 3 2 1
Paperback printed by WebCom, Inc., Canada
Hardback printed by Bridgeport National Bindery, Inc., United States of America
╇ v

CONTENTS

Series Editor Prefaceâ•…â•… ix

Editors’ Prefaceâ•…â•… xi
List of Contributorsâ•…â•… xvii

PART ONE: Agency as Flexible and Accountable Causality

1. Elements of Agencyâ•…â•… 3
N. J. Enfield
2. Distribution of Agencyâ•…â•… 9
N. J. Enfield
3. Gnomic Agencyâ•…â•… 15
Paul Kockelman
4. Semiotic Agencyâ•…â•… 25
Paul Kockelman

PART TWO: Agency of Institutions and Infrastructure

5. Agency in State Agenciesâ•…â•… 41
Anya Bernstein
6. Upending Infrastructure in Times of Revoltâ•…â•… 49
Julia Elyachar

PART THREE: Language and Agency

7. Brain-╉to-╉Brain Interfaces and the Role of Language
in Distributing Agencyâ•…â•… 59
Mark Dingemanse
8. Requesting as a Means for Negotiating Distributed Agencyâ•…â•… 67
Simeon Floyd
9. Social Agency and Grammarâ•…â•… 79
Giovanni Rossi and Jörg Zinken
vi

10. Distributed Agency and Action under the Radar

of Accountability   87
Jack Sidnell

PART FOUR: Economy and Agency

11. Distributed Agency and Debt in the Durational Ethics
of Responsibility   99
Jane I. Guyer
12. Money as Token and Money as Record
in Distributed Accounts   109
Bill Maurer

PART FIVE: Distributing Agency within Selves and Species

13. Distribution of Agency across Body and Self   119
Ruth H. Parry
14. Distributed Agency in Ants   131
Patrizia d’Ettorre

PART SIX: Social Bonding Through Embodied Agency

15. Group Exercise and Social Bonding   141
Emma Cohen
16. Social Bonding Through Dance and ‘Musiking’   151
Bronwyn Tarr

PART SEVEN: Agency and Infancy

17. Timescales for Understanding the Agency of Infants
and Caregivers   161
Joanna Rączaszek-​Leonardi
18. Movement Synchrony, Joint Actions, and Collective Agency
in Infancy   169
Bahar Tunçgenç

PART EIGHT: The Agency of Materiality

19. The Agency of the Dead   181
Zoë Crossland
20. Distributed Agency in Play   191
Benjamin Smith
21. Contingency and the Semiotic Mediation
of Distributed Agency   199
Eitan Wilf

[ vi ] Contents
  vii

PART NINE: The Place of Agency

22. Place and Extended Agency   213
Paul C. Adams
23. How Agency Is Distributed Through Installations   221
Saadi Lahlou

PART TEN: From Cooperation to Deception and Disruption

24. Cooperation and Social Obligations   233
David P. Schweikard
25. Deception as Exploitative Social Agency   243
Radu Umbres
26. Disrupting Agents, Distributing Agency   253
Charles H. P. Zuckerman

Author Index   271

Subject Index   275

Contents  [ vii ]
  ix

S E R I E S E D I T O R P R E FA C E

The human capacity for distributed agency allows us to dissociate social

units from the borders of the individual. Our social units—​the “agents”
who operate in people’s worlds—​can morph and shift in shape and size and
constitution. From fleeting moments of cooperation to historical move-
ments of community, people extend beyond the bounds of individuals. The
case studies presented in this volume provide a wide-​ranging set of entry
points into this captivating area of study. Together they show that when
we look at distributed agency we are looking at the cognitive and cultural
dynamics of human sociality.
N. J. E.
Sydney, September 2016
╇ xi

E D I T O R S ’ P R E FA C E

We never really act alone. Our agency is enhanced when we cooperate with
others, and when we accept their help. We benefit when we take credit
for people’s achievements, or when we free-╉ride on their ideas and their
strengths. And not only is our agency shared through action in these ways,
but we often have agency without having to engage in action at all. Even
when at rest, we are bound up in networks of cause and effect, intention
and accountability. The distribution of agency, for better or worse, is every-
where, and our species has perfected its arts.
This book presents an interdisciplinary inroad into the latest thinking
about the distributed nature of agency: what it’s like, what are its condi-
tions of possibility, and what are its consequences. The book’s 26 chapters
are written by a wide range of scholars, from anthropology, biology, cogni-
tive science, linguistics, philosophy, psychology, geography, law, econom-
ics, and sociology. While each chapter takes up different materials using
different methods, they all chart relations between the key elements of
agency: intentionality, causality, flexibility, and accountability. Each chap-
ter seeks to explain how and why such relations are distributed—╉not just
across individuals, but also across bodies and minds, people and things,
spaces and times. To do this, the authors work through empirical studies
of particular cases, while also offering reviews and syntheses of key ideas
from the authors’ respective research traditions.

IDEAS AND QUESTIONS GUIDING THIS BOOK

All creatures have some form of agency, but agency in humans is argu-
ably unique. Humans have a special capacity to share agency through joint
commitment and cooperation toward common goals, as well as through
coercive and parasitic practices by which we co-╉opt the capacities, and take
credit for the consequences, of others. We are the most flexible and creative
xii

actors of all species, as shown by our formidable and complex technolo-

gies, from levers, swords, and printing presses, to bridges, drones, and the
internet. Culture is itself the exemplar of relatively unconscious collective
creativity. (Or at least many human agents like to think!)
We can think of agency as having two key components. First, there is
flexibility. This is about who (or what) has a hand or say in the creation of
some effect, object, idea, or action: from theorems and inventions to floods
and revolutions, from commodities and crimes to conversations and crises.
In most cases, many “hands” contribute, at different degrees of remove,
with more or less effort and effect, and with greater or lesser visibility.
These hands are social as much as material; cognitive as much as somatic;
individual as much as collective; natural as much as artificial; flexibly cre-
ative as much as rigidly causal. Indeed, as soon as one starts tracking such
connections, the distinctions often break down. It is hard to see where cog-
nition ends and embodiment begins; where the divide between nature and
artifice is to be drawn; where to cut off an individual from a collectivity.
The second component of agency is accountability. We must ask how
those with agency are held accountable (usually by other kinds of agents)
for their contribution. Who realizes the cost of a product, or garners the
right to use it? Who is blamed for the failure of an action? Who is praised
for the genesis of an idea? Who or what benefits from a new adaptation?
Who berates themselves over an imagined sin? Through what mechanisms?
Accountability is identified in different ways, using judgments with varying
degrees of justness, rationality, and depth of vision. Accountability is both
moral and economic, causal and normative, grounded in natural selection
and cultural evaluation.
Current understandings of human agency suggest that our flexibility
and accountability are often not located in the individual alone, but are
radically distributed. This suggests a number of questions, which guided the
discussions that have shaped the contributions to the book.1 Among these
are the following:

• Why and how do we excerpt agents at certain scales, and how do differ-
ent understandings of agency alter such scales?
• When we cooperate with others, when is our agency weakened, and
when is it enhanced?
• In what ways is material culture an extension of the person?
• While artifacts and technologies increase our flexibility in obvious ways,
how do they affect our accountability?
• What are the causal forces that determine the scope and constraints of
our agency? How are these forces related to norms?

[ xii ]  Editors’ Preface

╇ xiii

• Why is it often considered appropriate to punish or reward a person for

something that someone else did?
• How is it that multiple individuals can behave as one?
• What is the role of distributed agency in human evolution, and in
ontogeny?

In short, who or what contributes to some action, idea, object, or outcome?

And who or what is held accountable? In what ways, and to what degrees?
With what conditions, and with what consequences? These are the ques-
tions we ask in this book.

CONTENTS

The ten parts of the book each group together essays that overlap in regard
to their empirical topics and analytic commitments. Part One, “Agency as
Flexible and Accountable Causality,” introduces key themes through four
interrelated chapters. In the first two chapters, Enfield theorizes some ele-
ments of agency, broadly understood as flexible and accountable causality;
and then describes how agency, thus framed, is distributed across actors
and activities. In the second two chapters, Kockelman reviews and synthe-
sizes some classic texts that lead to such an understanding of agency; and
then describes a variety of other possible ways of framing it.
The next nine parts take up interrelated questions from a variety of
stances, partly deriving from the cross-╉ disciplinary commitments of
authors, and partly from the empirical contents they analyze and the theo-
retical questions they answer.
In Part Two, “The Agency of Institutions and Infrastructure,” Bernstein
takes up agency in actual state agencies—╉in particular, legal and political
institutions that promulgate laws and regulations shaping the actions of
citizens—╉and Elyachar takes up agency in the context of political revolu-
tions, when critical energies are used to create, use, and destroy key forms
of infrastructure.
Part Three presents perspectives on “Language and Agency.” While
much previous work on language and agency has focused on the repre-
sentational and meta-╉representational capacities of humans, these four
chapters take us into the trenches of interaction: talk-╉in-╉interaction as it
unfolds in real-╉time practices, in face-╉to-╉face encounters. Dingemanse uses
the science fiction fantasy of brain-╉to-╉brain interfaces to shed light on the
natural distribution of agency in human communication; Floyd focuses
on requests as a key means for instigating and negotiating the sharing of

Editors’ Prefaceâ•… [â•›xiiiâ•›]

xiv

agency; also on this theme, Rossi and Zinken focus in more closely on the
fine-​grained grammatical categories that languages provide as tools for
mobilizing other people; and Sidnell focuses on a particularly important
but overlooked element of the shared construction of action in interaction,
namely the accountability that comes with our impositions on others.
In Part Four, “Economy and Agency,” Guyer examines the ethics of debt
as a way of distributing agency both across people and through time, and
Maurer looks at the infrastructures of monetary accounting, where the act
of “recording” an exchange can in fact be seen as effecting the very exchange
itself. In Part Five, “Distributing Agency within Selves and Species,” Parry
explores the distribution of agency in moving the human body in profes-
sional therapy sessions, and d’Ettorre—​on a very different scale—​shows
how agency is distributed across masses of individual bodies in the case of
eusocial organisms such as ants.
In Part Six, “Social Bonding Through Embodied Agency,” the theme of
agency in human groups is further developed, with an overview by Cohen
of some elements of social bonding through group exercise in humans, and
a focus by Tarr on the same in dance and music. And in Part Seven, “Agency
and Infancy,” Rączaszek-​Leonardi highlights the importance of differ-
ent timescales in an exploration of the understanding of agency between
infants and caregivers, and Tunçgenç stresses the importance of close tem-
poral coordination in the bodily movements of infants and caregivers as a
crucial step on the path from individual to collective agency in the lifespan.
Part Eight, “The Agency of Materiality,” presents three studies of how
agency gets into, and out of, inanimate objects: Crossland discusses the
agency of human remains, as understood through forensic anthropology;
Smith considers the agentive role of marbles in the lives of children; and
Wilf explores how computerized algorithms can be a source of contingen-
cies that are harvested to cultivate artistic creativity. In Part Nine, “The
Place of Agency,” Adams takes up agency through the eyes of a geographer
focused on spatially distributed selves, and Lahlou looks at it through the
lens of installations—​or spatially arranged ensembles of affordances that
shape the activities of those who act and think inside them. And in Part
Ten, “From Cooperation to Deception and Disruption,” Schweikard outlines
the central importance of intentionality and normativity in the coopera-
tive framework of joint action, while Umbres turns our attention to decep-
tion—​a decidedly antisocial form of distributed agency—​and Zuckerman
delves further into the dark side of agency, with a study of attempts to
sabotage others’ courses of action, and the post hoc framings and attribu-
tions of agency that then emerge.

[ xiv ]  Editors’ Preface

╇ xv

ENVOI

Human agency concerns the fundamental conditions and constraints

under which we pursue our goals, from the simplest everyday actions to
the greatest uses and abuses of power in society. But a deep understand-
ing of human agency is elusive because the concept of agency has been
understood and examined in such different ways. The meaning of the term
agency is often radically simplified, reduced to or conflated with that of
other terms: compare free will, choice, language, power, generativity, imagi-
nation, self-╉consciousness. There have been important advances in separate
disciplines, but there has been little opportunity to combine and build on
the collective achievements of research on human agency carried out using
sometimes radically different approaches. Our goals with this collection
of essays are to assemble insights from new research on the anatomy of
human agency, to address divergent framings of the issues from different
disciplines, and to suggest directions for new debates and lines of research.
We hope that it will be a resource for researchers working on allied top-
ics, and for students learning about the elements of human-╉specific modes
of shared action, from causality, intentionality, and personhood to ethics,
punishment, and accountability.

NOTE

1. The forum for most of these discussions was a retreat held at Ringberg Castle,
Tegernsee/╉Kreuth, Germany, April 13–╉17, 2014, attended by most of the book’s
contributors, and several others. The retreat was funded by a European Research
Council grant “Human Sociality and Systems of Language Use” (Grant number
240853, to Enfield). The discussion continued by correspondence and led to the
further invitation and inclusion of several authors whose work was important
to the project. We are grateful to all participants and contributors for their
invaluable input. We thank Angela Terrill at Punctilious Editing (http://www.
punctilious.net/) for first-class indexing. Please note the following conventions
for transcripts in a number of chapters: underline = stress; [ = beginning of
overlap; ] = end of overlap; (4.4) = 4.4 seconds of silence; text in ((double
brackets)) refers to visible bodily behavior; text in (single brackets) = not clearly
audible.

Editors’ Prefaceâ•… [â•›xvâ•›]

  xvii

CONTRIBUTORS

Paul C. Adams, Department of Geography and the Environment, University

of Texas at Austin
Anya Bernstein, Law School, SUNY Buffalo
Emma Cohen, Institute of Cognitive and Evolutionary Anthropology,
University of Oxford
Zoë Crossland, Department of Anthropology, Columbia University
Patrizia d’Ettorre, Laboratory of Experimental and Comparative Ethology,
University of Paris 13, Sorbonne Paris Cité
Mark Dingemanse, Max Planck Institute for Psycholinguistics
Julia Elyachar, Department of Anthropology, Princeton University
N. J. Enfield, Department of Linguistics, The University of Sydney
Simeon Floyd, SENESCYT (National Science Ministry), Ecuador; Radboud
University Nijmegen, and Max Planck Institute for Psycholinguistics
Jane I. Guyer, Department of Anthropology, Johns Hopkins University
Paul Kockelman, Department of Anthropology, Yale University
Saadi Lahlou, London School of Economics and Political Science
Bill Maurer, Department of Anthropology, University of California
at Irvine
Ruth H. Parry, Faculty of Medicine and Health Sciences, University of
Nottingham
Joanna Rączaszek-​Leonardi, Faculty of Psychology, University of Warsaw
Giovanni Rossi, University of Helsinki
xviii

David P. Schweikard, Department of Philosophy, Europa-​

Universität
Flensburg
Jack Sidnell, Departments of Anthropology and Linguistics, University of
Toronto
Benjamin Smith, Department of Anthropology, Sonoma State University
Bronwyn Tarr, Social and Evolutionary Neuroscience Research Group,
University of Oxford
Bahar Tunçgenç, Institute of Cognitive and Evolutionary Anthropology,
University of Oxford
Radu Umbres, Faculty of Political Sciences, National School of Political and
Administrative Sciences, Bucharest
Eitan Wilf, Department of Sociology and Anthropology, Hebrew University
Jörg Zinken, Institute for the German Language (IDS), Mannheim
Charles H.  P. Zuckerman, Department of Anthropology, University of
Michigan

[ xviii ] Contributors
  1

PART ONE

Agency as Flexible
and Accountable Causality
  3

CHAPTER 1

Elements of Agency
N. J. ENFIELD

O n October 15, 1965, on Whitehall Street in Lower Manhattan,

David J. Miller burned his draft card in protest at the Vietnam War
(Miller 2001):

The expectant crowd fell hush in front of me. The hecklers across the street
ceased their ranting and watched silently. An eerie stillness settled upon our
canyon as the last rays of the fall sun clung to the tops of the buildings. I said the
first thing that came to my mind. “I am not going to give my prepared speech.
I am going to let this action speak for itself. I know that you people across the
street really know what is happening in Vietnam. I am opposed to the draft and
the war in Vietnam.”
I pulled my draft classification card from my suit coat pocket along with a
book of matches brought especially for the occasion since I did not smoke. I lit a
match, then another. They blew out in the late afternoon breeze. As I struggled
with the matches, a young man with a May 2nd Movement button on his jacket
held up a cigarette lighter. It worked just fine.
The draft card burned as I raised it aloft between the thumb and index finger
of my left hand. A roar of approval from the rally crowd greeted the enflamed
card. This awakened the momentarily mesmerized hecklers and they resumed
their shouts.
As the card burned, I discovered that I had made no preparation for the card
to be completely consumed. I dropped the card as the flame reached my finger-
tips. At my trial in federal court, the unburnt corner of my draft card, with a
4

bit of my signature, was introduced into evidence. The FBI had been Johnny-​
on-​the-​spot in retrieving the charred remains of my card so as to assist in their
prosecution even though I  never denied that I  burned my card. Future card
burners used tongs or cans in order to complete the job.
… Three days later the FBI swooped in on me in Manchester, New
Hampshire… . I made bail the next day. I remained free till June 1968 when
the draft-​card burning case finally lost in the U.S. Supreme Court. I  served
22 months in federal prison in Pennsylvania from 1968 to 1970.

Miller’s story of publicly burning his draft card as a means of protesting

against the US government’s program of conscription during the Vietnam
War illustrates the core elements of agency: he had a degree of flexibility in
carrying out the behavior, and a degree of accountability. What does this
mean? Let us start with the elements of flexibility:1

(1) Agents have flexibility over meaningful behavior, insofar as:

(a) to some degree they control or determine that the behavior is
done at a certain place and time (thus, to some degree Miller
determined that his draft card would be burned at that place
and time; though with minor delays due to his malfunctioning
matches);
(b) to some degree they compose or design the behavior as a means
for a particular end; a thing to be done and a way to do it (thus,
Miller at some level had the plan to destroy this draft card as a pro-
test against the war, and he determined sub-​plans as means to that
greater end);
(c) to some degree they subprehend2 or anticipate how others could
view and react to the behavior; for instance to some extent they
may be prepared for certain interpretants—​in other words, ratio-
nal responses—​by others; they may be surprised or disposed to
sanction non-​anticipated interpretants (thus, Miller was not sur-
prised when hecklers tried to disrupt him, when FBI agents took
his burned card as evidence, or when they arrested him in the fol-
lowing days; nor when he was praised as a hero by members of the
antiwar movement).

These elements of flexibility are understood in terms of the semiotic process

that underlies an agent’s behavior (Kockelman 2007, 2013). Thus, control-
ling (1a) is about the production of perceptible behavior—​potential signs—​
without respect to what that perceptible behavior may be taken to mean, or

[ 4 ]  Agency as Flexible and Accountable Causality

  5

to what interpretants it might so elicit. In Miller’s case, controlling simply

had to do with instigating an event of fire destroying a piece of paper.
Composing (1b) is about the relation between an agent and an over-
all semiotic process that this agent instigates. Composing in this sense
concerns the degree to which an agent selects that behavior for a func-
tion, as a means to an end, and the degree to which her carrying out
of the behavior is responsible for success in achieving her goal. Miller’s
goal was to protest against the war; as an agent, he played a role in
determining which behavior would be produced and what that behavior
may stand for. His selection of tools (draft card, fire) as means toward a
local end (destroying the physical manifestation of a government order)
determined the conditions for success in his venture: without the card
actually catching fire and burning up, the planned action would not have
been consummated. As Miller describes, there were minor obstacles to
the execution of the card-​burning behavior, related to the affordances of
fire and paper: for one thing, the wind blew out his matches on his first
attempts, and for another, the heat of the burning paper caused him to
let go of the card and in the end not succeed in fully destroying it. For the
behavior to succeed as Miller composed it, the effective destruction of a
specific piece of paper was necessary to the consummation of a specific
political act.
Within composing, we can distinguish between execution and planning.
Differences in the manner of execution of a sign, which may be functions
of the skill or conditions of an agent, will have consequences for how effec-
tively a sign will succeed in standing for a particular object. While execu-
tion is the degree to which one’s physical production of the relevant sign is
done in such a way as to determine that the sign should be taken to stand
for a certain object, planning is deciding which sign-​object relations are
to be created in the first place. Miller planned to put fire to paper, but his
execution was lacking. In the end his plan was carried out, executed with
the help of a bystander with a lighter.
Subprehending (1c) is about the relation between the agent and the
overall semiotic process. Subprehension concerns the degree to which an
individual effectively foresees the interpretants—​in other words, the reac-
tions and responses—​that the sign event may evince. One way to measure
this would be in the ways a person might reveal that they have prepared
for a certain interpretant and not others. For example, if I ask What’s his
address? while at the same time pulling out a pen and paper to write the
address down, by pulling out the pen and paper I am showing evidence
of having subprehended that your interpretant (your answer) of my sign

E l e m e n t s of   Ag e n c y   [ 5 ]
6

(my question) will be to tell me the person’s address, which I will then be
already prepared to write down. Another way to measure subprehension in
this sense is to monitor the accountability involved. Is a person surprised
by or disposed to sanction certain interpretants? If I ask What’s his address?
and you look at me and say nothing, I might repeat the question or say Hey
I asked you something. An agent subprehends interpretants of a sign event
to some degree (especially the potential accountability): being prepared
for some interpretants and not others; being surprised by, or disposed to
sanction, some interpretants and not others. As noted above, Miller was
surely not surprised by the heckling, the arrest, the conviction; nor was
he presumably surprised by the praise, support, and imitation from other
activists. This indicates that his degree of agency on the measure of subpre-
hension was high.
With an agent’s flexibility comes accountability, as Miller anticipated or
at least subprehended:

(2) Agents have accountability for meaningful behavior, insofar as:

(a) they may be subject to public evaluation by others for their behav-
ior, where this evaluation—​in the form of interpretants like praise,
blame, or demand for reasons—​may focus on any of the distinct
components of flexibility given in (1), above (thus, Miller’s actions
were evaluated by many people from onlookers at the scene to fed-
eral judges in the US Supreme Court);
(b) they may be regarded as having some degree of entitlement to carry
out the behavior, and give reasons for it, and they or others may
invoke this entitlement; this may relate to any of the distinct com-
ponents of flexibility given in (1), above (thus, Miller’s behavior
was precisely designed to give him an opportunity to state his rea-
sons for action, with the intention to publicize those reasons for
action);
(c) they may be regarded by others as having some degree of obliga-
tion to carry out the behavior, and give reasons for it, and they or
others may invoke this obligation; this may relate to any of the dis-
tinct components of flexibility laid out in (1), above (thus, Miller’s
behavior could be used as a basis for motivating others, possibly
through a sense of moral obligation, to do the same).

Another way of framing accountability in relation to behavior in the sense

just defined is ownership of the behavior. Thinking in terms of accountabil-
ity foregrounds the possibility of blame or praise, while ownership fore-
grounds rights and obligations.

[ 6 ]  Agency as Flexible and Accountable Causality

  7

So, in sum, agency is the relation between a person and a course of

action and its effects. To summarize it in the leanest terms possible:

(3) With regard to some goal-​directed controlled behavior, agency consists

of the following:
(A) Flexibility
1. Controlling (determining that a perceptible/​physical behavior
occurs)
2. Composing (selecting the behavior, its function, and execution)
3. Subprehending (effectively anticipating interpretants of the
behavior)
(B) Accountability
1. Being evaluated (by others, on any of [A1–​3])
2. Being entitled (a right to do the behavior can be recognized and
invoked)
3. Being obligated (a duty to do the behavior can be recognized
and invoked)

There is a special relation between flexibility and accountability. Flexibility

is regimented both by natural laws and by social norms, and often by these
two in combination. One’s accountability is lower when one’s behavior is
more heavily constrained by natural laws, because less choice is involved.
Supporters of Miller are unlikely to sanction him for having delayed the
burning of the card before he was given the cigarette lighter; they would rec-
ognize that his flexibility was at that moment thwarted by natural causes,
and not by, say, some form of hesitation in anticipation of likely sanctions.
And so we arrive at Kockelman’s definitive equation: Agency equals flex-
ibility plus accountability. How flexible you are depends on how freely you
can determine the elements of a course of behavior and its outcomes, in
multiple senses: the physical carrying out of the behavior, the planning and
design of the behavior, the placing of the behavior in an appropriate context,
the anticipation or subprehension of likely effects of the behavior—​includ-
ing, especially, the reactions of others—​in that context. How accountable
you are depends on how much it can be expected or demanded that other
people will interpret what you do in certain ways, for example, by respond-
ing, asking for reasons, sanctioning, praising, or blaming you. With these
elements distinguished, we begin to understand why the concept of agency
is far from simple or primitive, and why it has resisted easy definition. With
the elements of agency distinguished, we may begin to understand the
many subtleties of the agency problem, something that is especially needed
when we want to take the next step and ask how agency is distributed.

E l e m e n t s of   Ag e n c y   [ 7 ]
8

ACKNOWLEDGMENTS

Thanks go to Paul Kockelman for his significant input to this work. Note that
this chapter draws partly on sections of my 2013 book Relationship Thinking
(Oxford University Press). I gratefully acknowledge support from the Max
Planck Society (through MPI Nijmegen), the European Research Council
(through grant 240853, “Human Sociality and Systems of Language Use”),
and the University of Sydney (through Bridging Grant ID 176605).

NOTES

1. This analysis of flexibility in agency is due to Kockelman (2007, 2013), though

I take one liberty with the terminology (Enfield 2013:104–╉117). Kockelman’s
term for (1c) is “commitment,” defined as “the degree to which one may
anticipate an interpretant, where this anticipation is evinced in being surprised
by and/╉or disposed to sanction unanticipated interpretants” (Kockelman
2007:380). I propose “subprehension,” to avoid confusion with other dominant
meanings of the word “commitment” (for example, in Kockelman 2007:153 and
passim, “commitment” has a technical meaning in the domain of accountability,
referring to “deontic obligation,” something one is obliged to do). On agency and
language specifically, see Ahearn (2010), Duranti (2004).
2. Subprehend may be defined as follows. If you subprehend something, it is as if
you anticipate or expect it, but not in any active or conscious way; rather, if you
subprehend something, when it happens you cannot say later that you had not
anticipated or expected it. Subprehension is thus close to the notion of habitus
(Bourdieu 1977).

REFERENCES

Ahearn, Laura M. 2010. “Agency and Language.” Society and Language Use 7: 28.
Bourdieu, Pierre. 1977. Outline of a Theory of Practice. Cambridge: Cambridge
University Press.
Duranti, Alessandro. 2004. “Agency in Language.” In A Companion to Linguistic
Anthropology, edited by D. Alessandro, 451–473. Malden, MA: Blackwell.
Enfield, N. J. 2013. Relationship Thinking: Agency, Enchrony, and Human Sociality.
Oxford: Oxford University Press.
Kockelman, Paul. 2007. “Agency: The Relation Between Meaning, Power, and
Knowledge.” Current Anthropology 48(3): 375–╉401.
Kockelman, Paul. 2013. Agent, Person, Subject, Self: A Theory of Ontology, Interaction,
and Infrastructure. Oxford: Oxford University Press.
Miller, David. 2001. “Reclaiming Our History: Memoirs of a Draft-╉Card Burner.”
Reclaiming Quarterly (82). http://╉www.reclaimingquarterly.org/╉82/╉rq-╉82-╉
draftcard.html Retrieved March 12, 2011.

[â•›8â•›]â•… Agency as Flexible and Accountable Causality

  9

CHAPTER 2

Distribution of Agency
N. J. ENFIELD

I t is easy to think that “an agent” should coincide exactly with an individ-
ual. But this is seldom, if ever, the case. One reason agents do not equal
individuals is that the elements of agency can be divided up and shared out
among multiple people in relation to a single course of action. When I get
you to pass the salt, it is me who plans the behavior but you who executes
it. Or when I report what a candidate said in yesterday’s speech, it is me
who speaks the words but the candidate who is accountable for what was
expressed. With distributed agency, multiple people act as one, sharing or
sharing out the elements of agency.
One person may provide the flexibility needed for meeting another per-
son’s ends—​as in slavery, factories, and armies—​or for meeting shared
ends—​as in team sports, co-​authorship, and joint enterprise. Similarly,
one person may bear the accountability entailed by another person’s flex-
ibility. It is hardly rare for a person to inherit the blame for someone else’s
punishable actions. Howitt (1904), writing on Aboriginal Australia, “men-
tions the case of an accused man pointing to his elder brother to take the
blame, because an elder brother stands for and should, ideally, protect a
younger” (Berndt and Berndt 1964:299). In law of this kind, “people are
categorized in units the members of which are interdependent … [and]
relations both within and between those units may not be framed only, or
predominantly, in kinship terms” (Berndt and Berndt 1964:303). This non-​
modern approach has difficulty fitting into today’s world. In 2009, there
was a campaign in the Australian Aboriginal community at Lajamanu in
10

the Northern Territory to find a solution to persistent social problems aris-

ing from the incompatibility of indigenous and government law. Here was
the problem. If people who committed crimes were jailed and thus sepa-
rated from their community before they could face traditional punishment
(known as payback) then this payback would be meted out on family mem-
bers of the transgressors, often leading to continuing feuds. The campaign
called for official recognition of traditional punishment, allowing that bail
be granted to indigenous offenders so that they might face punishment
(for example, being speared through the thigh) and thus resolve the matter
to the satisfaction of the community before being taken to jail.
Distributed accountability of the kind that the Lajamanu campaign
sought to avoid is the essence of the feud, a universal phenomenon. Take
for example the gjakmarrja—​or blood feuds—​of rural Albania. Blood feud-
ing was largely dormant during the communist period, but with the fall
of communism in the early 1990s, old feuds were rekindled, and people
found themselves being held accountable for things that were done before
they were born, sometimes by people they had never met. Here is a case
(Mustafa and Young 2008:99–​100):

A family living in a village outside Bajram Curri … came under a new threat,
even though the man from their family who had committed a blood feud killing
had himself died seventeen years ago. The perpetrator’s death did not satisfy
the indebted family who sent a warning that they expected “blood payment” in
the form of the life of a male member of the remaining family. The family under
threat consisted of a widow living with her two sons, their wives, and several
children. The family had already adjusted to the threatening situation: the eldest
son had gone into hiding elsewhere, and the younger remained indoors, unable
to leave the house. Obviously this situation had a drastic effect on the earning
capacity of the family.

Cases like this take us to the heart of distributed agency. One man commits
a misdeed against another, and yet revenge is taken years later between
the two men’s grandchildren, neither of whom was involved in the original
transgression. Here, someone is held to account for something that some-
one else chose to do. Here, agency, with its components of flexibility and
accountability, is divided and shared out among multiple individuals while
still being anchored in a single, sometimes decades-​long course of action.
Cases like this highlight one of the key reasons why agents do not equal
individuals: the locus of agency is the social unit, and social units are not
confined to individual bodies.

[ 10 ]  Agency as Flexible and Accountable Causality

  11

This is another, related reason agents do not equal individuals: social

units are often compound persons. People are able to merge together to
form single units of motivation, flexibility, and accountability. If you and
I agree to make sauce, we commit to a common course of behavior, which
means that we have joint reasons for action. We act as one. We may carry
out different subcomponents of the job of preparation—​you pour; I stir—​
but normally in the end we will agree to share the praise or blame for suc-
cess or failure. So just as an individual may act as a distinct unit of social
action and accountability, so may she be part of a larger unit, a unit that
incorporates other people.
The fission or dividing of agency between people is familiar from exam-
ples like speechwriters and messengers: two individuals own two parts of
the agency involved. Joint action is the inverse of this. Joint action is the
fusion or unifying of individuals into single, compound units of agency.
People can become socially fused in common action, especially with regard
to their accountability. But to define joint action, it is not enough to say
that two or more people do the action together. Because we are talking
about agency, we must work in terms of agency’s distinct elements. A com-
prehensive examination of joint action must distinguish joint agency at
its logically distinct levels:  joint controlling, joint composing, joint sub-
prehending, joint evaluability, joint entitlement, joint obligation. Multiple
individuals may together inhabit a single social unit with reference to any
of these elements of agency.
Numerous species display fission-​fusion social organization, “chang[ing]
the size of their groups by means of the fission and fusion of subunits …
according to both their activity and the availability and distribution of
resources” (Aureli et  al. 2008:627). These social dynamics arise naturally
from the fact that we live in large and intensive social groups yet we are
separate individuals. The fission-​fusion dynamic in humans can be fast-​
paced. Through the course of any day we move in and out of membership
of transient, often momentary pairs or groups. These changes may occur at
fine time scales, such as when we switch in conversation from the status of
speaker versus addressee and back again, or in a chess game when we alter-
natively have the status of being the one whose turn it is. Or the changes
may occur at longer time scales, less frequent in the life span and typically
with greater ceremony, such as when we change from single to married, or
from lay to initiated. Fission-​fusion dynamics in human social life must be
understood not just in terms of group size and location but also in terms of
the numerous if not innumerable relationship types that ultimately define
a society. These relationships may be called statuses, as defined originally

Di s t r i b u t i o n of   Ag e n c y   [ 11 ]
12

by the anthropologist Ralph Linton: sets of rights and duties that hold with
respect to certain others. Much of social life is about managing changes of
status. The navigation of status often involves moving in and out of mem-
bership in composite units of agency. Changes of status, at all levels of
grain, are closely associated with fission-​fusion agency.
For joint action, there needs to be joint commitment (Clark 2006). In
the philosopher John Searle’s terms, joint commitment might be defined
as a status function declaration of one’s acceptance of some form of shared
accountability (Searle 2010). Psychologist Herb Clark uses this notion to
reanalyze the notorious experiments conducted by Stanley Milgram in New
Haven in the 1960s (Milgram 1974). Volunteers in an experimental setting
were instructed to administer electric shocks to another volunteer when-
ever he made an error in a memory task. They were surprisingly willing to
inflict harm in this experiment, a fact that has been analyzed as having
to do with obedience to authority. Clark reinterprets the finding, suggest-
ing that joint commitment is what really accounts for why volunteers in
these experiments ended up behaving in ways they would otherwise have
abhorred. These volunteers were not merely doing what they were told. By
agreeing to participate in the experiment in the first place, they had made
a pact with the experimenter, and people are—​evidently—​deeply reluc-
tant to withdraw from social pacts. Clark uses Milgram’s extreme example
of joint commitment for illustration, but his point is that these pacts are
being made by all of us, all the time. They are a constant and essential part
of social life and they are implicit in our every move. But when you jointly
commit, what are you committing to? The answer is that you are agreeing
to merge, on some level, with another individual in carrying out a single
course of action. Clark’s point is that you then become socially and mor-
ally accountable for reneging on that agreement. This is cognitively under-
pinned by what students of human sociality, from philosophy (e.g., John
Searle, Raimo Tuomela) to biology (e.g., Michael Tomasello, Josep Call),
refer to as shared intentionality (Tomasello et al. 2005).
Distributed cognition (Hutchins 2006) is a type of distributed agency
in the sense meant here. Often, distributed cognition encompasses physi-
cal action and not just cognitive processes like reasoning. But since cog-
nition is more generally a kind of goal-​directed flexible behavior, then
distributed cognition can also be analyzed in terms of the three elements
of flexibility in agency: animating, authoring, and subprehending (and
their interpreter corollaries: perceiving, ascribing, and interpreting). We
should be able to improve our accounts of distributed cognition by explor-
ing implications of the ideas that sometimes the distribution of cognition
is in the realm of what we do with our bodies (controlling and executing);

[ 12 ]  Agency as Flexible and Accountable Causality

  13

sometimes it is in the planning or authoring of what is to be done and how

it is to be done (composing); sometimes it is in the figuring out of why
to do it or what will result (subprehending, anticipating, projecting); and
sometimes it is in associated aspects of accountability: evaluation, entitle-
ment, obligation.
People fission and fuse not just with other people but also with arti-
facts:  scallops, mosquitoes, and furniture can have agency too (Latour
2005). When the anthropologist Alfred Gell used the term “distributed
personhood,” he was pointing out that the locus of individuals’ agency is
not confined to the borders of their bodies (Gell 1998). His example was of
the soldiers who laid land mines in Pol Pot’s Cambodia. “The soldier’s weap-
ons are parts of him which make him what he is,” Gell wrote (1998:20–​21).
“As agents, they were not just where their bodies were, but in many dif-
ferent places (and times) simultaneously. Those mines were components
of their identities as human persons, just as much as their fingerprints or
the litanies of hate and fear which inspired their actions.” Accountability,
too, can be materially extended, but with limits. If a land mine has agency,
it is surely less than a man’s, and different in kind. At least, an inanimate
object may play a role in a causal chain of events with social consequences.
Laidlaw (2010) gives the example of a vase that falls off your apartment
window ledge, hurting a passerby. You, and not the vase, become account-
able for this “act.” You and the vase are not equal members of a fused unit
of accountability. Suppose you are caught stealing money and hiding it in
a vase. Could you propose to the court that your vase be held to account,
requesting that the vase be punished instead of you? While it is clear that
agency is radically distributed, extending to objects and artifacts, examples
like this show that there are strong asymmetries.
A large part of social interaction is about solving a me/​us problem, in real
time: Who is doing this, me or us? This requires us to navigate the fission
and fusion by which we exit and enter units of social agency together with
others. We need to do this because there is a persistent mismatch between
the immutable separateness of our physical bodies, on the one hand, and
the constant malleability of the units of social agency that we inhabit from
moment to moment, on the other. Am I  acting on my behalf alone? Or
are my current goals, my current reasons for action, shared with others as
well? Is this behavior by me or by us? Social interaction is a fission-​fusion
affair involving constant navigation of separateness and boundedness,
affiliation and disaffiliation, and a tacking back and forth between inhabit-
ing different social units. This has always-​relevant consequences for our
social relationships, both fleeting and enduring, and for the rights, duties,
and mutual dispositions that define these relationships.

Di s t r i b u t i o n of   Ag e n c y   [ 13 ]
14

ACKNOWLEDGMENTS

Thanks go to Paul Kockelman for his significant input to this work. Note that
this chapter draws partly on sections of my 2013 book Relationship Thinking
(Oxford University Press). I gratefully acknowledge support from the Max
Planck Society (through MPI Nijmegen), the European Research Council
(through grant 240853 “Human Sociality and Systems of Language Use”),
and the University of Sydney (through Bridging Grant ID 176605).

REFERENCES

Aureli, F., C. M. Schaffner, C. Boesch, S. K. Bearder, J. Call, C. A. Chapman, R. Connor,

A. Di Fiore, R. I. M. Dunbar, and S. P. Henzi. 2008. “Fission-╉Fusion Dynamics.”
Current Anthropology 49(4): 627–╉654.
Berndt, Ronald M., and Catherine H. Berndt. 1964. The World of the First Australians.
Sydney: Ure Smith.
Clark, Herbert H. 2006. “Social Actions, Social Commitments.” In Roots of Human
Sociality: Culture, Cognition, and Interaction, edited by N. J. Enfield and Stephen
C. Levinson, 126–╉152. London: Berg.
Gell, Alfred. 1998. Art and Agency. Oxford: Clarendon Press.
Howitt, Alfred William. 1904. The Native Tribes of South-╉East Australia.
Cambridge: Cambridge University Press.
Hutchins, Edwin. 2006. “The Distributed Cognition Perspective on Human
Interaction.” In Roots of Human Sociality: Culture, Cognition and Interaction,
edited by N. J. Enfield and Stephen C. Levinson, 375–╉398. Oxford, UK: Berg.
Laidlaw, James. 2010. “Agency and Responsibility: Perhaps You Can Have Too Much
of a Good Thing.” In Ordinary Ethics: Anthropology, Language, and Action, edited
by Michael Lambek, 143–╉164. New York: Fordham University Press.
Latour, Bruno. 2005. Reassembling the Social: An Introduction to Actor-╉Network-╉Theory.
Oxford: Oxford University Press.
Milgram, Stanley. 1974. Obedience to Authority: An Experimental View.
New York: Harper & Row.
Mustafa, Mentor, and Antonia Young. 2008. “Feud Narratives: Contemporary
Deployments of Kanun in Shala Valley, Northern Albania.” Anthropological
Notebooks 14(2): 87–╉107.
Searle, John R. 2010. Making the Social World: The Structure of Human Civilization.
New York: Oxford University Press.
Tomasello, Michael, Malinda Carpenter, Josep Call, Tanya Behne, and Henrike
Moll. 2005. “Understanding and Sharing Intentions: The Origins of Cultural
Cognition.” Behavioral and Brain Sciences 28(5): 664–╉670.

[â•›14â•›]â•… Agency as Flexible and Accountable Causality

  15

CHAPTER 3

Gnomic Agency
PAUL KOCKELMAN

T here are many different ways of framing agency, and thereby fore-
grounding different kinds of agents. My point in what follows is not to
endorse any particular frame, but merely to sketch some of the key features
of several pervasive frames. Such frames—​as ways of understanding and
interrelating flexibility, causality, and accountability—​have grounded the
intuitions of many influential thinkers. And so it is useful to understand, if
only to undermine, their characteristic assumptions.

ARISTOTELIAN AGENCIES

Aristotle (2001) famously described four kinds of causes that may underlie
any entity. There is the material cause (the substance something is com-
posed of, however heterogeneous), the formal cause (the way this substance
has been shaped, organized, or patterned), the final cause (the functions
such a formed substance may serve), and the efficient cause (that which
gives form to substance, often for the sake of some function). Aristotle’s
word for cause was aition, which is closely linked to notions of responsi-
bility. In some sense, such causes are responsible for the existence of an
entity. This is not to say that they should be held accountable in any legal
or moral sense, but only to say that we may make reference to such causes
when we try to account for such entities.
Such causes, broadly construed, provide one useful way of framing
agency. In particular, we might define (Aristotelian) agents as causes that
16

we can take account of. Acting alone, or in concert, they are a salient condi-
tion of possibility for the existence of entities, or the occurrence of events.
Natural selection is thus an agent. A worker is an agent. A tool is an agent.
An enzyme is an agent. Oxidation is an agent. Even lightning bolts, and
chance phenomena more generally, are agents. Note, then, that while it
is easy to make a splash by calling some non-​intentional actor an “agent”
(a land mine, dust, concrete, continental drift, worms, the trade winds,
a comet, etc.), this is only because so many people have a limited sense
of what an agent is. In this Aristotelian framing, in contrast, anything of
causal account is an agent and, indeed, interestingly so. Such a framing is
thus not limited to the kinds of causes enumerated by Aristotle, but may
include whatever causes are (deemed) worthy of account: ghosts and grav-
ity, sieving and serendipity, witches and wishes, global warming and politi-
cal apathy, semiotic practices and thermodynamic processes, and much
else besides.
Needless to say, most entities and events have many causes (each of
which can itself be an entity or event with many causes, and so on, ad infi-
nitum). And so agency is necessarily distributed in depth (each cause of
some effect is itself the effect of some cause), and necessarily distributed
in breadth (each effect has many causes; each cause has many effects). This
is not to say that all such causes are salient enough to show up in some
account: causal accounting, as the figuring (out) of agency, tends to focus on
those causes that are particularly relevant to the agents who are doing the
accounting: for example, those causes that are least expected; those causes
most easy or difficult to intervene in; those causes that are most powerful,
marketable, or useful; those causes most amenable to our intuitions; those
causes that seem most intentional; those causes with the most dire effects;
and so forth. To study agency is to study causal understandings, and imag-
inings, of the world: not just the conditions for, and consequences of, enti-
ties and events; but also the conditions for, and consequences of, particular
accountings (of such conditions and consequences). There is no interesting
account of agency that is not simultaneously an account of those agents who are
trying to account for agency.

BACONIAN AGENCIES

We may extend this framing, and get a better sense of its stakes, by apply-
ing some categories of Bacon (2000) to those of Aristotle. If knowledge
turns on the discovery of causes, power turns on the directing of causes.
Knowledge and power are thus necessarily coupled:  in directing known

[ 16 ]  Agency as Flexible and Accountable Causality

  17

causes, we may discover new causes; and we may use such discoveries in
our subsequent directings. From this standpoint, a Baconian agent exhib-
its a kind of meta-​agency: whatever has knowledge about, or power over,
the sorts of causes described above. Such an agent can offer an account of
such causes, so far as it can discover them; and it may make such causes
count, so far as it can direct them. Indeed, for such agents, we can often
invoke accountability in the strong sense. That is, not only can we account
for such agents (e.g., thematize them, characterize them, reason about
them), but we can also hold such agents accountable for their actions and
effects (e.g., hinder them or hasten them, reward them or punish them,
praise them or shame them). Finally, and reflexively, it is often the case
that such Baconian agents have yet to discover all the causes that direct
them. Such “blind spots” constitute one key sense of the unconscious: we
often have limited knowledge over precisely those causes that have power
over us.
This way of framing agency foregrounds not only the recursive and
reflexive nature of agency, but also the relative nature of agency. In partic-
ular, different agents have different degrees of knowledge and power, and
so different degrees of agency. But, that said, because there are so many
different kinds of causes, most kinds of power and knowledge are incom-
mensurate, so that it makes little sense to contrast them by degree. For
example, while we might compare the strength of two men, it is difficult
to compare the strength of one man with the speed of another. Similarly,
while we might compare how much you know about chemistry with how
much I know, it is difficult to compare your factual knowledge (know that)
with my practical knowledge (know how). And so on, and so forth. That is,
it only really makes sense to talk about different degrees of agency when
we can isolate out a shared dimension of agency (for example, how much
money, or credit, one has at one’s disposal). And for many, if not most,
kinds of agents, with their potentially heterogeneous suites of context-​
specific and event-​contingent causal capacities, that is impossible. Modes
of agency are as heterogeneous and incommensurate as the causes they
direct and discover, as the entities such causes enable, and as the events
such causes occasion.

INSTRUMENTAL AGENCIES

Returning to Aristotle, it is sometimes useful to set aside material causes

and formal causes, and instead focus on efficient causes (understood as
“means”) and final causes (understood as “ends”). In this more restricted

G n o m i c   Ag e n c y   [ 17 ]
18

framing, agents are relatively instrumental: they wield a variety of means

to achieve a variety of ends. Philosophers like James and Peirce went so
far as to define “lively agents” and “mindful agents” in related terms. For
example, James stated, “The pursuance of future ends and the choice of
means for their attainment are thus the mark and criterion of the pres-
ence of mentality” (1918:8). Indeed, in some sense, the Baconian framing
treats knowledge and power in terms of instrumental reasoning: to what
extent does an agent understand which causes lead to which effects; to
what extent can an agent create such causes (as means) in order to bring
about such effects (as ends).
Such agents might be considered more or less agentive as a function of
the range of means they have at their disposal, and as a function of the
variety of ends they may seek to achieve. That is, from the standpoint
of this particular framing, the more open or flexible an agent (in regard
to its means and ends), the more agentive that agent. Such means may
include tools, utterances, paths, forces, compositions, signs, representa-
tions, affects, traps, goods, and so forth. Such ends may include any kind of
resource or value potentially securable through such means: money, food,
sex, reputation, freedom, enlightenment, power, security, certainty, infor-
mation, territory, transcendence, immortality, revenge, and even agency
itself. And such openness or flexibility is often imagined to be causally
grounded in some putative human-​specific capacity: imagination, genera-
tivity, cognition, metaphor, displacement, the symbolic, language, mind,
choice, meta-​representations, culture, reason, and so forth.
That said, any number of distinctly “non-​lively” and “non-​mental” enti-
ties exhibit characteristics of instrumental agency:  they are means that
may be used to pursue ends (e.g., a rock used as a weapon); and they may
even have been made for the sake of achieving such ends (e.g., a bowl made
for holding soup). While such made or used agents reflect the purposes
of those agents who make them or use them, they don’t seem to under-
take purposeful actions themselves. For this reason, they are sometimes
understood as derivative agents, as opposed to originary agents. Note,
then, that from the standpoint of the first framing of agency, they would
be agents without qualification. But once we restrict ourselves to focusing
on means and ends, as a particular subset of Aristotle’s causes, such agents
are thereby demoted, showing up in causal accounts as “less than” fully
agentive. By way of contrast, the mindful or lively agents introduced above
often seem to exhibit a kind of reflexive agency:  they are auto-​technic
(using themselves as a means) and they are auto-​telic (having themselves
as an end). Indeed, it is often useful to unitize agency (and thereby enclose
“agents”) by reference to such modes of reflexive coherence.

[ 18 ]  Agency as Flexible and Accountable Causality

  19

EVALUATIVE AGENCIES

Just as any effect may itself be a cause that gives rise to another effect, any
end may itself be a means to a further end. And so we again find a rela-
tively recursive pattern: a means-​ends chain that stretches on indefinitely
in two directions at once. Or does it? We just saw how many affordances,
like the branch one uses to climb a tree, are means (insofar as one may
use them to undertake an action), but not themselves ends (insofar as the
branches were not made for the sake of climbing). Nature is often romanti-
cally understood to be constituted by this bottoming out of instrumental-
ity. Conversely, Aristotle argued that such means-​ends chains had to stop
somewhere, and thus “top in.” In particular, there must be some end that
is not itself the means for further ends. He called such a final end eudai-
monia, which is often translated as human flourishing, or happiness. And
he thought that this highest good, or supreme value, was of utmost impor-
tance to philosophy and politics.
Whether or not such higher goods are crucial to philosophy and politics,
they are certainly crucial to many understandings of agency, which are just
as prone to promote value-​oriented agents as they are to demote derivative
agents. In particular, a related tradition understands each of us instrumen-
tal agents as having too many ends, in that we have more desires than we
might ever hope to attain (given our finite time on this earth, and the finite
means at our disposal). And so a key question arises:  how do we decide
which desire we want to act upon, or which end we hope to achieve? To
which final end, or at least more distal end, should our intermediate ends
be oriented? Such a question presumes that we are not just instrumental
agents, who have some flexibility in regard to our means and ends, but that
we are also selecting and/​or economizing agents. That is, not only is there
more than one way to skin a cat, but there are more ways to spend our time
than skinning cats; and so we need to choose not just how to do things, but
what things to do. And so the question of evaluating agents arises: agents
who not only act instrumentally, but also evaluate instrumental acts in
reference to values and, in particular, in reference to values that could be
otherwise.
Such values are often understood to constitute some kind of relatively
shared standard that allows us to order means and ends according to their
relative desirability (Taylor 1989). Such standards help us determine which
course of action is more just, honorable, or efficient—​and hence which
course of action we should pursue if we are to be a just, honorable, or effi-
cient actor. Such a process is also recursive: one kind of agent may be rela-
tively flexible in regard to the means and ends it has at its disposal; another

G n o m i c   Ag e n c y   [ 19 ]
20

kind of agent, a kind of meta-​agent, may be relatively flexible in regard to

the standards of values it has at its disposal (to decide which means and
ends are best); and so on. That is, just as one can order desires relative to a
standard, one can order standards relative to a meta-​standard (and so on,
recursively). In such a situation, the question is not so much Which route
through life should I take, but rather Which map should I use to assess which
routes there are, how should these routes be weighted, and what constitutes the
terrain in the first place?
Such agents are often held accountable to a large degree: not only can
they account for their own values (in the sense of being able to thematize,
characterize, and reason about them), but they can also be held account-
able for upholding certain values (in the sense of being praised or pun-
ished as a function of whether or not they follow them). Indeed, quite
crucially, they can often be praised for pursuing the “correct” end, or pun-
ished for pursuing the “incorrect” end, even if they fail to achieve it. And
many such agents often hold themselves reflexively accountable in pre-
cisely these ways, feeling pride or shame (if not “indebtedness” or “guilt”)
as a function of whether or not they did, or at least tried to do, the right
thing. For these reasons, agents that can internalize standards, and/​or
the character judgments such standards license, are often understood to
be particularly agentive kinds of agents. Conversely, and somewhat pes-
simistically, many such agents are understood to have less agency than
they might otherwise have. In a critical tradition most forcibly articu-
lated by Marx, while such agents may freely choose among a range of
possibilities (or relative to a standard), they did not get to choose the
range of possibilities (nor the standard). To return to our notion of the
unconscious, we often seem to have little agency over the wellsprings of
our agency.
Crucially, just like Aristotelian agents, many such evaluative agents are
inherently distributed: while it may be a single human actor who under-
takes an action, the choice of which action to undertake, and the reason-
ing underlying why that’s the best choice, is usually done in interaction
with others (you and me in this conversation), by reference to shared val-
ues (we in this community, with some shared history), and more or less
implicitly or unconsciously. Such values and, in particular, disputes over
such values, often lead to collective actions in which a collective “we” not
only determines what to do, and why it should be done, but also actually
does it (however hierarchical, and often unfair, the distributed divisioning
of this determination). And so, at issue are not just collective behaviors like
working and playing, communing and communicating, but also wars and
revolutions, invasions and insurgencies.

[ 20 ]  Agency as Flexible and Accountable Causality

  21

TELEONOMIC AND TELEOMATIC AGENCIES

The foregoing tradition tends to focus on a certain class of final causes: those
that turn on the relatively self-​conscious strivings of intentional agents,
and hence teleological processes of a stereotypic sort. Just as interesting
are what philosophers like Mayr (1992) calls teleonomic agents: instincts,
traits, organs, and automata. For present purposes, teleonomic agents are
adaptive agents: they not only exhibit means-​ends behaviors (or character-
istics), but they were also selected, and thus shaped, for the sake of those
behaviors. Such agents that are selected (but cannot themselves select, or
“choose,” in the strong sense just described) are also considered derivative
agents—​insofar as their key characteristics are understood as having been
imposed by another agent (and not just by seemingly intentional agents
such as artists and engineers, but also by seemingly unintentional agents
like natural selection). Mayr thought that such teleonomic processes were
guided by an internal program (such as DNA, an algorithm, or some kind
of intricate engineering mechanism). And he characterized such programs
as being more or less “open” insofar as they were more or less sensitive to
contextual inputs. In this tradition, universal Turing machines are radi-
cally open (in that they can run any program you give them); and mem-
bers of the species homo sapiens are radically open (in that they have the
capacity to inhabit any culture they inherit, with each such culture having
distinctive suites of means and ends, distinctive modes of knowledge and
power, and distinctive causal capacities and evaluative rationalities). The
relative “openness” of such agents is, of course, yet another way of imag-
ining their relative agentiveness. And again, such an agentiveness is not
necessarily agentive in any stereotypic sense insofar as it might have been
“programmed” by an external agent—​your parents or teachers, your cul-
ture or society, a particular religion or ideology, and so forth. (Recall the
admonishments of Marx.)
This sense of agency usefully points to another sense of accountability.
All of the products of natural selection can be held accountable in the first
sense: we can account, if only partially, for their coming to be. The world
itself, in the sense of an “environment,” can hold organisms accountable
for their adaptations—​selecting the most fit from the least fit, and thereby
changing the adaptiveness of a population of organisms over time. And a
few such organisms, which like to think of themselves as being on the top
of a ladder, are uniquely accountable in the moral sense (or so they tell
themselves, in their self-​accounts). Of course, many try to offer evolution-
ary accounts (and, perhaps more often, non-​evolutionary accounts, often
by reference to some kind of god, qua meta-​meta-​agent) of the origins of

G n o m i c   Ag e n c y   [ 21 ]
22

that seemingly unique kind of moral accountability. But, that said, many
will argue that teleological processes (that undergird moral accountability)
are themselves the products of teleomatic processes (which are distinctly
non-╉intentional), which are themselves the products of non-╉telic processes
of various kinds (such as sieving and serendipity). In other words, it is not
clear, once various forms of selection have been properly understood (from
selecting among different values, qua choice, to natural selection of dif-
ferent variants, qua evolution), that all agents are not derivative agents
somehow—╉so far as their own agency is causally distributed in such ways.
There seems to be an enormous contingency underlying the complex,
non-╉intentional causalities that lead to reflexively-╉instrumental, meta-╉
evaluational, and self-╉accounting modes of agency.

AGENCY ENCLOSED

Given all these different ways of framing agency, and thereby figuring
different kinds of agents, we may ask why human agents are so prone to
frame and figure in particular ways. Why do they ascribe agency along
specific dimensions—╉the causal, the flexible, the reflexive, the evaluative,
the teleological, the moral? Why do they order (along such dimensions)
by reference to particular degrees—╉more or less causal, flexible, reflexive,
evaluative, teleological, moral, and so forth? Why is it so easy to excerpt
agentive figures from multi-╉causal backgrounds? Look, there is an agent,
qua figure (all else becoming background, and so of “no account”). Why do
they excerpt and enclose particular units—╉drawing figures within particu-
lar boundaries: skin, self, team, society, and so forth?
Different collectivities, with different histories and cultures, frame and
figure in different ways; and, indeed, are framed and figured in different
ways by themselves and others. To some degree, a large chunk of politics
turns on who gets to determine the relevant dimensions and degrees of
agency, what causes and capacities are considered important, what order-
ings are appropriate, what units can be excerpted, where to draw boundar-
ies between figures and grounds. In some sense, the most consequential
forms of agency reside in who or what determines what counts as agency,
and thus who or what should be held accountable as an agent.
When we stress these kinds of inherently reflexive, political, and gene-
alogical concerns alongside the foregoing and more canonical kinds of
questions and commitments regarding teleonomic processes, and their
conditions and consequences, we are opening up inquiry into what might
best be called “teleognomic agency,” or simply gnomic agency.

[â•›22â•›]â•… Agency as Flexible and Accountable Causality

╇ 23

REFERENCES

Aristotle. 2001. Physics. The Collected Works of Aristotle, edited by Richard McKeon,
214–╉394. New York: The Modern Library.
Bacon, Francis. 2000. The New Organon. Cambridge: Cambridge University Press.
James, William 1918. The Principles of Psychology. Vol. 1. New York: Dover.
Mayr, Ernest W. 1992. “The Idea of Teleology.” Journal of the History of Ideas
53: 117–╉135.
Taylor, Charles. 1989. Sources of the Self. Cambridge, MA: Harvard University Press.

G n o m i c   Ag e n c y â•… [â•›23â•›]
  25

CHAPTER 4

Semiotic Agency
PAUL KOCKELMAN

S emiotic processes are usually understood to have three key compo-

nents: a sign stands for an object and gives rise to an interpretant. For
example, you point (a sign), to something (an object), and I turn to look (an
interpretant). In what follows, we will be interested in the semiotic agents
that seem to lie at the center of such processes: those entities that do not
just signify (by expressing signs of objects) and interpret (by expressing
interpretants of signs), but also get signified and interpreted (and thus
constitute an object of their own and others’ semiotic processes). See
Figure 4.1. That is, such agents not only have the capacity to “point” and
“look” (broadly construed), they also have the capacity to be “pointed to”
and “looked at” by other such agents (which may even include themselves,
at some degree of remove).
For a semiotic agent to be an “agent” requires that it is also a semiotic
object (from the standpoint of another such agent). In what follows we will
look at several different modes of semiotic agency (and hence several dif-
ferent modes of semiotic objectivity). Rather than seeing semiotic agents
as simple effects of, or deep conditions for, semiotic processes (such as a
Foucauldian subject or a Cartesian ego), we will see them as reflexively con-
stituted creatures, and hence causally concomitant with such processes.
Just as you can’t understand such agents without reference to semiotic
processes, semiotic processes cannot be understood without reference to
such agents.
26

S I

Figure 4.1╇ Semiotic Processes and Semiotic Agents.

AGENCY THROUGH THE LENS OF SENSATION

AND INSTIGATION, SELECTION AND SIGNIFICANCE

Vervet monkeys are semiotic agents. Upon sensing a predator, one such
monkey can instigate an alarm call. And upon sensing the call, another
such monkey can instigate an escape. Here there are at least two semiotic
agents engaged in overlapping semiotic processes. See Figure 4.2. For the
first semiotic agent (A1), the object (O1) is the predator (say, an eagle). The
sign (S1) is an iconic index of that eagle (say, a characteristic wing shape
silhouetted against the sky). And the interpretant (I1) is the alarm call. For
the second semiotic agent (A2), the object (O2) is still the eagle, but now
at some degree of indexical remove. The sign (S2) is the alarm call. And the
interpretant (I2) is the action of running into the underbrush. Instead of
one agent simply seeing and running; we have two coupled agents, the first
seeing and calling, and the second hearing and running.
Crucially, we may also consider the object (qua predator) to be a rela-
tively derivative agent (A3)—╉one who is not just seen and pointed to (by
the first agent), but also looked at (or “heard”) and run from (by the second
agent). Such a predatory agent can, of course, also sense and instigate in
its own right. Indeed, its ability to see and swoop is one of the key reasons
it constitutes such an important object in the semiotic processes of the
vervet monkeys.
There is also a fourth agent (A4) in this scenario—╉one that is inher-
ently distributed, insofar as it is not just composed of the first two agents,
but also partially created by the third agent. In particular, the first two
agents arguably constitute a genetic unit of accountability, by reference to
processes like inclusive fitness, and thus function as a kind of extended
organism.
Moreover, so much of the sensing and instigating behavior of this dis-
tributed agent (framed as an organism, qua bottom ellipse) makes sense
only in reference to the sensing and instigating behavior of the third agent

[â•›26â•›]â•… Agency as Flexible and Accountable Causality

  27

A3/O3

O1 O2

S1 I1/S2 I2
A5

A1 A2

A4

Figure 4.2  Communication Between Conspecifics.

(framed as an environment, qua top ellipse). The capacities and propensi-

ties of prey are often best understood in terms of the capacities and propen-
sities of predators. Organism and environment are, indeed, so inseparable
when functionally understood that it is tempting to call such a unit an
envorganism, and to see it as a fifth agent (A5, qua encompassing circle).
Such envorganisms, as the locus of those causal processes we must take
into account to understand natural selection, might even be considered the
originary agents of all evolutionary scenarios (notwithstanding how dis-
tributed, emergent, contingent, and confusing they might at first seem).
In short, agency is distributed not only across two cooperative agonists
(A1 and A2, together equivalent to A4), but also across such agonists and
their antagonist (A3). And it is distributed not only across such interact-
ing agents (framed as waypoints along evolutionary paths), but also across
the longue-​durée processes (A5) that give rise to such agentive precipitates
(such as natural selection).
Let us set aside this larger ensemble of agencies, and focus on the capac-
ities of a single semiotic agent, generically understood. What are its key
features? First, such an agent is capable of sensing and instigating. In par-
ticular, for something to constitute a sign, it must be able to not just stand
out in an environment (be a difference) but also be sensible to an organ-
ism. And for something to constitute an interpretant, it must be able to
not just stand up in an environment (make a difference) but also be insti-
gatable by an organism. And so we can inquire into the range of qualities

Se m i o t i c   Ag e n c y   [ 27 ]
28

and events, practices and processes, relations and interrelations, actions

and affects, that an agent can sense or instigate. What kinds of sensory
and instigatory capacities does it have? What scales is it sensitive to? How
do various kinds of media (from tools to techniques) enable it to extend
its capacities or shift its scales? Relatedly, what is the range of semiotic
objects such an agent might relate to via its signs and interpretants, its
sensations and instigations? Given its current environment, what counts
as an object for it? And given its capacities, and the affordances of poten-
tial environments, what could count as an object for it? Some agents have a
wide range of actual, or at least potential, signs, objects, and interpretants;
some agents have a relatively small range. What are the conditions for, and
consequences of, various ranges?
Second, not only do semiotic agents sense and instigate, and thereby
relate to objects via their signs and interpretants, but they are also caught
up in selection in a variety of senses. To foreground two extremes, we might
say that agents are capable of being selected, and/​or capable of selecting,
such that their instigations (interpretants) make sense in the context of
their sensations (signs), given the features of their objects and their inter-
ests as agents. As we saw in the example of vervet monkeys, agents might
have been naturally selected to produce certain interpretants (such as call-
ing and fleeing) in the context of certain signs (such as wing shapes and
calls), given the objects those signs and interpretants correlate with (such
as predators), and given their own interests as agents (say, to survive and
reproduce under conditions of predation). And, as we saw in the chapter
on gnomic agency, agents might also engage in “choice” in a stereotypic
sense: given a range of possible and desirable interpretants (of some sign,
correlated with some object, itself critical to the interests of the agent),
such an agent can select which one is the most desirable, and thus worthy
of instigating (given the agent’s evaluative standard).
To be sure, these examples barely scratch the surface of various modes
of selecting and being selected. There are agents capable of undergoing,
and undertaking, artificial selection. There are agents capable of making
and using tools, and being made into or used as tools. There are agents that
can write, and run, algorithms. And so on, and so forth. As always, we can
argue about whether such agents are more or less “originary” or “derived”
(and, indeed, whether they should be called “agents” at all). And we can
inquire into the relative “openness” of their processes of selection. This is
not so much the question of what range of signs and interpretants they
can sense and instigate, but rather the relative flexibility, or contextual
sensitivity, of their interpretant-​sign relations. What kinds of “learning”

[ 28 ]  Agency as Flexible and Accountable Causality

╇ 29

or “programming” or “enculturation” or “evolution” are they capable of, or

amenable to? What complex adaptations can they (hope to) achieve? What
feats of strength and imagination, calculation and communication, sympa-
thy and sociality, caring and killing, are they capable of? To what degree are
they capable of distributing, and thereby concatenating and concentrating,
their individual agencies? To what degree can they select options that will
increase the range of their possible selections?

AGENCY THROUGH THE LENS OF ONTOLOGIES

IN TRANSFORMATION

So much for semiotic agents understood through the intersection of sens-

ing and instigating, on the one hand, and significance and selection, on the
other. Now let us turn to a relatively overlapping topic, agency through the
lens of ontologies in transformation. Table 4.1 characterizes agents in rela-
tion to indices, individuals, kinds, and ontologies. The key issue here is the
way that ontologies contribute to interpretations. Table 4.2 characterizes
several ways that the ontologies of agents might transform over time. The
key issue here is the way that interpretations contribute to ontologies.
Suppose, for example, that you are the individual and I am the agent.
I watch you engage in some kind of performance, or trial of strength (an
index), and infer that you have some kind of competence, or power (a kind).
The performance might be a piano recital, a dissertation defense, a con-
versation in German, a chemical assay, a race, whatever. I could therefore
expect other performances from you, as complex indices, that would be
keeping with that kind—╉for example, success at another trials of strength
of a similar nature (e.g., if you played that sonata, I bet you can also play
this sonata).

Table 4.1 ╇ ONTOLOGIES IN TRANSFORMATION

Index Any quality that is relatively perceivable (to some agent).

Kind Any projected propensity to exhibit particular indices.
Agent Any entity that can perceive such an index and project such a kind (itself often an
individual).
Individual Any entity that can evince indices (to an agent) and thereby be a site to project
kindedness (by that agent).
Ontology The assumptions an agent has as to the indices, kinds, and individuals that
constitute a particular world.

Se m i o t i c   Ag e n c y â•… [â•›29â•›]
30

Table 4.2   TRANSFORMATIONS IN ONTOLOGIES

1. Indices (and signs more generally) may change an individual’s kind irrespective of an agent’s
ontological assumptions.
2. Indices may change an agent’s ontological assumptions regarding the kinds that constitute a
particular individual.
3. Indices may change an agent’s ontological assumptions regarding the indices that constitute
a particular kind.
4. Indices may change an agent’s ontological assumptions regarding the indices, individuals,
kinds, and agents that constitute a particular world.
5. Changes in an agent’s ontological assumptions about a world (in the foregoing ways) may
change the world about which the agent makes assumptions.

This would be a simple case of transformativity number 2: having seen a

connection between an index and an individual, an agent connects an indi-
vidual to a kind (given a set of ontological assumptions in regard to the
behavior of various kinds). Other kinds of transformativity are also possi-
ble. For example, in the case of transformativity number 3, the agent might
hold fast to its assumption that the individual is of a certain kind and, upon
perceiving it behave a certain way, update its understanding of the underly-
ing propensities of such kinds (e.g., I guess babysitters don’t always behave
as I thought). Inductive reasoning is of this type. In the case of transfor-
mativity number 4, the agent might hypothesize a new kind (as a way to
make sense of anomalous behavior), or start tracking indices it had never
noticed before (perhaps because of the advent of a new medium, such as
a microscope), or attend to a novel individual (whose behavioral patterns
seem newly relevant and potentially predictable). Or, in the case of transfor-
mativity number 5, the individual (itself an agent in another frame) might
internalize the assumptions of the perceiving agent and change its behavior
accordingly—​say, a patient trying, however unconsciously, to fit (or thwart)
a doctor’s diagnosis. (Indeed, an illness, or diagnostic category, is a quint-
essential kind.) Lastly, there is transformativity number 1, the fact that
individuals are caught up in multiple causal processes that help create them
and their kindedness, more or less irrespective of the ontological assump-
tions of particular agents: chemical reactions (creating kinds like “acid” and
“base”), evolutionary processes (creating kinds like “dog” and “raccoon”),
socialization processes (creating kinds like “punk” and “mod”), and so forth.
From the standpoint of such an analytic framework, an agent has many
basic capacities. It can perceive indices and project kinds. It can hold (or at
least exhibit) ontological assumptions (which enable it to project as a func-
tion of what it perceives). And these assumptions can themselves transform

[ 30 ]  Agency as Flexible and Accountable Causality

╇ 31

over time. For example, such agents can be more or less sensitive to the fact
that their assumptions are in error and thereby come to update them.
For any such agent, we can thereby inquire into the range of indices it
can perceive, the richness of the kinds it can entertain, the diversity of
the individuals onto which it can project such kinds, and the number of
different worlds it can imagine. Concomitantly, we may inquire into the
complexity of the inferences such an agent can engage in, the ease with
which it may update its ontological assumptions, the degree to which it
can detect and correct errors in its assumptions, and the extent to which
its assumptions can change the world. And we can inquire into its access to
forms of media that extend such capacities (or buffer itself from the effects
or limits on such capacities).
Crucially, such an agent can itself be an individual that exhibits indi-
ces that are perceivable to other agents (including itself); and these other
agents can project kinds onto it, and thereby come to interact with it in
particular ways. That is, such an agent is not just a source of perception
and projection; it is also a site of perception and projection. And just as
its own ontological assumptions about the world can transform (through
its indexical encounters with various individuals), the ontological assump-
tions that other individuals have about it can transform (via their indexical
encounters with it). And, indeed, as per transformativity number 5, such
an agent might even be able to internalize the ontological assumptions
that other agents have about it, and thereby come to behave according to
their beliefs about its various kinds. Note, then, how radically “distributed”
such agents can be.
Finally, this whole framework easily scales to a meta-╉level, for we have
just described a particularly important kind in our own ontology—╉the
agent. And we can project such a kind (or power) onto various individuals
as a function of the indices they express (as the evincing and exercising of
that power). Such agents are a particular kind of kind: one that can per-
ceive and project in ways that conform to ontologies and transform ontolo-
gies. In short, and to return to our opening concerns, such agents not only
ontologize entities in the world, but they are also ontologized as entities
in the world.

AGENCY THROUGH THE LENS OF RESIDENCE IN,

AND REPRESENTATIONS OF, THE WORLD

We might characterize comportment as any behavior that involves heeding

affordances, wielding instruments, undertaking actions, performing roles,

Se m i o t i c   Ag e n c y â•… [â•›31â•›]
32

or fulfilling identities. And we might characterize one kind of agent as any

entity (process, event, actor, assemblage, etc.) that enables and constrains
one’s comportment insofar as it makes a difference in regard to: (1) what one
does and how one does it; (2) where and when one engages in such a doing;
and (3) why one does it, and what effects arise because one does it. From
one perspective, an agent is whatever alters the contents (what and how),
contexts (where and when), and consequences (why and to what effect) of
one’s comportment—​even if only retroactively. From another perspective,
an agent is whatever has some measure of control (where and when), compo-
sition (what and how), and commitment (why and to what effect) in regard
to its own comportment. See Table 4.3. And just as the latter kind of agent
need not be an “individual,” the former kind of agent includes imaginaries
and institutions as much as interactions and infrastructure. In this framing,
agents are not just whatever “has” some degree of control, composition and
commitment (semiotically speaking), but also whatever enables and con-
strains the contents, contexts, and consequences of such “havelings.”
Such would be a way of understanding agency as a mode of residence
in the world. We would also do well to understand agency as a mode of
representing the world (itself always already built on ways of residing in
the world). Such agencies enable and constrain the contents of our propo-
sitions: not just what we are talking or thinking about (as topics, or refer-
ents), but also what we are saying or thinking about such topics (as foci,
or predicates). And such agencies enable and constrain our reasoning with
such propositions: not just the forms of evidence and inference that lead to
such propositions (as roots), but also the forms of evidence and inference
that flow from such propositions (as fruits). From one perspective, a repre-
sentational agent is whatever has some capacity to topicalize, characterize,
and reason. From another perspective, a representational agent is whatever
enables and constrains the contents of our propositions, the conclusions
we draw from them, or the conclusions that drew us to them. See Table 4.4.
And again, just as the former kind of agent need not be an “individual,” the
latter kind of agent includes imaginaries and institutions as much as inter-
actions and infrastructure. In short, both residential and representational

Table 4.3  RESIDENTIAL AGENCY

Control the contexts (where and when) of one’s behavior

Compose the contents (what and how) of one’s behavior
Anticipate the consequences (why and to what effect) of one’s
behavior

[ 32 ]  Agency as Flexible and Accountable Causality

╇ 33

Table 4.4 ╇REPRESENTATIONAL AGENCY

Determine Topics (of Representations)

Determine Foci (that Apply to such Topics)
Determine Arguments (that Lead to, or Follow from, such Topic-╉Focus
Relations)

modes of agency are inherently multidimensional, graded, embedded, and

distributed phenomena.

AGENCY THROUGH THE LENS OF CAUSAL PROCESSES

As should be clear from both the foregoing perspectives, agents stand at

the center of causal processes—╉as targets as much as sources. Such causal
processes can, for present purposes, be understood in relatively simple and
stereotypic terms: one event (E1) leads to another event (E2) in the context
of a field of forces. See Figure 4.3a. How such force fields condition such
event sequencing is not of concern here (and, indeed, would take most of
human knowledge to explicate fully). Rather, I  am interested in drawing
out the repercussions of such processes for our understanding of semiotic
agents (insofar as such agents “understand” such processes).
Suppose, for example, that an agent is more or less aware of the causal
relation between two such events;1 and suppose that an agent is more or
less able to sense and/╉or instigate such events. In regard to instrumental
agency, such an agent might instigate E1 as a means to bring about E2
as an end. In regard to inferential agency, such an agent might predict E2
(having sensed E1), or retrodict E1 (having sensed E2). In other words, to
such an agent, E1 and E2 relate not just as cause and effect, but also—╉
at least potentially—╉as means to ends, sign of object, and object of sign.
Instrumental and inferential practices don’t just constitute causal pro-
cesses in their own right, they also depend on them to function and, in
particular, to malfunction.
Such causal processes (and hence such instrumental and inferential
practices, and hence the semiotic agencies themselves), are subject to vari-
ous reframings. First, any particular event is caught up in a myriad of force
fields, and so is (partially) causal of many other events, and (partially)
caused by many other events. See Figure 4.3b. Moreover, any causal pro-
cess may be reframed as one link in a longer causal process; or as a longer
causal process that is made up of many links, each of which is a smaller

Se m i o t i c   Ag e n c y â•… [â•›33â•›]
34

(a)
C E

E1 C1
(b) E2 C2
C E
E3 C3

(c)
C1 E1/C2 E2/C3 E3/C4

E2

(d) (where effect of C1 is C2-E2 relation)

C1

C2

Figure 4.3  Embedding and Enchaining of Causal Processes.

causal process. See Figure  4.3c. Which specific events, force fields, and
scales an agent attends to are, in part, a function of what events it can
sense and instigate, and what force fields it is aware of. And they are, in
part, a function of what it is currently engaged in—​either instrumentally
or inferentially.
Second, just as semiotic processes can incorporate causal processes
(i.e., a sign-​object relation, in Figure 4.1, can be a cause-​effect rela-
tion), causal processes can incorporate semiotic processes (i.e., a sign-​
interpretant relation, in Figure 4.1, can be a cause-​effect relation). As
an example of the first case, I  can infer fire from smoke (and thereby
use a causal process to engage in a semiotic practice). As an example of
the second case, I can raise my hand in order to get you to answer my
question (and thereby use a semiotic process to engage in a causal prac-
tice). Indeed, the key objects of many semiotic processes are precisely
causal processes:  many semiotic agents can signify and interpret such
processes; and thereby communicate, collaborate, and compete in regard
to such processes; and thereby help or hinder others in their ability to
direct or discover such processes.
Finally, a particularly important kind of effect (E2) is the setting up,
removing, or rechanneling of a force field that links two other events (E3
and E4). See Figure 4.3d. In particular, an agent that instigates E1 in order
to cause E2 may thereby ultimately govern the instrumental and inferen-
tial processes of other agents (who are caught up in E3 and E4). Causing

[ 34 ]  Agency as Flexible and Accountable Causality

╇ 35

causality is closely linked to conducting conduct, and hence a key mode of

power or governance.
In short, to know about the causal, inferential, and instrumental pro-
cesses that an agent is attentive to (and/╉or subject to), is to know much of
what there is to know about that agent. Agents not only discover and direct
such processes, but their own processes can be discovered and directed.
With such points in mind, we can now ask a series of questions
about particular semiotic agents. What is the range of causal processes
they can sense and instigate (and hence direct and discover)? How do
such causal processes allow such agents to have agency over larger or
smaller swatches of space-╉time (insofar as E1 and E2 are spatiotempo-
rally “near” or “far” from each other). In other words, to what degree
can such agents both act at a distance and with precision? Which such
causal processes are relatively portable, insofar as the relevant force
fields can be expected to hold wherever (and whenever) such agents go?
To what degree can agents make them more portable—╉by creating an
infrastructure, or built environment, in which such force field can be
made relatively reliably present? To what degree can agents make causal
chains relatively reflexive insofar as the effects they induce act back on
them as causes? To what degree can agents discover, and perhaps come
to direct, the causal processes that created them as effects (for exam-
ple, converting natural selection into artificial selection)? In short, or
perhaps more generally, to what degree are semiotic agents, through
their causal processes, and hence in regard to their instrumental and
inferential practices, relatively flexible—╉in the sense of self-╉reflective,
framing, governing, displaceable, precise, portable, self-╉reflexive, and
so forth?

SPHERES OF INFLUENCE

We may often usefully distinguish between an agent’s relatively immedi-

ate and relatively mediate spheres of causal (inferential and instrumental)
influence. The immediate sphere (or network, or infrastructure, or context)
consists of the range of events they can (more or less) directly sense or
instigate. The mediate sphere consists of the range of events they can sense
or instigate only indirectly, by tapping into various force fields, and the
causal processes these condition and enable. In other words, the mediate
sphere consists of all the events an agent can instigate or sense only by way
of the causal processes it is caught up in, and attentive to; and hence what

Se m i o t i c   Ag e n c y â•… [â•›35â•›]
36

they can instrumentally instigate, or inferentially know, at one (or more)

degrees of causal remove.
Crucially, the distinction between immediate and mediate spheres,
like the notion of causality more generally, is relatively frame-╉dependent,
requiring that we delimit where the agent ends and the environment
begins, or where the agent’s relatively inalienable media ends (e.g., its eyes
and hands) and its relatively alienable media begins (i.e., its telescopes
and rocket launchers). In this way, it consists of a flexible and contestable
boundary, which may shift whenever new media, new environments, new
forms of knowledge, new techniques, new technologies, and so forth come
into play. Indeed, to tie together two earlier topics, it may even (and per-
haps more often) shift as a function of new ontological assumptions as to
where an organism ends and its environment begins, or which capacities
are essential to an organism versus ancillary, or which force fields in an
environment are crucial versus contingent.
A key limit on the agent is thereby delimited: where its effects (qua insti-
gations) would end without one or more channels versus where its effects
can end given such force fields (which take those instigations as causes, and
help generate further effects). In some sense, agency is radically distrib-
uted for the simple reason that without all those force fields, or channels,
an agent is relatively provincial. Its sphere (or rather network) of influence,
and hence the agent itself, is only as small, or as large, as the causal pro-
cesses it can direct and discover.

ACCOUNTABILITY

Those events that an agent can sense and instigate, instrumentally direct
or inferentially discover, via the causal processes it is caught up in, do
not—╉by themselves—╉delimit the agent, or even its sphere of influence.
This is because a significant chunk of agency rests in accountability: of all
the effects an agent can have in the world, given all the causes it is caught
up in, only some are directly attributed to it (by other agencies), such that
it can be held accountable for them. We are punished or rewarded only
for particular events and through particular events, even if we are causally
(inferentially, instrumentally) entangled in a much wider range of pro-
cesses. And so such particularly consequential events loom large in and for
our ontology. That is, such events (for which we are accountable) are not
just key objects in our ontologies (as part of their contents, so to speak);
they are also a key condition of possibility for our ontologies (whatever
their particular contents).

[â•›36â•›]â•… Agency as Flexible and Accountable Causality

╇ 37

A wide range of agents can internalize such consequentiality. That is,

they can take into account the way their inferences and instigations have
effects back on them, for better or worse, through the regimenting agents
that hold them accountable; and they can thereby act (infer and/╉or insti-
gate) in ways that channel such consequentiality:  stopping it, redirect-
ing it, modulating it, minimizing it (and sometimes even fostering it).
This ability of an organism to internalize the consequences of its own
actions, by means of such regimenting agencies, is one key part of our
understanding of self-╉consciousness. But that said, such agencies need
not be intentional:  we can internalize the sharpness of knifes and the
swiftness of predators as much as the gaze of states and the attitudes
of parents. Indeed, not only our learned behavior, on developmental
timescales, turns on internalization as such; most of our innate behav-
ior also turns on the internalization of the reliable causal pathways that
constitute our environments on evolutionary timescales. Every organ-
ism’s body/╉mind/╉habitus/╉imaginary is a diagram of the (salient, reliable,
pervasive, shocking) force fields of its environment. (That said, what we
have internalized (from past environments) may be out of skew with the
current environment—╉so we are not necessarily well fitted to where we
find ourselves.)
With this understanding of causality and semiosis in hand, we may now
highlight various senses of accountability. Semiotic agents not only count
on (or take into account) causal processes, but they also make them count,
both inferentially and instrumentally. Some can, to some degree, offer
accounts of such causal processes: they can point to them, predicate prop-
erties of them, and reason about them. Some can, to some degree, control
the when and where of causal unfoldings (qua E1), the what and how (qua
field of forces), and the why and to what effect (qua E2). Reciprocally, semi-
otic agents are themselves the source and target of semiotic processes, so
other semiotic agents can count on, and offer accounts of, them. Finally,
as a function of all of this, and as just laid out, agents can be held account-
able for the causal processes they are caught up in (by other agents)—╉an
accounting that, to some degree, constitutes the agent.

NOTE

1. Not necessarily aware in the sense of “conscious of,” but rather in the sense of
“has ontological assumptions regarding.” For more on each of these accounts
of agency, see Kockelman (2012, 2014a, 2014b, 2015). For more on vervet
monkeys, see Cheney and Seyfarth (1990). See Gell (1992) for a different take on
spheres of influence.

Se m i o t i c   Ag e n c y â•… [â•›37â•›]
38

REFERENCES

Cheney, Dorothy L., and Robert M. Seyfarth. 1990. How Monkeys See the World.
Chicago: University of Chicago Press.
Gell, Alfred. 1992. Art and Agency. Oxford: Oxford University Press.
Kockelman, Paul. 2012. “The Ground, The Ground, The Ground: Why Archaeology Is
So ‘Hard.’” Yearbook of Comparative Literature 58: 176–​183.
Kockelman, Paul. 2014a. Agent, Person, Subject, Self. Oxford: Oxford University Press.
Kockelman, Paul. 2014b. “The Anthropology of an Equation: Sieving Spam,
Algorithmic Agents, and Ontologies in Transformation.” Hau: Journal of
Ethnographic Theory 3 (3):33–​61.
Kockelman, Paul. 2015. “Four Theories of Things: Aristotle, Marx, Heidegger, and
Peirce.” Signs in Society 3 (1):153–​192.

[ 38 ]  Agency as Flexible and Accountable Causality

  39

PART TWO

Agency of Institutions and Infrastructure

  41

CHAPTER 5

Agency in State Agencies

ANYA BERNSTEIN

T he democratic state is an administrative state. It is often portrayed,

for convenience, as a legislature, a court, or a chief executive. But the
actual work of representative governance is done primarily in administra-
tive agencies, which interpret and implement the often vague ambitions
inscribed in statutes. Agencies translate legislative pronouncements into
rules that set standards and regulate conduct; investigate and ensure com-
pliance; adjudicate disputes and prosecute violators.1 Agencies, in short,
both interpret and implement the laws that legislatures pass. Whether we
are interested in consolidated sovereignty or diffuse pathways of power, it
is administrative agencies that enact much of the state’s effect in the world.
When we talk about agency in the state, then, we must primarily be talking
about agency in agencies.
That may seem odd. Bureaucracy seems like the absence of agency: just
mechanistic gear-​grinding continuing things begun by other, distant, pow-
erful actors. Where can agency find a foothold amid the faceless people, the
featureless buildings, the infinite red tape, the endless unread files? Few
have captured this un-​agency more grippingly than Hannah Arendt (1994),
who shows in consternating detail how bureaucracy melts individuals into
a mass, subordinating them to an unstoppable process set in motion by
others elsewhere—​actual agents—​and negating their accountability for
even their most extreme acts. In bureaucratese, the things that happen
happen in the passive voice.
42

Here there is no room for conduct that creates, starting off the uncon-
trollable iterations of response and reinterpretation that characterizes true
political action (Arendt 1958). It would be a travesty to conflate the agonis-
tic, agentive sphere of politics, where any answer is always subject to further
contestation, with that of administration, which reduces human activity to
a striving for smooth functioning and conclusive answers (Honig 1993).
So, another oddity:  administration has world-​changing effects, but
seems bereft of agents. Infuriatingly—​ yet conveniently—​ bureaucracy
appears as an undifferentiated entity exerting power that cannot be held
to account. It is the decider’s enforcer: an implementing thug who carries
out orders but lacks any creative abilities of his own.
This common image, it turns out, distorts our understanding of both
agencies and agency. It conceals the complex distribution of possibility and
responsibility within bureaucracy, which involves individual subjectivi-
ties, interpersonal relations, and socially structured decision-​making (Blau
1963; Bernstein 2008). And it obscures the varied ways that accountability
for bureaucratic action is structured by different social arenas. What kind
of accountability is available, it turns out, depends on the position from
which one does the accounting. Here, I unpack one administrative process
to show how units of agency emerge and blend in the ongoing process of
differentiation and subsumption that characterizes bureaucratic action.
I  then explain how one particular social arena—​litigation—​provides a
scaffolding for bureaucratic accountability that, like all scaffoldings, both
enables and constrains.
Bureaucracy’s complex distribution of agency and the legal scaffoldings
for its accountability have two linked implications. One, bureaucracy may
be a paradigmatic case of distributed agency. If we conceive of agency not
as a heroic individual exerting power over others, but as an always inter-
dependent social activity, then bureaucracy’s subsumption of individual
impulses in mass results looks less like agency’s nemesis and more like its
instantiation. Rather than seeing agency in agencies as an oddity, then, we
may see in it an oddly forthright presentation of the nature of agency itself.
Two, identifying legal scaffoldings that enable and constrain bureau-
cratic accountability suggests that accountability is perspectival not only
in agencies, but in general. Whether accountability is available, and what
it looks like when it is, depends not on some underlying relation between
actor and action, but on the scaffoldings that structure the interpretation
of action in particular social arenas. Understanding accountability scaffold-
ings in particular social realms is thus central to understanding the nature
of accountability generally. Indeed, it may be more accurate not to speak of
accountability generally, but only of local tropes of accountability.

[ 42 ]  Agency of Institutions and Infrastructure

╇ 43

VARYING, AND VALUING, DISTRIBUTION

American administrative agencies are bewilderingly variegated in their

mandates, their hierarchical structures, and their operational traditions.
Among the most important, and most opaque, things agencies do is to
implement the statutes that legislatures pass. They do so largely through
rules of conduct that have legally binding force. In the United States, the
Administrative Procedure Act and the legal cases interpreting it lay out
the public face of this rulemaking power. An agency charged with imple-
menting a statute solicits public comments on a proposed rule and then
publishes a final rule, which responds to significant comments, justifies
significant choices, and sets out constraints on conduct by those under its
purview.
Beyond those constraints, the internal process is up to the people who
run the agency. One brief example can demonstrate how different social
units—╉sub-╉individual, individual, group—╉emerge and recede as agents at
different stages of a bureaucratic process. In the Environmental Protection
Agency (EPA), mid-╉level administrators may convene “workgroups” of
scientists, engineers, economists, policy analysis, lawyers, enforcement
officers, and political negotiators to develop a regulation. Colleagues from
the initiating office gather information and draft primary documents, but
workgroups use their range of expertise and perspective to analyze the evi-
dence, design the rule, and resolve conflicts.
Workgroup conveners are advised to avoid “bastards”—╉participants
who pursue goals and grind axes unrelated to the rule production,
thereby delaying or derailing the process (McGarity 1991:73). After
higher-╉level administrators allow the proposed rule to be published and
public comments have been received, individual workgroup members
respond to comments about the parts of the rule they drafted. The conve-
ner submits these responses, workgroup recommendations for the final
rule, and dissents from those recommendations to higher-╉level adminis-
trators, who resolve remaining conflicts before approving the final rule
(McGarity 1991).
Thus, what appears in public as a monologic pronouncement by an undif-
ferentiated organization emerges from ongoing social relations and itera-
tive interactions that both individuate and collectivize human action (Hull
2012). The external world that provides the evidence on which administra-
tors base their rules influences regulatory developments (Bernstein 2016).
Particular contributions are attributable to human individuals as they take
positions within the workgroup. That individuation is largely erased as the
group reaches compromise positions. But the erasure is not complete, since

A g e n c y i n S tat e Ag e n c i e s â•… [â•›43â•›]

44

unresolved conflicts move up for resolution at higher levels. The notice of

proposed rulemaking briefly presents the proposal as a univocal, undiffer-
entiated text addressed to the public. But it is promptly disaggregated again
as administrators respond to comments on the particular pieces of the rule
they drafted, and workgroup members stake out positions on emerging
conflicts. Those specificities are then gradually erased again as higher-​level
administrators resolve those conflicts, leading to the final rule—​another
seemingly univocal text.
Individual contributions thus appear, meld into a mass, and become
re-​articulated at different stages of the process in ways visible differently
to differently situated observers. For a private party, the proposed rule
appears to have been produced by the agency as a whole. Agency employ-
ees who craft the rule and respond to comments, in contrast, see their
colleagues’ individual contributions. But neither of these is the only,
or the real, agent. Each individual’s actions have effects in the world,
but only insofar as they pass through the congealing process that undif-
ferentiates them into collective action. At the same time, the collective
could not act absent each individual’s contributions, which influence
the final product. What emerges in public view, naturally, are only the
final products of internal differentiation. The image of bureaucracy as
an impenetrable morass may thus owe something to bureaucracy’s self-​
presentation in public products—​and to observers’ sometimes gullible
acceptance of it.
Within the agency, similarly, individuals are ascribed professional com-
petencies and characterological proclivities—​scientists, bastards—​yet also
treated as a collective mass—​the workgroup. Not quite the ideal arena of
Weberian interchangeability, actual bureaucracies take into account how
individuals, with their backgrounds and personalities, contribute to or
detract from collective action. Yet this also highlights the enduring inter-
dependence of bureaucratic actors, whose individual contributions are
irrelevant, uninterpretable, unimplementable without the whole that
subsumes them.
This inherent interdependence frustrates both participants and theo-
rists of bureaucracy. Yet interdependence is in the very nature, not just of
bureaucracy, but of agency itself. The idea that agencies’ agency is undif-
ferentiated masks internal complexity in much the same way as the idea
that individuals’ agency is indivisible. Bureaucracy, thus, may provide a
viable image of how agency works generally—​an image that is unusually
forthcoming about its distributed nature, building the interdependence of
action not only into how it works, but also into what it values.

[ 44 ]  Agency of Institutions and Infrastructure

╇ 45

STRUCTURING ACCOUNTABILITY

Who is responsible for agency action? It depends who’s asking, and in

what way. One very visible way American society assigns accountability
for agency action is through litigation. Litigation is not the only route to
accountability, of course. On the contrary, it shows how particular, and
peculiar, the scaffoldings for accountability in different realms may be. In
litigation, the legal status of the action at issue—╉not the proximal actor or
decision-╉maker—╉determines who may be accountable and how.
For instance, you can sue a federal employee who violates your rights
under the United States Constitution. If you win, the court assigns the
defendant responsibility for harming you and usually requires that defen-
dant to pay you money. Such “constitutional tort” lawsuits allow for one
particular kind of accountability. It is textually bounded:  constitutional
tort lawsuits can assign responsibility only for violations of rights guar-
anteed by the federal Constitution. It is backward-╉looking and restitution-
ary, assigning blame and monetary compensation for past action. And it
is individualistic. While the conduct at issue in a constitutional tort law-
suit may arise from agency policy, the litigation can target only individuals,
not agencies. Constitutional tort lawsuits enable accountability only for
individuals who personally undertake an unconstitutional act, not for the
bureaucratic structures underlying their decision to do so.
You can also sue a federal employee who harms you in some more ordi-
nary way, such as by destroying your property, under the Federal Tort
Claims Act (FTCA). The FTCA also structures a backward-╉looking and res-
titutionary accountability, providing money damages for past action. It
provides accountability only for the universe of actions the statute speci-
fies: torts for which private parties would be liable.2 But in contrast to con-
stitutional tort lawsuits, accountability under the FTCA is collective, not
individualistic: the government is the only defendant available. If you sue
an individual government employee under the FTCA, the statute requires
courts to substitute the United States as defendant.
Finally, you can challenge an agency’s rule under the Administrative
Procedure Act (APA), which allows lawsuits “seeking relief other than
money damages.” By prohibiting monetary restitution for past action, the
APA provides scaffolding for a future-╉oriented accountability:  courts do
not punish agencies for passing bad rules, but can prevent an existing rule
from having further effects. APA lawsuits hold neither individuals nor the
government as a whole to account, but the agency itself. Accountability
here is limited to acts taken in the name of, and held out to represent, the

A g e n c y i n S tat e Ag e n c i e s â•… [â•›45â•›]

46

agency. And the consequences it imposes adhere not to an individual per-

son but to a bureaucratic position. The court’s judgment addresses whoever
leads the agency, whether or not that person had anything to do with the
offending rule. Thus the APA’s accountability treats the agency as a unitary
organism that endures through time irrespective of personnel.
These three legal regimes allow for the construction of accountabilities
with different temporal orientations; different consequences; and different
units of responsibility. These differences depend neither on the closeness
of the actor to the act, nor on the act’s organizational complexity. Rather,
they depend on how the conduct at issue looks from the perspective of
each particular legal regime:  constitutional law, tort law, administrative
law. Litigation channels legally distinguishable acts into separate struc-
tures that define the kinds of legal accountability that can be constructed
for them.
This legal accountability enabled by litigation, of course, covers neither
all acts undertaken by or in agencies, nor all the forms of accountability
available for any particular act. One version of accountability, for instance,
is conspicuously missing: this regime provides no way to reward the exer-
cise of agency in agencies, only to punish or limit it. Litigation only works
in one direction. But one can hypothesize that accountability scaffoldings
will always be constructed with a particular slant. Other social arenas will
have their own ways to channel, define, and delimit the accountabilities
they enable.
Within the agency, for instance, workgroup members can construct
accountabilities that hinge not on the legal status of their acts but on other
values and forms of judgment. They may hold one another responsible for
the quality of their contributions or their positions in a conflict. Superiors,
in turn, will have different avenues for constructing accountability for
their subordinates. And so on out. Legislators may impose kinds of pub-
lic accountability based in political values removed from both legal status,
group contribution, and job description. Historians may use altogether dif-
ferent structures to attribute accountability. Who is accountable for agency
action, and how they can be held to account, depends on the position from
which that accounting is done. Different social arenas offer different scaf-
foldings for constructing accountability, giving accountability different
parameters, consequences, and meanings.
Just as the distributed nature of bureaucratic agency may illuminate
a broader characteristic of agency in general, the perspectival nature of
bureaucratic accountability may have something to say about account-
ability elsewhere. Insofar as accountability depends on an attribution of
responsibility, what it looks like may always depend on the local structures

[ 46 ]  Agency of Institutions and Infrastructure

╇ 47

that make such attributions comprehensible and effective—╉the social

conditions that render a particular attribution of responsibility felicitous
(Austin 1962). My brief sketch suggests that even the relatively tightly
clustered social realm of litigation provides different scaffoldings that
serve as conditions of possibility for, as well as limits on, legal accountabil-
ity. Such scaffoldings for accountability do not simply allow accountability
to be recognized. They allow it to be created in the first place.

NOTES

1. Recent historical work shows that this was the case even in the early
United States, generally considered a latecomer to the administrative game
(Mashaw 2012).
2. The doctrine of sovereign immunity prevents governments from being
sued without their consent. The FTCA provides this consent for the limited
circumstances it specifies.

REFERENCES

Austin, John. 1962. How to Do Things with Words. Oxford: Oxford University Press.

Arendt, Hannah. (1963) 1994. Eichmann in Jerusalem: A Report on the Banality of Evil.
New York: Penguin Books.
Arendt, Hannah. 1958. The Human Condition. Chicago: University of Chicago Press.
Bernstein, Anya. 2016. “Differentiating Deference.” Yale Journal on Regulation
33(1): 1–╉53.
Bernstein, Anya. 2008. “The Social Life of Regulation in Taipei City Hall: The
Role of Legality in the Administrative Bureaucracy.” Law and Social Inquiry
33: 925–╉954.
Blau, Peter M. (1955) 1963. The Dynamics of Bureaucracy: A Study of Interpersonal
Relations in Two Government Agencies. Chicago: University of Chicago Press.
Honig, Bonnie. 1993. Political Theory and the Displacement of Politics. Ithaca,
NY: Cornell University Press.
Hull, Matthew. 2012. Government of Paper: The Materiality of Bureaucracy in Urban
Pakistan. Berkeley: University of California Press.
Mashaw, Jerry L. 2012. Creating the Administrative Constitution: The Lost One Hundred
Years of American Administrative Law. New Haven, CT: Yale University Press.
McGarity, Thomas O. 1991. “The Internal Structure of EPA Rulemaking.” Law and
Contemporary Problems 54(4): 57–╉111.

A g e n c y i n S tat e Ag e n c i e s â•… [â•›47â•›]

  49

CHAPTER 6

Upending Infrastructure in Times

of Revolt
JULIA ELYACHAR

“We never really act alone,” Enfield and Kockelman remind us in the pref-
ace to this volume. When millions of people come to the street from their
homes and workplaces to fight for a political goal, that fact is easy to see.
But who (or what) has agency in situations that seem to be the opposite
of “acting alone”? Who (or what) has agency in a revolution forestalled?
Millions of Egyptians took to the street in 2010–​11 during the January 25
Revolution in Egypt to overthrow President Hosni Mubarak and to demand
bread, freedom, and social justice. They succeeded in forcing Mubarak to
resign, with the support of the armed forces. Debates immediately began
about whether this was really a revolution and whether it had succeeded
or failed. Egyptians called it the January 25th Revolution after Mubarak
resigned and continue to do so. Four years later President Morsi of the
Muslim Brotherhood had been elected and deposed; chief of the Egyptian
armed forces under President Morsi, General Sisi, had become president
in turn. Opinions about what had happened, how it had happened, how
events should be named, and who was to be credited—​and blamed—​for the
ensuing course of events remain heated and will for years to come. In this
brief chapter I do not try to explain these historic events, ask whether the
revolution succeeded or failed, or pass judgment on the political agenda or
methods of any party. Rather I ask: what can the January 25th Revolution
and its aftermath teach us about agency and about distributed agency in
50

particular? I  do this by thinking about agency through infrastructure in

times of mass revolt. My thinking is informed by Kockelman and Enfield’s
preface to this volume, by Kockelman’s earlier essay on agency (Kockelman
2007), and by a growing body of social science research on infrastructure
(for a review, see Larkin 2013).
Agency is often paired with structure. In the simplest of fashion, such
pairing allows us to ask, to what degree do we have agency over our own
life? Does an “internal self” make us who we are? Or is our social status—​
our place in social structure—​more decisive than a self we only imagine is
our own? As scholars, we complicate such questions. As anthropologists,
we draw on ethnographic knowledge gained through live interactions with
people in determinate settings to think these things through. Anthropology
of the Middle East has been a rich place for such debates. The Middle East
is a fuzzy geographic concept that became common only during World
War II. Many western observers thought of it as a place where individu-
als (especially women) had little agency. The January 25th Revolution was
a surprise to almost all. It showed that agency could overcome structure
and tradition after all. Things soon got more complicated. But rather than
pursue the terms of this debate, I  shift its terms to think about agency
together with infrastructure in a time of mass revolt.
Infrastructure long seemed a matter for technocrats, far from concerns
about agency or political revolt. Infrastructure usually lies in the back-
ground, becoming visible only upon breakdown (Star 1999). Economies
cannot function without infrastructure. This commonplace is known in
all kinds of economic theory. In Volume II of Capital, for example, Marx
devotes a great deal of attention to how infrastructure is central to the
circulation and realization of value. The creation and maintenance of infra-
structure is not itself directly productive of value and yet is essential to the
capitalist system of production. Nor, from the standpoint of a neoclassi-
cal theory of value, does infrastructure create price. But if you cannot link
a product to the market, then that product will spoil and become worth-
less. If you cannot link a buyer to a seller, then a market cannot function.
Infrastructure—​roads, airports, ports, and bridges—​allows producers to
realize the potential economic value of a product. Linking buyers and sell-
ers entails more than physical transportation of goods in space over roads
and railroads, as in classic accounts of the role of railroads in 19th-​century
economic expansion.
Infrastructure is a classic public good, a set of resources available to all
and the use of which does not decrease its availability to others. This clas-
sic definition of a public good in economics often needs to be fought for
in practice. Infrastructure built by colonial states in the 19th century was

[ 50 ]  Agency of Institutions and Infrastructure

  51

never available for all. Postcolonial movements held out the promise of
infrastructure as a public good for all citizens. Failures of infrastructure
in the postcolonial era, in turn, made visible—​and came to symbolize—​
failures of the postcolonial state. Privatization of infrastructure began
in postcolonial states under structural adjustment policies in the 1970s.
Infrastructure began to be privatized in wealthy countries under neoliberal
policies of the 1990s. All this makes infrastructure a terrain and object of
political contestation.
Revolts are collective, willed, and condensed events. They are the out-
come of many different kinds of social action and political intent. This is
different from the kind of hacking into infrastructure on a daily basis that
is a normal part of life in poor countries and communities. During mass
revolts, infrastructure is repurposed by groups or actively broken down.
Roads, phones, subways, and wires become channels through which people
are persuaded to act and bodies are motivated to move. Infrastructure is
also rendered visible in revolt when people are blocked from moving—​
when streets are blocked off or squares suffused with people—​or when
signals cannot be transmitted across computer or mobile phone networks.
Many early commentators on the January 25th Revolution in Egypt
gave agency to technological infrastructure. They called it a “Facebook
Revolution.” This notion went together well with Orientalist stereotypes
of the Egyptian people as endlessly and forever patient. Patience is indeed
a virtue, as the Egyptian saying goes. But the virtue of patience in popular
Egyptian culture had never stopped Egyptians from political action in the
past. Years of organizing went into making the January 25th Revolution
as well. Many unions, individuals, leaders and parties were involved. Hosni
Mubarak and his advisors also thought that the internet had agency—​if
less than their own agency. They tried to “turn off” the internet during
the revolution in an effort to stop social action. While others might have
tried in the past to throttle or shut down parts of the internet that reach a
particular polity, no government had ever been able to turn off the internet
in an entire country and to go off the digital map. Egypt came close, sup-
pressing 91 percent of internet networks and all mobile phones by January
29th (Peterson 2011).
Communications infrastructures had been nationalized in Egypt
together with banks and large industry nine years after the 1952 Free
Officers Coup. Egypt began to espouse open market policies in the mid
1970s, and free market neoliberal economics in the 1990s and 2000s.
Through all of this, Egypt kept tight control over banking and telecommu-
nications. It instituted a private sector telecommunications industry only
in the mid-​2000s, when it granted mobile phone licenses to two private

U p e n di n g I n f r a s t r u c t u r e i n T i m e s of Re volt   [ 51 ]
52

sector firms to begin operation. Mobile phone use quickly skyrocketed.

Signed contracts included a stipulation that telecom and internet service
could be shut down at the discretion of authorities in case of national
emergency. That stipulation was invoked during the revolution, but the
attempt to keep people from communicating, coordinating, and mobiliz-
ing backfired.
Egyptians quickly turned to and patched together alternative platforms
for communication. These included older platforms such as landlines. Even
fifty years after the 1952 Free Officers Coup promised a redistribution of
wealth to Egypt’s poor, and a remaking of the Egyptian economy, only the
wealthy, well connected, or foreigners had access to infrastructure such as
landlines. (Lack of landlines on the African continent—​including Egypt—​
made it prime terrain for rapid and deep mobile phone penetration.) Some
of those landlines were concentrated in the downtown heart of colonial
Egypt, where the most pitched battles of the revolution were fought.
Landlines in apartments of activists and their sympathizers became stra-
tegic channels—​sites around which people visited, gossiped, and planned.
Communicative infrastructures in Egypt were not created by Facebook.
But they were crucial in the January 25th Revolution. Efforts to turn off the
internet and mobile phones inadvertently made visible the relational and
multilayered nature of infrastructures of communicative channels. New
communications technologies are integrated moves into existing commu-
nications ecologies (Horst and Miller 2006) or communicative infrastruc-
tures (Elyachar 2010). Infrastructure always has the “smell of the public”
(Robbins 2007), but in times of revolt this is illuminated for all to see.
One image from the revolution shows a group of young men in Tahrir
Square at night, crouched over a group of mobile phones, in a tangle of
extension cords, wired into stores opened up by owners around Tahrir
Square, under spotlights patched by activists into the state grid of street-
lights, charging their mobile phones. This captures the way in which poor
people in Cairo and elsewhere patch their way into infrastructure of the
state on a daily basis (Simone 2004). This kind of slow, long-​term, appar-
ently apolitical encroachment of the poor onto land and into infrastructure
(Bayat 2009) assumes overt political meaning in times of revolt. Activists
also brought out retired technologies like boat phones and satellite com-
munications. They created pirate and alterative infrastructures (Larkin
2013, Simone 2004), patching together whatever worked to keep up con-
tact with one another and with the outside world. They tapped into peer
geek networks around the world. With mobile phones turned off, the infra-
structure that lay behind them came into view. This included other kinds of
communicative channels, those created and reproduced over time through

[ 52 ]  Agency of Institutions and Infrastructure

  53

practices of visiting, gossiping, socializing, and strategizing. This multilay-

ered infrastructure of communicative channels jumped out from the back-
ground and became part of the action itself.
In such a moment, the politics of platforms becomes clear. Platforms of
technologies such as Facebook are neither neutral nor purely technological.
Competing models of the market, socio-​technologies of market life, tech-
nological innovations, engineering dilemmas, and historical layering of
infrastructures in Egypt were all wielded together as a communicative infra-
structure and functioning platform in which Facebook operated. Platforms
themselves became a key terrain of political contestation. Thousands of
activists and hackers around the globe got involved in the Egyptian revolu-
tion via their concern about platforms and an open internet.
The Muslim Brotherhood was not important in the lead-​up to the rev-
olution. They joined in only at the last moment. The Brotherhood had
little agency in the revolution in the sense of “the creation of some effect,
object … or action” (Kockelman and Enfield, this volume, p. xii).
But when, as the only organized political party able to take the state,
their candidate was elected president, it assumed accountability for infra-
structure and the public good for which it speaks. Who had agency for
the breakdown that ensued? Many insistent answers to that question
circulated. Nothing worked. Everyday life had become more precarious.
Potholes got bigger; streets were impassible, services were collapsed,
infrastructure in even more disrepair. Strikes continued, even in the out-
skirts of the city, far to the south and the north, in offices even of the
most important oil infrastructure companies that generated so much of
the country’s wealth.
The January 25th Revolution was centered around battles on a few iconic
squares. In the year to follow, venders moved onto the street, artists into
the squares, and strikers into administrative offices, claiming space for the
expression of agency newly emergent in the experience of revolt. Chants
from the revolution still rang out:  “The people want the downfall of the
regime!” “The workers want resignation of the management!” was another
common call, such as the call I heard one day in southern Cairo, with work-
ers occupying offices an oil infrastructure firm towards the end of the
working day. Along the low brick walls of a telecom company that owned
much of the communicative infrastructures of the city splashed graffiti of
the Ultras, Egypt’s football support clubs who had taken to the streets to
support the revolution and paid a high price with their lives. A sign warn-
ing the public to stay away merged into drawings of a fist raised high, the
image of a martyred young revolutionary, and the phrase “Long live the
revolution.” That uplifted fist appeared again in graffiti from the Ultras.

U p e n di n g I n f r a s t r u c t u r e i n T i m e s of Re volt   [ 53 ]
54

That single upended fist is an iconic symbol of resistance. It points up in

the air, rather than down toward the ground like a fist wielding tools. This
work of upending infrastructure is another, if very different, way of mak-
ing visible that which was “infra,” down below, stand upright for all to see.
That which had been infrastructure becomes another channel along which
messages can be run and claims made.
Mass revolts do more than voice resistance to an all-​powerful will. They
do more than take the state—​and then maybe lose it. Mass revolts upend
infrastructure and make clear the distributed nature of agency in its differ-
ent aspects. They leave dramatic traces of collective and often intentional
causalities. They create channels for new claims to accountability: who, in
fact, did what? Why? To what effect? Who can offer an account of what
happened? Which account will hold? Who will be held accountable? Who
can be counted on, furthermore, and whose voice counts in the process
to come?
Massive revolts upend infrastructure, creating new channels for the con-
veying of meaning, for dreaming of futures and reclaiming the past. They
stand up, as a standard to be reached and a flagpost to be taken. Further
kinds of agency—​of names to be given, claims to be made, and futures
to be dreamed—​conjoin around them. To say, then, that the revolution
failed—​or that it succeeded—​is both too easy and too wrong. We cannot
find the one place where real agency lies: with the people, with the mili-
tary, or with Facebook. Agency—​in both its components of flexibility and
accountability—​is always distributed, as we can see so well in the January
25th Revolution and its aftermath.

REFERENCES

Bayat, Asef. 2009. Life as Politics: How Ordinary People Change the Middle East.
Stanford, CA: Stanford University Press.
Elyachar, Julia. 2010. “Phatic Labor, Infrastructure, and the Question of
Empowerment in Cairo.” American Ethnologist 37(3): 452–​464.
Horst, Heather, and Daniel Miller. 2006. The Cell Phone: An Anthropology of
Communication. Oxford, UK: Berg.
Kockelman, Paul. 2007. “Agency: The Relation Between Meaning, Power, and
Knowledge.” Current Anthropology 48(3): 375–​401.
Larkin, Brian. 2013. “The Politics and Poetics of Infrastructure.” Annual Review of
Anthropology 42: 327–​343.
Peterson, Mark Allen. 2011. “Egypt’s Media Ecology in a Time of Revolt.” Arab
Media and Society 14 (Summer). Accessed August 10, 2016. http://​www.
arabmediasociety.com/​?article=770.

[ 54 ]  Agency of Institutions and Infrastructure

  55

Robbins, Bruce. 2007. “The Smell of Infrastructure: Notes toward an Archive.”

Boundary 2 34(1): 25–​33.
Simone, AbdouMaliq. 2004. “People as Infrastructure: Intersecting Fragments in
Johannesburg.” Public Culture 16(3): 407–​429.
Star, Susan Leigh. 1999. “The Ethnography of Infrastructure.” American Behavioral
Scientist 43(3): 377–​391.

U p e n di n g I n f r a s t r u c t u r e i n T i m e s of Re volt   [ 55 ]
  57

PART THREE

Language and Agency
╇ 59

CHAPTER 7

Brain-╉to-╉Brain Interfaces and the Role

of Language in Distributing Agency
MARK DINGEMANSE

A VISION OF THE FUTURE

In 1994, Nobel Prize–╉winning physicist Murray Gell-╉Mann wrote, “Some

day, for better or for worse … a human being could be wired directly to
an advanced computer, (not through spoken language or an interface like
a console), and by means of that computer to one or more other human
beings. Thoughts and feelings would be completely shared, with none of
the selectivity or deception that language permits” (Gell-╉Mann 1994:224).
Only two decades later, the scientific literature records the first proofs of
concept of such brain-╉to-╉brain interfaces (Grau et al. 2014; Rao et al. 2014),
touting a new era in which brains are directly connected and promising
to “revolutionize how humans communicate and collaborate” (Rao et  al.
2014:11). This chapter explores some of the assumptions about human
interaction implicit in this line of research, and juxtaposes them with a
view of language as the ultimate brain-╉to-╉brain interface:  an interactive,
selective, negotiable system that enables individuals with separate bodies
to achieve joint agency without giving up behavioral flexibility and social
accountability.
A key step on the way to brain-╉to-╉brain interfaces has been the devel-
opment of brain-╉to-╉machine interfaces: devices that can be controlled with
neural activity. The attraction of such interfaces is that they can bypass
some of the constraints of human bodies, a goal that resonates in popular
60

culture but also offers practical applications, for instance in the form of neu-
rally controlled prosthetic devices that enable paralyzed patients to regain
limb functions. A more ambitious program is to link such brain-╉to-╉machine
interfaces together to yield brain-╉to-╉brain interfaces (Nicolelis 2013). The
applications, medical or otherwise, are somewhat less clear here, but a com-
monly stated interest is to bypass verbal communication in order to avoid
its articulatory bottlenecks and expressive limitations (Rao et al. 2014) or
indeed its possibilities for selectivity and deception (Gell-╉Mann 1994).
Implicit in these ideals of sidestepping the constraints of bodies and
language is a narrow view of human agency and communication, and
a limited recognition of how material and cultural artifacts have long
extended our minds (Clark and Chalmers 1998). “Can information that
is available in the brain be transferred directly in the form of the neu-
ral code, bypassing language altogether?” ask Rao et  al. (2014:1). This
reveals a sender-╉receiver model of communication in which externaliza-
tion is seen as a mere obstacle to the goal of sharing private processes—╉
the possibility that it may provide much-╉needed filters and calibration
mechanisms is not considered. Rao et al.’s experiment consists of record-
ing brain activity in one person using electro-╉encephalography and
delivering it to another person using a transcranial magnetic stimula-
tion device positioned such that a pulse results in an involuntary upward
jerk of the hand. The other has no choice, but is controlled by brute force.
This is communication stripped of any possibilities for negotiating joint
commitment or shared intentionality (Gilbert 1989). It is akin to push-
ing someone out of the room rather than persuading him or her to leave.
Current conceptions of brain-╉ to-╉
brain interfaces thus appear to be
more about the involuntary unloading of information than about the co-╉
construction of shared understanding, and more about control by brute
force than about cooperation by consent—╉all at a loss of individual agency.
Collaboration, of course, always involves some redistribution of individual
agency to a larger social unit, creating joint agency. How is this achieved?
A brain-╉to-╉brain interface implements it physically; language does it socially.
The two need not exclude each other. A goal of this chapter is to put forward
some insights about human language that may be incorporated in the design
of ethically responsible brain-╉to-╉brain interfaces (Trimper et al. 2014).

LANGUAGE AND HUMAN AGENCY

To understand how language shapes and constrains distributed agency,

there is no better place to look than its primary ecological niche: everyday

[â•›60â•›]â•… Language and Agency

  61

conversation. The following fragment was recorded in Ghana, but its anal-
ysis draws on insights that apply to social interaction everywhere (Sacks
1992). Some family members have just finished processing some newly
harvested maize and are now preparing to play a board game. A large metal
tub holding the maize is still in the way, making it hard for others to join
the game.

Siwu (Neighbours_​3093575)
1 ((Ben is holding a board game for four; a large metal tub with
processed maize stands in the way of others joining))
2 Sesi kãrãnpɔ laa puta lò yedza ìyo katɔ̃ mmɔ.
right I’ll lift this and put it in storage there
3 Ben be tè mì puta mì yedza iyo:?
what are you going to put in storage?
4 Sesi àtita
the maize
5 Ben ta mà ba mà su àtitabi wangbe kɔ̃rɔ̃kɔ̃rɔ̃.
this maize will be picked up right away
6 Sesi ã, ma kɛlɛ̀ gu kala
ah, they’re taking it down ((to the lower end of the village))
((some intervening turns omitted))
12 Sesi ɛ̀h, nyua bò Afua.
uh, wait, Afua
13 Afua ((looks up))
14 Sesi mi mɔɛ gu mɛ̀ si bò su bò sia ngbe.
grab this with me so we put it right here
15 ((1.6 seconds of silence))
16 Sesi bò su bo yedza i kuruɛ nɛ te mi ba mia sɛ ngbe mi pɛ iraɔ̀ ni.
let us put this to the side so you can come and play this thing
17 ((Sesi & Afua lift tub together and place it one meter to the side))
18 Sesi kailɛ.
that’s good.
19 ((people sit down to play the game))

Sesi offers to put away the tub in the storage room. Ben asks for clarifica-
tion, and when Sesi makes clear he refers to the maize, Ben notes that it
will be taken elsewhere soon. This leads to a slight change of plans. Now

B r a i n - B r a i n I n t e r fac e s , La n g uag e , a n d   Ag e n c y   [ 61 ]

62

all Sesi needs to do is move the tub to the side. He recruits the help of his
sibling Afua. Once the obstacle is out of the way, people sit down to play
the game. Exchanges like this happen all over the world every day—​and it
is in this kind of context that we see some of the most fundamental roles
of language in social life.
Most importantly, language helps distribute agency. Take Sesi’s request
for help (lines 12–​18): he addresses Afua by name to secure her attention,
then produces a first version of the request. When there is no immediate
uptake, he reformulates his request, adding a reason which turns it from a
mere practical matter into a joint project that may also benefit her (“let’s
put this to the side so you can come and play this thing,” line 16). His
actions show that language provides a systematic organization of linguis-
tic resources to interactively manage the many concerns at play when one
person asks another to do something. The ultimate effect is that Sesi suc-
ceeds in recruiting Afua’s body as a “tool” for carrying out a joint action.
This is somewhat like the participants in the Rao et al. (2014) experiment,
but with the crucial difference that it is jointly negotiated rather than
achieved by brute force. Two features in particular help language distribute
agency: its selectivity and its negotiability.
The selectivity of language manifests itself in two ways. First, speakers
can select the linguistic means for making their ideas public, allowing them
to foreground or background certain informational or relational concerns.
To this end, language offers a large and renewable supply of words, expres-
sions, and conversational structures. Sesi can refer to the tub of maize with
“this” (relying on shared context) or with “the maize” (spelling out the under-
specified reference). He can refer to his sister as “Afua” (highlighting her sta-
tus as an individual) but also include her in the pronoun “us” (highlighting
her status as part of a joint social unit created in that very moment). This
selectivity makes language highly efficient by enabling it to be underspeci-
fied when possible, yet specifiable when necessary (Levinson 2000). A sec-
ond sense in which language is selective is that speakers can select what to
make public and what to keep private. Language provides us with ways to
control and filter what we share with whom. Perhaps it is this sense of selec-
tivity that Gell-​Mann deplored, as it opens up the possibility for deception.
However, this is easily outweighed by the social benefits: some things are
better left unsaid. More generally, sharing and withholding information are
among the most important ways in which we manage our social relation-
ships. The selectivity of public language means that it offers a set of filters
through which we can efficiently and tactfully connect private worlds.
The other key feature of how language distributes agency is its negotia-
bility: the fact that meaning and mutual understanding are always open to

[ 62 ]  Language and Agency

╇ 63

negotiation. Mutual understanding between people is built bit by bit in con-

versation. Every turn at talk is a social action as well as a take on the social
action implemented by the prior one. So Sesi’s “ah, they’re taking it down” dis-
plays one possible interpretation of Ben’s “this maize will be picked up right
away” (lines 5–╉6) in an instance of the common process of grounding (Clark
and Brennan 1991). A more explicit case of the negotiation of understand-
ing is Ben’s request for clarification and Sesi’s response to it (lines 3–╉4). This
illustrates “conversational repair,” the machinery that people use to deal with
small but frequent rifts in mutual understanding (Dingemanse et al. 2015).
Don’t such misunderstandings exemplify the kind of imperfections that
brain-╉to-╉brain interfaces aim to rid us of? To the contrary: it is precisely in
the tussles of negotiating mutual understanding that language affords a
great degree of agency. Here, for instance, conversational repair provides
Ben with a way to have a hand in subtly revising Sesi’s plans without uni-
laterally imposing another course of action. In other instances, repair can
provide ways to resist something (such as divulging personal information
or agreeing with another’s statement), empowering people to commit to
different degrees of participation without rupturing the social fabric (Sacks
1992). The interactive construction of shared understanding provides peo-
ple at every turn with opportunities for signaling consent or dissent. This
amounts to a set of checks and balances through which people can calibrate
and revise the inferences that are constantly made.
People have separate bodies. While brain-╉to-╉brain interfaces may some-
what dilute this separateness, language has long bridged it. Never merely
individuals, we are always part of a wide range of larger social units, some
fleeting and diffuse (like the 4-╉second long unit of Sesi and Afua moving
the tub out of the way, or the more dispersed unit of “readers of this essay”)
and others more strong and durable (like a close friendship or a kin rela-
tion). Language is the main tool through which we are able to navigate this
mosaic of social relations, constantly switching frames between “me” and
the many different senses of “us” (Enfield 2013). Language allows us to
maintain our individual identity and agency while merging and joining
forces at other levels.

TOWARD A MORE PRODUCTIVE VIEW OF LANGUAGE

FOR BRAIN-╉T O-╉B RAIN INTERFACES

Current conceptions of brain-╉to-╉brain interfaces strive to connect brains

directly, “bypassing language altogether” (Rao et al. 2014), hoping to bring
about the complete sharing of thoughts and feelings as Gell-╉Mann envisioned

B r a i n - B r a i n I n t e r fac e s , La n g ua g e , a n d   Ag e n c y â•… [â•›63â•›]

64

it. A recurrent dream in this work is the prospect of a “mind meld,” in which
a super-╉network of brains may come to be the substrate for something like
a meta-╉consciousness (Nicolelis 2013). From a biological perspective this is
not novel. Mind melds —╉or complex networks in which large numbers of
individual agents together achieve a form of sentience—╉have independently
arisen several times in evolution, for instance in eusocial insect societies and
in cellular slime molds, the latter capable of impressive feats in maze-╉solving.
The real challenge for brain-╉to-╉brain interfaces is not to achieve some inter-
linking of brains; it is to harness this technology in a way that does not reduce
human participants to the level of amoeba in a slime mold.
Seen in this light, the selectivity and negotiability of language are not
bugs, they are features. Thanks to these features of language, we can main-
tain a complex web of social relations by managing what we share with
whom, and we can join forces in larger social units without indefinitely
relinquishing individual agency. To throw this into sharper relief, con-
sider what happens when selectivity and negotiability are systematically
diminished or punished, as in religious indoctrination or interrogation
under torture. Here, the very defining features of individual agency are
taken away, and it is no coincidence that we describe such circumstances
as dehumanizing or inhumane. Language is what makes us human. It is
not merely a conduit for information. We might try exchanging it for a
high-╉throughput physical connection to optimize the flow of some types of
information—╉but we would do so at the tremendous cost of throwing away
a rich infrastructure for organizing social agency.
Could brain-╉to-╉brain interfaces be designed in a way that incorporates
this infrastructure? None of the points made above logically depend on the
spoken, face-╉to-╉face version of language that is its most prevalent form
today. In fact, language is modality-╉independent to some degree, as shown
by the fact that it can be realized wholly visually as in the signed languages
of the deaf. So it is not inconceivable that there can be some useful form
of social interaction in the substrate of brain-╉to-╉brain interfaces. Whatever
the modality of communication, the two design features that matter most
for creating a truly humane form of brain-╉to-╉brain interfaces are: (1) selec-
tivity, giving people control over the relation between public words and
private worlds; and (2) negotiability, giving people systematic opportuni-
ties for calibrating and revising joint commitments.

CONCLUSIONS

The Hitchhiker’s Guide to the Galaxy records the case of the Belcerebon
people of Kakrafoon, who were inflicted by a Galactic Tribunal with “that

[â•›64â•›]â•… Language and Agency

╇ 65

most cruel of social diseases, telepathy” (Adams 2002[1980]:252–╉3). It was

a punishment with unforeseen consequences: “in order to prevent them-
selves broadcasting every slightest thought that crosses their minds …
they now have to talk very loudly and continuously about the weather,
their little aches and pains, the match this afternoon, and what a noisy
place Kakrafoon has suddenly become.” Current conceptions of brain-╉to-╉
brain interfaces appear to be on their way to replicating the fate of the
Belcerebon people. In contrast, the human condition is enabled by a flex-
ible communication system that saves us from an all too unconstrained
sharing of private processes while still helping us to cooperate and achieve
shared goals in ways unmatched anywhere in the animal kingdom.
Brain-╉to-╉brain interfaces are often presented as a system for sharing
information and organizing joint action that is superior to natural lan-
guage. However, directly connecting one individual’s mental life to that
of another has the effect of robbing both of a great degree of individual
agency. There may be gains in agency elsewhere, but without the checks
and balances provided by a system like language, they will likely lie at lev-
els beyond the direct control of the individuals. Language serves as a filter
between the private and the public, and as an infrastructure for negotiating
consent and dissent. As research into brain-╉to-╉brain interfaces matures, it
is my hope that it will find ways to incorporate selectivity and negotiability,
so as to extend human agency in meaningful ways.

ACKNOWLEDGMENTS

Thanks to Nick Enfield, Steve Levinson, and Paul Kockelman for helping me
develop my thinking on the topics raised here. I received helpful comments
on draft versions from Nick Enfield and Roel Willems, and I thank other
contributors to the Foundations of Social Agency meeting at Ringberg
Castle for stimulating interactions. I  gratefully acknowledge the support
of an NWO Veni grant and of the Max Planck Society for the Advancement
of Science.

REFERENCES

Adams, Douglas. 2002. The Ultimate Hitchhiker’s Guide to the Galaxy.

New York: Del Rey.
Clark, Andy, and David Chalmers. 1998. “The Extended Mind.” Analysis 58(1): 7–╉19.
Clark, Herbert H., and Susan E. Brennan. 1991. “Grounding in Communication.”
In Perspectives on Socially Shared Cognition, edited by Lauren B. Resnick, J. M.
Levine, and S. D. Teasley, 13:127–╉149. Washington, DC: APA Books.

B r a i n - B r a i n I n t e r fac e s , La n g ua g e , a n d   Ag e n c y â•… [â•›65â•›]

66

Dingemanse, Mark, Seán G. Roberts, Julija Baranova, Joe Blythe, Paul Drew,
Simeon Floyd, Rosa S. Gisladottir, Kobin H. Kendrick, Stephen C. Levinson,
Elizabeth Manrique, Giovanni Rossi, N. J. Enfield. 2015. “Universal Principles
in the Repair of Communication Problems.” PLOS ONE 10(9): e0136100.
doi:10.1371/​journal.pone.0136100.
Enfield, N. J. 2013. Relationship Thinking: Agency, Enchrony, and Human Sociality.
Oxford: Oxford University Press.
Gell-​Mann, Murray. 1994. The Quark and the Jaguar: Adventures in the Simple and the
Complex. New York: W. H. Freeman.
Gilbert, Margaret. 1989. On Social Facts. London/​New York: Routledge.
Grau, Carles, Romuald Ginhoux, Alejandro Riera, Thanh Lam Nguyen, Hubert
Chauvat, Michel Berg, Julià L. Amengual, Alvaro Pascual-​Leone, and Giulio
Ruffini. 2014. “Conscious Brain-​to-​Brain Communication in Humans Using
Non-​Invasive Technologies.” PLOS ONE 9(8): e105225. doi:10.1371/​journal.
pone.0105225.
Levinson, Stephen C. 2000. Presumptive Meanings: The Theory of Generalized
Conversational Implicature. Cambridge, MA: MIT Press.
Nicolelis, Miguel. 2013. Beyond Boundaries: The New Neuroscience of Connecting Brains
with Machines—​and How It Will Change Our Lives. New York: Henry Holt and
Company.
Rao, Rajesh P. N., Andrea Stocco, Matthew Bryan, Devapratim Sarma, Tiffany M.
Youngquist, Joseph Wu, and Chantel S. Prat. 2014. “A Direct Brain-​to-​Brain
Interface in Humans.” PLOS ONE 9(11): e111332. doi:10.1371/​journal.
pone.0111332.
Sacks, Harvey. 1992. Lectures on Conversation. 2 vols. London: Blackwell.
Trimper, John B., Paul Root Wolpe, and Karen S. Rommelfanger. 2014. “When
‘I’ Becomes ‘We’: Ethical Implications of Emerging Brain-​to-​Brain
Interfacing Technologies.” Frontiers in Neuroengineering 7: 4. doi:10.3389/​
fneng.2014.00004.

[ 66 ]  Language and Agency

  67

CHAPTER 8

Requesting as a Means for Negotiating

Distributed Agency
SIMEON FLOYD

REQUESTING ASSISTANCE FROM OTHERS IN SOCIAL INTERACTION

People frequently achieve goals that they are unable (or sometimes unwill-
ing) to accomplish alone by requesting assistance from others. This possibil-
ity for distributing agency across multiple individuals, some expressing the
goals and others accomplishing the actions, is a basic ingredient of human
sociality. It makes possible all types of collective action, from the macro-​
level of large-​scale institutions to the micro-​level of face-​to-​face interaction.
Much of our collective action is managed by pre-​existing arrangements; for
example, nobody needs to be specifically asked to report to work on a given
day if there is already a contract establishing the required activities over a
period of time. However, at the micro level of co-​present social coordination,
it is often necessary for a person to use some explicit semiotic means, lin-
guistic or otherwise, to tell another that he or she wishes for something to
be done at a given time (see Dingemanse, ­chapter 7 in this volume). In such
cases, people make requests. A “request sequence”1 in face-​to-​face interac-
tion can be identified on the basis of two moves2 involving two participants:

Move A:  Participant A does something participant B can perceive.

      e.g., Pass the salt. ((points at salt))
Move B:  Participant B does something for participant A.
      e.g., Here you go. ((passes salt to A))
68

Practices of requesting have been studied under a number of frame-

works,3 not all of which look at them in situ to see how they work in
natural speech or, importantly, how they are responded to. This chapter
discusses the distribution of social agency in light of findings from a
study of requesting in a corpus of video recordings of everyday house-
hold and community interaction among the indigenous Chachi people
of northwest Ecuador (with a sample of 205 request sequences), part
of a cross-​linguistic project including fieldworkers working all over the
world.4 Specifically, it discusses the tension between a characterization
of requesting practices as prosocial and cooperative, in that people step
in to help one another, and a characterization of request sequences
as inherently unequal, in that one person gives the command and
another acts.5

MAKING REQUESTS

One common way that people formulate requests is by using the mor-
phosyntactic resources made available to them by the different lan-
guages they speak. While grammatical systems vary to some degree,
they usually organize the types of formulations they make possible
into “sentence types” that are useful for cross-​linguistic comparison.6
Languages tend to distinguish making a statement (“declarative”),
asking a question (“interrogative”), and making a command (“impera-
tive”). This last type is primarily dedicated to expressing a desire to
direct the actions of others, and as such it is the most frequently used
sentence type for requests in many languages (e.g., “Pass the salt”).
Speakers also use the other major sentence types for requests:  inter-
rogatives (“Can I  have the salt?”), declaratives (“I need the salt”), and
sometimes other formats including utterances with no verb and thus
no sentence type, strictly speaking (“Salt!”). Languages commonly allow
for different sentence types to be used flexibly for requesting, often
with relation to more “indirect” interrogative and declarative strategies
of requesting that may be considered more polite,7 but languages may
also vary greatly with respect to the different frequencies with which
the different sentence types are used for requests. The types of indirect
request strategies that have been observed for English are also seen in
Cha’palaa, but are relatively rare, and imperatives are the most common
format (71% of cases).

[ 68 ]  Language and Agency

  69

(1) Imperative request

A inu jayu kude aamama shipijcha M1
  i-​nu    jayu   ku-​de      aamama shipijcha
  me-​to   a.little  give-​imperative  grandma madroña.fruit
  Give me  a little  madroña grandma ((reaches toward B))
B ((turns toward A; begins passing fruit)) M2

(2) Interrogative request

A serruchu tsutyuu M1
  serruchu  tsu-​tyu-​u
 saw  lie-​ negative-​interrogative
  Is (your) saw not there?
B enku (.) tanami ibain (.) M2
 enku ta-​ na-​
mi         i-​ bain
  here have-​position-​declarative  me-​also
  Here, I have it. ((goes to get saw))

(3) Declarative request

A tu-​tu’pushujuntsaa kanu juaa M1
  tu-​   tu’pu-​shujunsta-​ya   ka-​nu  ju-​ya
  nail  nail-​that.part-​focus grab-​to be(declarative)-​focus
  It is necessary to grab the part that was nailed.
B ((grabs and moves board)) M2
Based on their predicates, the examples above represent imperative (1),
interrogative (2)  and declarative (3)  request types. Other spoken mate-
rial added to the main sentence types may include “vocatives” (“hey”),
“mitigators” (“a little salt”), “strengtheners” (“the damn salt”), accounts
or explanations (“the soup is not salty enough”), among other elements.
In addition to spoken material, people also frequently combine speech
with visual bodily practices, most frequently reaching out to receive
objects, holding out objects relevant for the request, and pointing to places

Here, put the

Give me soap away.
that.

Give me the
old razor.

Figure  8.1 Different types of multimodal request formats. Left: Holding out the hand
(16%). Middle: Holding out an object (27%). Right: Pointing (42%).

Re q u e s t i n g a s a Mea n s f or Ne g o t i at i n g   Ag e n c y   [ 69 ]
70

or things involved in the request. Under certain conditions, people are also
able to successfully accomplish requests entirely nonverbally, such as when
there is an already agreed upon joint activity.8

RESPONDING TO REQUESTS

How people formulate their requests is only half of the issue, however.
While most work on requesting has focused on request formats in iso-
lation, cues from work in conversation analysis points toward seeing
requesting as a sequence of moves in which the initial move provides an
opportunity for a subsequent, responsive move. One important aspect
of these contingent moments is that they are socially delicate, as the
addressee is in a position to potentially reject the request. Work on
politeness and “face” in interaction notes how people actively avoid and
mitigate this type of face-╉threatening act.9 Indeed, looking at the differ-
ent ways Cha’palaa speakers responded to requests, fulfilling the request
is by far the most frequent type of response while rejection is the least
frequent.
One way to interpret this finding, which is surely appropriate to some
degree, is that people share a generally cooperative stance, and so they max-
imize the potential for affiliative interactions with no “face-╉threatening”
consequences. People aim for request sequences like those in (1), (2), and
(3), above, in which the next move is B accomplishing the requested action,
and avoid sequences like (4), in which A’s request is not fulfilled.

80
70
60
50
40
30
20
10
0
Response

Fulfillment Rejection Other

Figure  8.2╇Types of responses to requests in Cha’palaa that received sequence-╉closing

responses. Fulfillment was 71%, rejection was 6%, and other actions (i.e., initiating
repair: e.g., “What?”) represented 22%.

[â•›70â•›]â•… Language and Agency

  71

(4) Rejection of a request

1A lemu tsutyuu, lemu deii M1
lemu  tsu-​tyu-​u  lemu  de-​i-​i
lime  lie-​NEG-​Q  lime CMPL-​become-​Q
Are there no limes? Are the limes gone? ((turns toward B))
2B lemu jutyu kaa ruku M2
lemu  ju- ​tyu  kaa-​ruku
lime  be-​NEG  DIM-​man
There are no limes little husband.
((reaching into basket to check))

The few rejections that do occur are often, like this one, managed through
indirect requests and mitigating, polite answers (in this case, with the
use of a diminutive), and rarely is there any overt contention. However,
there may be more behind this surface of apparent pervasive cooperative,
affiliative behavior. One of the options in Brown and Levinson’s original
formulation of how people manage face in social interaction is to avoid
the face-​threatening act in the first place, if there are potential negative
consequences.10 For the study of requesting, this means that in addition
to reflecting a cooperative stance, the low rate of overt rejection may also
reflect the fact that in most instances people only make requests that have
a high chance of success.
Not all members of a society have equal rights to request others to do
things in any given time or place, and these rights are tied to the nature
of the activity at hand, the setting, and the participants involved.11 In the
context of everyday interaction under discussion here, the different rights
and duties expected of individuals are in part structured by their household
and community roles. In the collaborative project that the investigation of
Cha’palaa requests was part of, each field researcher, based on long-​term
experience in the communities where the natural speech data was recorded,
was asked to provide ethnographically grounded information about the
social (a)symmetry of each dyad involved in a given request sequence,
when possible. These rationales for identifying such asymmetries varied.
In Chachi households in Ecuador, where men often rest in hammocks while
women attend to cooking, cleaning, and childcare, husband-​wife dyads
often show dynamics of gender inequality. Children and adolescents are
also expected to take orders from adults to do different household tasks. In
other cultural contexts, gender or age might not be as relevant in commu-
nity contexts as status as host versus guest, for example, or the birth order

Re q u e s t i n g a s a Mea n s f or Ne g o t i at i n g   Ag e n c y   [ 71 ]
72

120

100

80

60

40

20

0
(A)symmetry

A>B A<B A=B

Figure 8.3╇ Social (a)symmetry in dyads involved in request sequences in Cha’palaa. Most

requests in dyads of equal status (53%), but a further 38% are from higher-╉status to lower-╉
status individuals, compared to only 9% for lower-╉status to higher-╉status individuals.

of siblings, in places where this is held as important. When it was possible

to characterize dyads as asymmetrical based on knowledge of Chachi soci-
ety more generally, there were always more cases of a higher-╉status indi-
viduals requesting from lower-╉status individuals than vice versa.
These findings add another layer to the simple interpretation of high suc-
cess rates of requests as straightforward cooperation. One important aspect
of the apparent tendency toward cooperation is that people tend to ask oth-
ers to do things when they have more social rights to do so, and so cases of
outright rejection are rare since people may not even make socially inap-
propriate requests in the first place. This perspective leads to the apparently
paradoxical conclusion that the cooperative principles that allow individuals
to share collective agency also introduce the possibility of social asymmetry,
when some actors more frequently take the role of the command-╉giver and
others more frequently accomplish the action requested.

REQUESTING AS COOPERATIVE OR ASYMMETRICAL?

Human behavioral and evolutionary sciences have made much of prosocial,

altruistic types of practices seen in the behavior of humans and sometimes
in other social species. Some researchers debate how altruistic traits may
represent evolutionary fitness through the benefits that individuals obtain
by being parts of social collectives.12 Others focus on understanding human
capacities for intersubjective-╉type thinking, in which people consider the per-
spectives and desires of others in their own behavior.13 Such basic properties

[â•›72â•›]â•… Language and Agency

  73

of humans are no doubt significant, but research on altruism and prosocial-

ity sometimes seems to portray an overly rosy view of social relations.
In the social sciences, researchers have long attempted to understand
social inequality, focusing as much as on how people dominate each other
as on how they assist each other. For example, a situation like that in the
Chachi communities under discussion here in which members of society
perform labor and receive benefits of labor disproportionately resonates
with the type of inequalities emphasized in Marxian analysis. However,
Marx was mainly interested in broad social stratification based on eco-
nomic class and had much less to say about everyday community-​or family-​
internal inequalities.14 Weberian analysis of social stratification introduces
more varied intersecting planes beyond economic class, one of which being
the more local social status of the individuals involved.15 The types of social
asymmetries most relevant in the everyday contexts under discussion here
mainly concern people’s status in terms of age, gender, and kinship rela-
tions, although these planes interact with others at different levels of scale.
When Weber defined social domination as “the probability that a command
with a given specific content will be obeyed by a given group of persons”16
he may not have been thinking of commands so mundane as “Pass the salt,”
but these types of requests are in some senses fundamental to larger scale
issues. It is a bit of a jump to move from a man in an indigenous Chachi
community in Ecuador telling his wife to bring him salt to, say, a mining
company requesting access to ancestral Chachi territory to extract miner-
als. Yet Weber’s definition includes terms like “command” and “obey” that
seem amenable to being applied to both levels of scale.
Analysts of mundane everyday interaction have resisted the imposition
of abstract categories to micro-​interactional instances unless there is direct
evidence for them in the data, as seen in Schegloff’s classic critiques of social
scientists’ use of abstract concepts like power and social identity without
showing empirical evidence based on social interaction.17 However, given
the interactionally grounded finding that the labor and benefits involved
in requesting are not shared equally in everyday interaction in Chachi com-
munities, it is perhaps possible to link such findings to broader social issues
beyond specific interactions.18 Indeed, if we look at how Foucault, one of
the social sciences’ most influential theorists of power, formulates power
not as a static relationship but as a way in which certain actions modify
others,19 it seems to fit quite well with conversation analysts’ conception of
the social as emergent from successive sequential actions:

In effect, what defines a relationship of power is that it is a mode of action which

does not act directly and immediately on others. Instead, it acts upon their

Re q u e s t i n g a s a Mea n s f or Ne g o t i at i n g   Ag e n c y   [ 73 ]
74

actions: an action upon an action, on existing actions or on those which may arise
in the present or the future…â•›.[A]â•„power relationship can only be articulated
on the basis of two elements which are each indispensable if it is really to be a
power relationship: that “the other” (the one over whom power is exercised) be
thoroughly recognized and maintained to the very end as a person who acts; and
that, faced with a relationship of power, a whole field of responses, reactions,
results, and possible inventions may open up.20

While perhaps it is a more micro-╉level context than is usually considered

in current, more processual understandings of inequality in the social sci-
ences more broadly, a response in a request sequence is one place where
we can directly observe at least some of the kinds of responsive actions
Foucault refers to. Addressees are in a position to make the request socially
meaningful by responding to it, either by fulfilling the request or contest-
ing it. When looking at peoples’ overwhelming tendency to fulfill requests,
we can see this as a (possibly evolved) trait of altruism and prosociality, but
we can also complicate this understanding by showing how this tendency
may also reflect asymmetrical social relations in terms of who requests and
who responds. While seeming to generate a paradox of concurrent coopera-
tion and inequality, the introduction of this complication does not make
humans’ social capacity for cooperation any less remarkable. Instead, it
highlights this capacity’s complex relation to other, sometimes more con-
tested dimensions of sociality.

NOTES

1. Current research on requesting places such practices in a broader context of

“recruitments,” practices for “making an immediate physical need, problem or
wish overt and publicly available, thereby providing an opportunity for a co-╉
participant to assist in meeting the need, resolving the problem, or fulfilling the
wish” (Drew and Couper-╉Kuhlen 2014:28). In this chapter, the more familiar
term “request” is used instead of “recruitment,” but it is meant in a broad sense
that encompasses “requesting” in a narrow sense as well as related practices of
“recruiting” assistance from others in interaction. Comparable cases of request/╉
recruitment sequences are identified using the “natural” or “sequential control”
method (Dingemanse and Floyd 2014).
2. In conversation analysis, a basic unit is the “adjacency pair,” consisting of two
“turns” fitted to each other (Sacks, Schegloff, and Jefferson 1974). Request
sequences are in many ways an example of an adjacency pair, although since the
“turn-╉taking” rules are formulated for speech and not for the types of practical
actions involved in such sequences, here the term “move” can help capture
broader type of multimodal “composite” sequential actions (Enfield 2009), so
“move” will be employed instead of the traditional “turn” in this chapter.

[â•›74â•›]â•… Language and Agency

  75

3. There has been a great deal of research on requests in pragmatics,

sociolinguistics, psychology, and related fields. Conversation analysis has put
the most emphasis on requesting as part of a request-​response; see Drew and
Couper-​Kuhlen (2014) for a review in a recent collected volume.
4. This study of Cha’palaa was conducted in the context of the Recruitments
Subproject (2010–​2014), part of the European Research Council project “Human
Sociality and Systems of Language Use,” directed by Nick Enfield and hosted
by the Language and Cognition Department (Stephen Levinson, director) of
the Max Planck Institute for Psycholinguistics. While this chapter specifically
discusses Cha’palaa, some of the ideas stem from group collaborations with
Julija Baranova, Joe Blythe, Mark Dingemanse, Paul Drew, Nick Enfield, Kobin
Kendrick, Giovanni Rossi, and Jörg Zinken, and will be further elaborated on
in future collective publications. As a coordinator of this subproject, the author
would like to acknowledge the contributions of all project members to expanding
our understanding of the domain of requests/​recruitments.
5. See Rossi and Zinken, ­chapter 9 in this volume, on the different roles of parties
involved in request sequences.
6. Sentences types, which are discussed in a huge array of literature, including one
well-​known summary by Sadock and Zwicky (1985), are also sometimes referred
to with the term “clause types.”
7. Indirect (non-​explicit, non-​imperative, off-​record, etc.) requests have long been
a topic of research concerning speech acts, politeness and inferential processes
in social interaction (Grice 1975, Searle 1975, Clark 1979, Brown and Levinson
1987, among many others).
8. See Rossi (2014) for a detailed account of the conditions of fully nonverbal
requesting based on data from Italian. The rate of fully nonverbal requests in the
Cha’palaa sample was low (4%).
9. Goffman 1955; Brown and Levinson 1987.
10. Brown and Levinson 1987:69.
11. Curl and Drew 2008 refer to an orientation assuming a high probability that a
request will be fulfilled as “entitlement.” One factor that relates to whether people
take such stances in request sequences (among others, like the cost or “contingency”
of the activity, in Curl and Drew’s terms) is the relative social (a)symmetry of a dyad
in a given context.
12. Wilson 1975, Sober and Wilson 1999, Bowles and Gintis 2011, and many others.
13. See, for example, Tomasello 2009.
14. After Marx’s death, Engels applied Marx’s analysis to some more family-​internal
and community-​internal dynamics in his 1884 work The Origin of the Family,
Private Property and the State, and researchers have gone on to find more
connections to current understandings of multiple intersecting planes of social
identity beyond economic class.
15. For Weber (1978), “status” was one of three major dimensions of social
stratification, along with economic class and “party” (i.e., religious-​type affiliations).
16. Weber 1978:53.
17. This line is expressed in much of Schegloff’s work (e.g., Schegloff 1993).
18. Some work has taken up Schegloff’s challenge to seek to empirically ground
social categories like gender in interactional analysis; see, for example,
Kitzinger 2013.
19. Here is the full quote: “The exercise of power is not simply a relationship
between partners, individual or collective; it is a way in which certain actions

Re q u e s t i n g a s a Mea n s f or Ne g o t i at i n g   Ag e n c y   [ 75 ]
76

modify others. Which is to say, of course, that something called Power, with or
without a capital letter, which is assumed to exist universally in a concentrated
or diffused form, does not exist. Power exists only when it is put into action,
even if, of course, it is integrated into a disparate field of possibilities brought to
bear upon permanent structures.” Foucault 1982:788, author’s emphasis.
20. Foucault 1982:789, author’s emphasis.

REFERENCES

Bowles, Samuel, and Herbert Gintis. 2011. A Cooperative Species: Human Reciprocity

and Its Evolution. Princeton, NJ: Princeton University Press.
Brown, Penelope, and Stephen C. Levinson. 1987. Politeness: Some Universals in
Language Usage. Cambridge: Cambridge University Press.
Clark, Herbert H. 1979. “Responding to Indirect Speech Acts.” Cognitive Psychology
11(4): 430–╉477.
Curl, Traci S., and Paul Drew. 2008. “Contingency and Action: A Comparison of Two
Forms of Requesting.” Research on Language & Social Interaction 41(2):129–╉153.
Dingemanse, Mark, and Simeon Floyd. 2014. “Conversation across Cultures.” In
The Cambridge Handbook of Linguistic Anthropology, edited by N. J. Enfield,
Paul Kockelman, and Jack Sidnell, 447–╉480. Cambridge: Cambridge
University Press.
Drew, Paul, and Elizabeth Couper-╉Kuhlen. 2014. “Requesting—╉From Speech Act
to Recruitment.” In Requesting in Social Interaction, edited by Paul Drew and
Elizabeth Couper-╉Kuhlen, 1–╉34. Amsterdam/╉Philadelphia: John Benjamins.
Enfield, N. J. 2009. The Anatomy of Meaning: Speech, Gesture, and Composite Utterances.
Cambridge: Cambridge University Press.
Foucault, Michel. 1982. “The Subject and Power.” Critical Inquiry 8(4): 777–╉795.
Goffman, Erving. 1955. “On Face-╉Work: An Analysis of Ritual Elements in Social
Interaction.” Psychiatry 18(3): 213–╉231.
Grice, H. Paul. 1975. “Logic and Conversation.” In Syntax and Semantics, edited by
Peter Cole and Jerry L. Morgan, 3:41–╉58. New York: Academic Press.
Kitzinger, Celia. 2013. “Conversation Analysis and Gender and Sexuality.” In The
Encyclopedia of Applied Linguistics, edited by Carol A. Chapelle, 996–╉1000.
Oxford, UK: Wiley-╉Blackwell.
Rossi, Giovanni. 2014. “When Do People Not Use Language to Make Requests?” In
Requesting in Social Interaction, edited by Paul Drew and Elizabeth Couper-╉
Kuhlen, 301–╉332. Amsterdam: John Benjamins.
Sacks, Harvey, Emanuel A. Schegloff, and Gail Jefferson. 1974. “A Simplest
Systematics for the Organization of Turn-╉Taking for Conversation.” Language
50(4): 696–╉735.
Sadock, Jerrold M., and Arnold Zwicky. 1985. “Speech Act Distinctions in Syntax.” In
Language Typology and Syntactic Description, edited by Timothy Shopen, 155–╉
196. Cambridge: Cambridge University Press.
Schegloff, Emanuel. 1993. “Reflections on Talk and Social Structure.” In Talk and
Social Structure, edited by Deirdre Boden and Don H. Zimmerman, 44–╉70.
Cambridge, UK: Polity Press.

[â•›76â•›]â•… Language and Agency

  77

Searle, John R. 1975. “A Taxonomy of Illocutionary Acts.” In Language, Mind, and

Knowledge, edited by K. Günderson, 344–​369. Vol. 7, Minneapolis Studies in
the Philosophy of Science. Minneapolis: University of Minneapolis Press.
Sober, Elliott, and David Sloan Wilson. 1999. Unto Others: The Evolution and
Psychology of Unselfish Behavior. Cambridge, MA: Harvard University Press.
Tomasello, Michael. 2009. Why We Cooperate. Cambridge, MA: MIT Press.
Weber, Max. (1922) 1978. Economy and Society: An Outline of Interpretive Sociology.
Berkeley: University of California Press.
Wilson, Edward O. 1975. Sociobiology: The New Synthesis. Cambridge, MA: Belknap
Press of Harvard University Press.

Re q u e s t i n g a s a Mea n s f or Ne g o t i at i n g   Ag e n c y   [ 77 ]
78
  79

CHAPTER 9

Social Agency and Grammar

GIOVANNI ROSSI AND JÖRG ZINKEN

O ne of the most everyday and conspicuous moments in which human

agency is distributed is when we ask others for help. This happens
all the time in daily life:  we constantly rely on others for small and big
practicalities such as getting the salt, moving a sofa, or cooking a meal.
Requests have long been a topic of interest in the social and behavioral sci-
ences, including research concerned with the distribution of human agency
(see Floyd, ­chapter 8, this volume; Dingemanse, ­chapter 7. this volume).
Some of our everyday requests are motivated by an individual goal (e.g.,
“can I have a glass of water? I’m so thirsty”), others by a common goal (“give
me the cards so I can deal them to everyone”). Some requests are isolated
from whatever else is going on at the moment, while others are bound up
in ongoing activities. What is common to all actions of requesting is that
a person’s behavior is recruited by another to achieve a certain goal. In the
domain of practical action, this typically involves manipulating the mate-
rial environment or changing the course of one’s bodily conduct.
Research on requesting has documented a variety of strategies that
people of different cultures use to get another person’s help (see Drew and
Couper-​Kuhlen 2014, Rossi 2015, for recent reviews). Not surprisingly,
these strategies often involve language (Floyd, Rossi, and Enfield under
review). Language has a number of properties that aid in making requests.
One is the facility to encode symbolic meaning, which allows a speaker to
specify the action requested with a verb (e.g. “pass”) and the object impli-
cated in the action with a noun (e.g. “salt”). Language can also specify who
80

is to participate in the action requested by means of person marking. Since

requests are typically made by an individual to another, linguistic forms
like imperatives (e.g., “pass the salt”) and recurrent types of interrogatives
(e.g., “can you move over a little?”) either encode or conventionally imply a
second person as the requestee. Second person marking has obvious con-
sequences for the distribution of agency in a requesting event, in that it
designates another person as the animator of the action requested—​the
participant who will physically carry it out—​while the speaker usually
maintains the role of author and possibly principal of the action—​the
participant who decides what is to be done and who is responsible for its
effects (see Goffman 1981: ch. 3; Enfield, ­chapter 2, this volume). This is
the typical way in which agency is distributed in a requesting event. But it’s
not the only one. In this chapter, we show that languages offer grammati-
cal structures that afford alternative ways of distributing the agency of an
action. We focus in particular on impersonal deontic declaratives like “it is
necessary to x,” which are used in several languages to express the need for
an action without tying it to any particular individual (see Malchukov and
Siewierska 2011).
From a grammatical point of view, impersonal deontic declaratives have
three core components. One is deontic meaning, which invokes a social
obligation that makes relevant the doing of something. Another is declara-
tive form, which—​unlike an imperative or an interrogative—​doesn’t con-
strain how people should respond to it (Vinkhuyzen and Szymanski 2005).
Third, and most important, the statement doesn’t specify the agent of the
necessary action. In languages like Italian and Polish, this can be done by
using either an impersonal verb, such as Italian bisogna or Polish trzeba (“it
is necessary to”), or an intransitive construction with a nonhuman subject
(e.g. “the rubbish has to be thrown out”).
From an interactional point of view, the grammatical makeup of an
impersonal deontic declarative has two main consequences. One is that
the same statement can be used to accomplish actions as different as (1) a
request for someone to do something or (2) an account of the speaker’s doing
that something. In these two scenarios, while the action is formulated as
necessary for anyone in the current situation, only one person becomes
involved in it (see Rossi and Zinken in press for an analysis of how these
alternatives come about). The second property is that, at other times, the
participation in the necessary action can be negotiated or avoided. These
are cases in which the distribution of agency emerges interactionally and
can be observed in vivo. This will be our focus in the rest of the chapter.
Consider a first case from Italian. Sergio, Greta, and Dino are chatting in
a kitchen while Sergio dyes Greta’s hair. Just before the fragment begins,

[ 80 ]  Language and Agency

  81

Greta has asked Sergio to remove a “thingy” from her forehead, which he
identifies as a wisp of hair (line 1). Sergio realizes that the hair has glued up
because some dye has run down on Greta’s forehead. This leads him to state
that: “it is necessary also to wipe away the dye from the forehead.”

Fragment 1 Tinta_​2051380 (Italian)

1 Sergio [questo-​((gets hold of wisp of hair))
this-​
2 Greta [(eh non lo so) c’ho un coso
(pcl not 3s.a know-​1s) ex=have-​1s a thingy
(well dunno) I’ve got a thingy
3 Sergio scusa sì bisogna pu[lire:: anche la crema dalla fronte
sorry yes necessitate-​3s clean-​inf also the cream from-​
  the forehead
sorry yes it is necessary also to wipe away:: the dye from
the forehead
4 Dino [((turns and reaches))  faccio io
do.1s 1s.n
I’ll do it
5 Greta [((reaches out for kitchen paper))

Sergio’s statement presents wiping away the dye as the right thing to do
for anyone in the current situation. Here, the action could be done by any
of the three participants, including Sergio himself. He is the person who
is most directly involved in the dyeing process, and who is responsible
for having let the dye drip on Greta’s forehead, for which he apologizes
(“sorry”). While saying bisogna “it is necessary to,” Sergio moves his hand
in the direction of the table, possibly toward the kitchen paper, but then
hesitates. At the same time, he gazes at Dino, and in so doing prompts him
to get involved. Dino is arguably in a better position to wipe Greta’s fore-
head, because Sergio’s hands are stained with dye. Also, Dino has already
assisted Sergio more than once earlier in the interaction. Here again, Dino
volunteers his help (“I’ll do it,” line 4). But, as he reaches for the kitchen
paper, Greta gets to it before him and goes on to wipe her own forehead.
In this case, two recipients of an impersonal deontic declarative find
themselves “competing” for the animatorship of the necessary action.
Although Dino makes an explicit bid for it (“I’ll do it”), it is Greta who

S o c i a l Ag e n c y a n d G r a m m a r   [ 81 ]
82

eventually takes it on without much explanation—​after all, it is her

own body.
The indeterminacy of animatorship left by an impersonal deontic declara-
tive is shown even more clearly in cases where this becomes a matter of
explicit negotiation. In fragment 2, from Polish, the family are beginning
their lunch. After wishing everyone an enjoyable meal, Jacek (the father)
moves to sit down, while Ilona (the mother) surveys the laid table (line 3).
She then says something to get Jacek’s attention (“eh you know what,”
line 4) and begins an impersonal deontic statement about the need of doing
something with the juice.

Fragment 2 PP2-​4 00:19 (Polish)

1 Ilona smacznego
tasty-​G
enjoy your meal
2 Jacek smacznego (  )
tasty-​G
enjoy your meal (  )
3 (1.2)/​((Ilona touches gravy jug and napkin holder, then
   raises left arm))
4 Ilona y wiesz co kom[potu¿     (.)
PCL know-​2S what juice-​G
eh you know what, juice
5 [((reaches right arm across the table for the
  juice jug))
6 Ilona trzeba ((touches jug))
necessitate
it is necessary
7 Jacek już [ja naleję]
already 1S.N PF-​pour-​1S
right now, I’ll pour it
8 [((Ilona lifts the jug))]
9 Ilona [wziąć] dob[rze
take-​INF  good
to take  good
10   [((Ilona puts the jug down))

[ 82 ]  Language and Agency

  83

11 Jacek [ty nalejesz¿ ja

you PF-​pour-​2S 1S.N
you’ll pour it? I’ll
12 na[leję
PF-​pour-​1S
pour it
13 Ilona [to weź już to nalej
   then take-​IMP already then PF-​pour-​IMP
   then go ahead already, pour then
14 (.)/​((Jacek takes the jug))
15 Jacek będzie mi wygodniej ((picks up jug))
be.FUT-​3S 1S.D convenient-​CMPR
it will be more convenient for me
16 Jacek ((pours juice into everyone’s glasses))

As Ilona begins the statement (“juice … l”), she reaches across the table
toward the juice jug, thus making a bid to take charge of the need she is in the
process of formulating. As Ilona continues the statement (“ … it is neces-
sary … ”) and her hand reaches the jug (line 6), Jacek begins responding by
saying that he will do the necessary action “right now” (line 7). Immediately
after this, while Ilona lifts the jug and completes the statement (“ … to
take”), Jacek continues his response by explicitly volunteering to pour the
juice (“I’ll pour it,” line 7). Ilona then accepts the offer (“good,” line 9) and
puts the jug back down (line 10). Meanwhile, Jacek has started to retract
(‘you pour it?,” line 11), but after Ilona’s acceptance he repeats his offer (“I’ll
pour it,” line 12), which is then ratified by Ilona (line 13). While proceeding
to pour the juice, Jacek adds an explanation for why he is taking over the job
from Ilona: “it will be more convenient for me” (line 15).
An impersonal deontic declarative doesn’t constrain participation in
the necessary action, but leaves it open to any relevant contributor. In
fragment 2, the interaction initially favors a certain individual to take
on the action (Ilona), but its subsequent development leads to a change
of animatorship. When Ilona begins the statement, she is projectably
engaged in pouring the juice. At the same time, however, the preface to
her statement “you know what” (line 4) marks it as relevant for Jacek and
calls for his attention to what is happening. This complex configuration
leads to an explicit negotiation of who will pour the juice. The result is that

S o c i a l Ag e n c y a n d G r a m m a r   [ 83 ]
84

the animatorship of the necessary action is eventually taken over by the

participant who is best positioned to do it.
The last case to be discussed illustrates a reverse situation from the first
two, one in which, instead of competing for the animatorship of the nec-
essary action, people try to avoid becoming involved. Michele and Remo
are looking for a utensil to pour tea into mugs. In line 1, Michele notices a
soup ladle in the sink, immersed in a dirty pot (“well there is this one”). As
Michele gets hold of the ladle and inspects it, Remo states that “it would be
necessary to wash it.”

Fragment 3 Camillo_​1241239 (Italian)

1 Michele cè ci sarebbe questo
pcl ex be-​cnd-​3s this
well there is this one
2 (2.4)/​((reaches towards soup ladle in the sink))
3 Michele c’è un po’ di pizzoccheri (nel recipiente) però ((grabs ladle))
ex=be.3s a bit of name (in-​the container) but
there’s some pizzoccheri (in the container) but
4 Remo allora bisognerebbe lavarlo
then necessitate-​cnd-​3s wash-​inf=3p.a
then it would be necessary to wash it
5 (2.6)/​((Michele shakes ladle and lifts it out of pot))
6 (0.5)/​((Remo walks to kitchen door))
7 Michele ma vuoi farlo tu ((turns to Remo))
but want-​2s do-​inf=3s.a 2s.n
do you want to do it?
8 (0.3)
9 Remo come
how
sorry?
10 Michele ( ) lavarlo (volevi lavarlo)
wash-​inf=3s.a (want-​ipf-​2s wash-​inf=3s.a)
( ) wash-​(you wanted to wash it)?
11 Remo no no no ( ) perché stavo ( ) perché io ho un impegno
no no no because stay-​ipf-​1s because 1s.n have-​1s one
commitment
no no no ( ) because I was ( ) because I’ve got
something to do

[ 84 ]  Language and Agency

╇ 85

As in the two cases above, before the impersonal deontic statement is

made, none of the participants is particularly responsible for the action
in question. One element that may at first glance tilt the balance is that,
at the time of the statement, Michele is holding the ladle (line 3), which
puts him in a better position to go ahead and wash it. But note that this
manipulation is motivated by his just-╉prior noticing and doesn’t imply any
commitment to wash it. Also, Remo is standing just beside Michele (up
until line 6), with visibly nothing to do. This appears to warrant a negotia-
tion of animatorship. In line 6, Remo walks away from the sink, making a
bid to leave Michele in charge. Michele then attempts to devolve the wash-
ing to him (“do you want to do it?”), bringing into the open the matter of
designating the agent. After giving a rather opaque excuse (line 11), Remo
dodges becoming involved, and Michele is left to do the washing.
Negotiations of animatorship such as those we have examined in this
chapter generally do not occur with forms that designate another person
as the animator of an action (e.g. “pass the salt” or “can you move over a
little?”). An impersonal deontic declarative, on the other hand, affords such
negotiations by virtue of its grammatical design, which doesn’t restrict ani-
matorship to a single individual. Although there are a number of cases in
which the responsibility for the action in question falls on a specific person
(see Rossi and Zinken in press), impersonal deontic declaratives have the
potential to generate complex interactions in which the identity of the ani-
mator must be sorted out. The three cases we have examined also illustrate
the kinds of reasons at play in these interactions, including the presence of
obstacles or concurrent commitments, the relative easiness of the action
for different people, the rights over one’s body.
The case of impersonal deontic statements demonstrates the impor-
tance of grammatical detail for social action. What may at first glance
appear only subtle, differences of expression turn out to put constraints
on what people can or should do in a given situation. Moreover, given the
great diversity among languages, grammatical variation will be conse-
quential also for social interaction across cultures. Studying grammar as a
resource for social action, and in particular for the distribution of agency, is
therefore a crucial piece of the science of human cooperation and sociality.

REFERENCES

Drew, Paul, and Elizabeth Couper-╉Kuhlen. 2014. “Requesting—╉From Speech Act

to Recruitment.” In Requesting in Social Interaction, Studies in Language and
Social Interaction, edited by P. Drew and E. Couper-╉Kuhlen, 1–╉34. Amsterdam/╉
Philadelphia: John Benjamins.

S o c i a l Ag e n c y a n d G r a m m a r â•… [â•›85â•›]
86

Floyd, Simeon, Giovanni Rossi, and N. J. Enfield, eds. Under review. Getting Others to
Do Things: A Pragmatic Typology of Recruitments. Berlin: Language Science Press.
Goffman, Erving. 1981. Forms of Talk. Philadelphia: University of Pennsylvania Press.
Malchukov, Andrej Lvovich, and Anna Siewierska, eds. 2011. Impersonal
Constructions: A Cross-​Linguistic Perspective. Amsterdam/​Philadelphia: John
Benjamins.
Rossi, Giovanni. 2015. “The Request System in Italian Interaction.” Ph.D. dissertation,
Radboud University, Nijmegen, the Netherlands.
Rossi, Giovanni and Jörg Zinken. 2016. “Grammar and Social Agency: The Pragmatics
of Impersonal Deontic Statements.” Language 92(4).
Vinkhuyzen, Erik, and Margaret H. Szymanski. 2005. “Would You Like to Do It
Yourself? Service Requests and Their Non-​Granting Responses.” In Applying
Conversation Analysis, edited by K. Richards and P. Seedhouse, 91–​106.
New York: Palgrave Macmillan.

[ 86 ]  Language and Agency

  87

CHAPTER 10

Distributed Agency and Action

under the Radar of Accountability
JACK SIDNELL

C ommunicative, purposive social action is semiotic and involves one

person getting another to recognize her intention.1 As Sidnell and
Enfield (2014) write, “A social action’s effect is an interpretant of some-
one’s sign, not a mere effect of someone’s cause. Social actions only bring
about effects by virtue of someone recognizing what is intended, and react-
ing appropriately.” Along these lines, conversation analysts have shown
that much of what gets done in interaction is accomplished by the use of
“practices of speaking,” which, when deployed in some specific sequential
context, are recognizable as implementing a particular action. So when,
for instance, in the course of a telephone call, Janet says, speaking of her
daughter who is at Ann’s house, “Do you want me to come and get her?” this
is understood, by Ann, to implement an offer rather than as merely asking
about her wants and so on. In this [[practice of speaking in context] -​>
[recognizable action] -​> [interactional goal]] scenario, speakers are under-
stood as accountable for having done whatever action is so effected. This
accountability is realized in various ways. In perhaps the most basic case,
a next speaker produces a turn that treats the previous speaker as having
done the action for which this turn is an appropriately fitted response (e.g.,
next speaker produces an acceptance to an “offer,” or an answer to a “ques-
tion”). So for instance in the example mentioned, Ann eventually accepts
Janet’s offer. Alternatively, a next speaker may name the action he or she
88

understands the first speaker to have done, saying, for instance, “thank you
for offering” and so on.2
So by making it recognizable to a recipient what action(s) one intends
to accomplish, a speaker thereby necessarily becomes accountable for hav-
ing done that action. But are there ways to achieve some interactional goal
without having to do the action that would lead to that outcome by virtue
of a recipient’s recognition? Relatedly, is there a way to achieve an interac-
tional goal without, in the process, becoming accountable as the agent who
effected it? In what follows I will describe three kinds of cases that suggest
an affirmative answer to these questions—​there are indeed methods by
which a participant in interaction can accomplish interactional goals while
flying beneath the radar of accountability.3

GETTING A CALLER’S NAME BY GIVING ONE’S OWN

First, let us consider a case, described by Sacks (1995), in which insti-

tutional call-​takers want to get a caller’s name. Sacks noted that in calls
to a suicide prevention hotline requests for the caller’s name were often
met with suspicious resistance and questions such as “Why do you need
to know?” In order to avoid this eventuality, call-​takers would routinely
adopt a different strategy, introducing themselves as, for example, “Mr.
Smith” and thereby eliciting reciprocal name identifications without going
on record as having asked for this information.
Sacks noticed, then, that there was a “fittedness” relation between the
two parts of an adjacency pair, so that if the first person says “Hello,” then
so does the second; if the first person says “This is Mr. Smith,” then the
second tends to say “This is Mr. Brown,” and so on. If there is a tendency to
fit the form of the return to the form of the initiating action in any of these
greeting (“hello”) or introduction (“This is Mr. Smith”) pairs, then there
exists a way of getting someone’s name without asking for it. Rather than
asking “What is your name?” one can say “This is Mr. Smith,” and this will
establish the relevance of the other giving his or her name.
Sacks goes on to note that if the staff member asks “Would you give me
your name?” the caller can reply “Why?” or “What for?” That is, the caller
can ask the staff member to provide a reason or account for asking for the
name. Sacks (1995: 4) notes that “what one does with ‘Why?’ is to propose
about some action that it is an ‘accountable action’. That is to say, ‘Why?’
is a way of asking for an account.” Sacks goes on to suggest that accounts
“control activities.” The important point here is that “This is Mr. Smith,

[ 88 ]  Language and Agency

  89

may I help you?” may be a way of getting the other person to provide her
name, but it is not an accountable action in that respect; one cannot say
in response “Why do you want my name?” because the name was never
asked for.

FISHING FOR INFORMATION BY TELLING “MY SIDE”

A second example is provided by what Pomerantz (1980, following Sacks

1995) described as a “my-​side telling” in which one participant merely
tells of her limited access to some state of affairs thereby eliciting infor-
mation about it again without having to ask explicitly. So in the following
well-​known case, all Emma has to do is report “Yer line’s been busy,” and
thereby convey her “limited access” to some state of affairs that the recipi-
ent knows about. This prompts Nancy to provide information about how
it was that her phone was busy, who she was talking to, and why she was
talking for such an extended period.

(1) NB:II:2:R
01 Nancy: Hel-​lo:,
02 Emma: .hh HI::.
03 (.)
04 Nancy: Oh:’I::: ‘ow a:re you Emmah:
05 Emma: FI:NE yer LINE’S BEEN BUSY.
06 Nancy: Yea:h (.) my u.-​fuhh h-​.hhhh my fa:ther’s wife ca:lled me,h
07 .hhh So when she ca:lls me::,h .hh I always talk fer a lo:ng
08 ti:me cz she c’n afford it’n I ca:n’t.hhh[hhh]°huh°]
09 Emma: [OH:] : : : : : ]:=
10 =my [go:sh ah th ]aht my phone wz outta order:
11 Nancy: [-​-​Ah :::::::, ]

Emma is able to elicit information without having to ask for it and thus
without having to do something that might be characterizable as “nosey”
and so on. Where the telling of limited access does not elicit the information

A g e n c y a n d A c t i o n u n de r Rada r of Ac c o u n ta b i l i t y   [ 89 ]
90

wanted, a participant may follow up with an on-​record, explicit question,

as in the following case:

(2) Geri-​Shirley
01 Geri: Oka:y d’dju just hear me pull up?=
02 Shirley: =.hhhh -NO:. I wz -TRYing you all day.’n the LINE wz busy
03 fer like hours.
04 Geri: Ohh:::::::, ohh:::::. .hhhhhh We::ll, hhh ah’m g’nna c’m
05 over ‘n a li’l while help yer brother ou:t,
06 Shirley: Goo[:d.
07 Geri: [.hh Cz I know he needs some he::lp, ((mournfully))
08 Shirley: .khh Ye:ah. Yeh he’d mention’ that t’day.=
09 Geri: =M-​hm,=
10 Shirley: =.hhh Uh:m, .tch.hhhh Who w’yih -ta:lking to:
11 (0.6)
12 Geri: Jis no:w?
13 Shirley: .hhhh No I called be-​like between ele[ven en
14 Geri: [I: wasn’talkeen tuh
15 a:nybuddy. (b) Bo-​oth Marla’n I slept ‘ntil about noo:n,=
16 Shirley: =O[h.
17 Geri: [.hhhh en w’n I woke up, I wannid tih call my mother.
18 Shirley: Mm-​[hm,
19 Geri: [.hhhh en I picked up the pho:ne, e:n I couldn’t dial
20 out.’n[I thought ar phone wz out’v order.[‘n I-​.hhh
21 Shirley:  [Oh: (  ),   [Yeh,
22 Geri: said w’l oyll I: bet I know w’t happened,=
23 =[(so) I walked-​]
24 Shirley: =[Marla left th’]phone o[ff the hook,]=

Here, when Shirley’s initial conveyance of limited access does not elicit
information about Geri’s use of the phone, she eventually formulates a
direct inquiry, asking, at line 10, “Who w’yih -​ta:lking to:.” Sticking with
examples in which one participant reports trouble in reaching or find-
ing another, we can see that not only is this utterly pervasive, it is also
quite tightly structured. And so when, in example (3), Penny reports,
“I triedju all morning long’n it wz really busy,” prompting Pat’s “Yeh I
wz sleeping my mother wz answering,” Penny then goes on to elaborate
her “side.”

[ 90 ]  Language and Agency

╇ 91

(3) Houseburning
01 Penny: =.hh I said tih myself I triedju all morning long’n it
02 wz really busy. .hhhh[hhh]=
03 Pat: [Yea][h,
04 Penny: [=en: : :[d,
05 Pat: ╅╅╅╅╇[Yeh I wz sleeping my
06 mother wz answering, hh[hh
07 Penny: [.hh So-╉Uhri:ght.
08 Then, right, that’s good.That’s good. .hh
09 But so I’ve been trying all along en I thought maybe thet
10 chih took the phone off the hook.

My-╉side tellings serve as a technique for eliciting information without hav-

ing to ask a direct (indeed, any) question. Given that actions are account-
able, a person asking a question should be entitled to ask it and moreover
have a good reason for doing so. Responses to questions that take such
forms as “why do you want to know?,” “why do you ask?,” or even “it’s none
of your business” hint at this accountability of actions as do reports about
a person being a “gossip,” “nosey parker,” “busy-╉body,” and so on. With
this is mind we can see how the practice Pomerantz describes might have
important interactional uses since with it a person may elicit information
about some specific state of affairs without being accountable for having
asked a question.

ELICITING AN OFFER BY DESCRIBING A TROUBLE

Finally, we can see a similar kind of pattern across quite different examples
in which rather than make a request, a speaker merely describes a trou-
ble, thereby inviting the other to offer what is wanted (see, for discussion,
Pomerantz and Heritage 2012). Routinely, description of a trouble leads
the recipient to offer a remedy, but again, if the offer is not made or is not
sufficient, the speaker can then “go on record” and make the request, as in
the following case in which a landlady initially establishes the presence of
a laundry machine in her basement apartment before going on to report
the problem of having “no water” to the tenant. When the tenant offers a
solution that would merely fix the immediate problem, Alice ends up going
on record with the request saying, “Well I don’t want you to shut it off mid-╉
load, just don’t run it anymore.”

A g e n c y a n d A c t i o n u n de r Rada r of Ac c o u n ta b i l i t y â•… [â•›91â•›]

92

(4) YYZ-​Shawn & Alice

01 Alice: [.hh Do you guys have a washer en dry:er er somethi:ng?
02 (0.6)
03 Shawn: ah::=yeah,
04 we got a little washer down here. (.) ogoino.
05 Alice: ↓oh::.
06 ok. [I didn’]-​I didn’t know thet you guys ha:d tha:t
07 Shawn: [water]
08 (0.4)
09 Shawn: kay=
10 Alice: =an u::m and I’ve got no water in my house. now.
11 Alice: I-​see-​I  [jus]
12 Shawn: [ ok]
13 Alice: I don’t have a big enough water heater for that.
.
14 [ hh]
15 Shawn: [ ok]
16 Alice: in the house o(h)k(hh)?,
17 (0.4)
18 Shawn: onoproblem.o=
19 Alice: [=than-​ ]           [Tha:nks]
20 Shawn: [=I’ ll] ge Carrie to shut it off [o(som e] water)o
21 (.)
22 Alice: well, I mean, don’t-​I don’t want you to shut it off mid-​load
23 bu:t (.) just don’t run it anymore
24 (0.1)
25 after this one.
26 Shawn: (.) oAri:ghto
27 Alice: Ok thanks
28 Shawn: mbye
29 Alice: Bye.

And in the following well-​known case, Donny sets up his telling of a trou-
ble with the pre-​announcement “Guess what” in line 6. The problem is
then described in incremental fashion (lines 8 and 10), but this does not
elicit more of a response than the information receipt in line 11. At line
15 Donny seems to be on the way to formulating a request but leaves this
unfinished in order to continue with the telling of the trouble. It is at

[ 92 ]  Language and Agency

  93

this point that Marcia interjects, indicating that she has understood what
Donny wants and that, if possible, she would comply.

(5) Donnie & Marcia

01 Marcia: Hello?
02 Donny: ‘lo Marcia,=
03 Marcia: Yea[:h
04 Donny:   [=(‘t’s) Donny.
05 Marcia: Hi Donny.
06 Donny: Guess what.hh
07 Marcia: What.
08 Donny: o
hh My ca:r is sta::lled.
09 (0.2)
10 (‘n) I’m up here in the Glen>
11 Marcia: Oh::.
12 [(0.4)]
13 Donny: [ohhh] a:nd.hh
14 (0.2)
15 I don’t know if it’s possible, but [ohhh] see
16 I haveta open up the ba:nk.hh
17 (0.3)
18 a:t uh: (.) Brentwood?hh=
19 Marcia: =Yeah: en I know you want-​(.) en I
20 would, but-​except I’ve gotta in aybout five
21 min(h)utes[(hheh)
22 Donny:   [Okay then I gotta call somebody
23 else.right away.
24 (.)
25 Donny: Okay?=
26 =Okay [Don  ]
27 Donny:      [Thanks] a lot.=Bye-​.
28 Marcia: Bye:.

And finally, in example (6) Ann is talking to Janet, whose daughter is at

Ann’s house for a visit. Here, all Ann has to do is remark that she is “run-
ning a bit late” to elicit from Janet the offer to “come and get her.”

A g e n c y a n d A c t i o n u n de r Rada r of Ac c o u n ta b i l i t y   [ 93 ]
94

(6) xtr. 1
01 Ann: =Hello::
02 Janet: Oh=hi:_╉=it’s Janet_╉ [Shelby’s mo]m
03 Ann: [hi:  Janet]
04 How eryou(h) .h hh
05 Janet: I’m goo:d,how are y [ou
06 Ann: â•…â•…â•…â•…â•…â•…â•…â•…[I’m fi:ne.h[we’re actually: uhm
07 Janet: â•…â•…â•…â•…[ogoodo
08 (0.2)
09 Ann: we’re running bit late=but we’re
10 (.)
11 on our wa(h)y:
12 Janet: Do you want me to come an’ get her?
13 Ann: Uhm:, it doesn’t matte:r, like(hh)
14 (0.4)
15 .hhhhh
16 Janet: I-╉ I could.it’s very easy.
17 so rather than you h:av(h)e(h)
18 (.)
19 you know (.) tuh get everybody ou[t
20 â•…â•…â•…â•…â•…â•…â•…â•…â•…â•…â•…â•…â•…â•…[.hhh
21 I’ll justa
22 (0.2)
23 come dow:n.=

Across these examples then we see that the telling of a trouble or even the
mere hint of one (as in the last example) can prompt from the recipient an
offer of assistance or at least an account for not producing such an offer.

CONCLUSION

Examples such as these show us then that participants are acutely aware of
the accountability that attaches to the actions they are recognized as doing
and, in certain circumstances at least, seek to avoid that accountability by
achieving their interactional goals by alternate means. Across all the cases
discussed here there seems to be an interactional short-╉circuit whereby
speakers are able to bypass the action/╉recognition node of the [practice] -╉>
[action/╉recognition] -╉> [interactional goal] chain such that they are able to
recruit the agency of another without being accountable for having done

[â•›94â•›]â•… Language and Agency

╇ 95

so. Recruited others provide their own names, provide information, or pro-
vide offers of assistance and thereby commit themselves as agents to some
joint venture.
I have suggested that much of what gets done in interaction is accom-
plished by the use of practices of speaking that, when deployed in some
specific sequential context, are recognized as implementing a particular
action. So when Janet says, “Do you want me to come and get her?” this is
understood to implement an offer rather than as merely asking about what
Anne wants or desires. When speakers implement actions in this way, using
practices that are designed to be recognizably implementing them, they “go
on record” and are thus accountable for having done the action in question.
Recipients, or nonrecipient participants, can thus ask why this action is
being done (“why do you ask?,” “why do you say that?,” “why do you want
to know?,” etc.), or they can report that the action has been done (“she
asked if I was going,” “she complained that it was too hot,” etc.). Obviously,
formulations of prior actions can be challenged or contested (see Sidnell
and Enfield 2014) but that does not change the underlying fact that par-
ticipants become accountable for the actions that they are recognized as
having done. Now of course there are various circumstances in which a
participant might want to “fly under the radar of accountability”—╉that is,
achieve some interactional outcome or goal without doing the action that
would effect this and thus without being accountable for having done it.
In this chapter I have briefly considered some cases in which we see how
participants are able to do this.

ACKNOWLEDGMENT

I am indebted to Paul Kockelman for very helpful comments on a previous

version of this chapter.

NOTES

1. By “communicative” I mean to distinguish what I am talking about here from

various kinds of purposive social action that do not depend on intention
recognition (e.g., Weber’s woodcutter [1978:8–╉9]). By “purposive” I mean to
distinguish what I am talking about from various kinds of communicative
(typically, indexical) “effects” that are a consequence of acting (for more
discussion of this distinction see Sidnell and Enfield 2014).
2. And there are many other ways in which the accountability may be made
manifest. A report may be made about what happened for instance (“She

A g e n c y a n d A c t i o n u n de r Rada r of Ac c o u n ta b i l i t y â•… [â•›95â•›]

96

offered me … ,” “She asked if I wanted her to do it … , ” and so on). This sense

of accountability is more expansive than one that emphasizes sanctions (e.g.,
Radcliffe-╉Brown 1954). Rather, on this view, an accountable action is one that
can be the object of inquiry—╉“Why did you do that?” (see Sacks 1995) and
that can be the subject of report (see Garfinkel [1967:1], “When I speak of
accountable … I mean observable-╉and-╉reportable, i.e. available to members as
situated practices of looking-╉and-╉telling”).
3. Some of the issues I develop here were originally raised by Pomerantz (1980)
and, of course, by Sacks (1995). Rosenblum’s (1987) discussion of “When is a
question an accusation?” also bears on many of these issues although she does
not herself thematize them.

REFERENCES

Garfinkel, H. 1967. Studies in Ethnomethodology. Englewood Cliffs, NJ: Prentice-╉Hall.

Pomerantz, A. M. 1980. “Telling My Side: ‘Limited Access’ as a ‘Fishing Device.’”
Sociological Inquiry 50: 186–╉198.
Pomerantz. A. M., and J. Heritage. 2012. “Preference.” In The Handbook of ConÂ�
versation Analysis, ed. J. Sidnell and T. Stivers, 210–╉228. Oxford, UK:
Wiley-╉Blackwell.
Radcliffe-╉Brown, Alfred R. 1954. “Social Sanctions.” In Structure and Function in
Primitive Society, 205–╉211. New York: Free Press.
Rosenblum, K. 1987. “When Is a Question an Accusation?” Semiotica 65(1–╉2):
143–╉156.
Sacks, H. 1995. Lectures on Conversation. 2 vols. Edited by Gail Jefferson. Oxford:
Blackwell.
Sidnell, J., and N. J. Enfield. 2014. “The Ontology of Action in Interaction.” In
The Cambridge Handbook of Linguistic Anthropology, 423–╉446. Cambridge:
Cambridge University Press.
Weber, M. 1978. Economy and Society: An Outline of Interpretive Sociology. Berkeley/╉
Los Angeles: University of California Press.

[â•›96â•›]â•… Language and Agency

  97

PART FOUR

Economy and Agency
  99

CHAPTER 11

Distributed Agency and Debt in

the Durational Ethics of Responsibility
JANE I. GUYER

INTRODUCTION: MODERN TIME

Walter Benjamin’s concept of “modern time” as “homogeneous” and

“empty” (Benjamin 1968:261) depicts the creation of time as a continuous
variable, which can then open up to marking and coordination in terms
of chosen punctuations and the dated time of a universal calendar (Guyer
2007). It provides the template for the temporal framing of contracts, which
can then enter into formal model-​building and projections, especially for
formalizations that can be measured and calculated as rates:  of interest,
of depletion/​erosion, of growth, and so on. The actual temporal durations
relevant to a particular instance can then be inserted. Calculation and pro-
jection are fundamental to the intricate synchronization of a mechanically
and bureaucratically mediated society and economy. What anthropolo-
gist Marc Augé (1995) calls the “non-​places” of supermodernity, such as
airports, would screech to a halt without modeled projections of tempo-
ral coordination for a vast variety of material variables, that also marshal
people, seen as solitary, homogeneous individuals, into temporary crowds.
In social sciences such as economics, where quantification and projec-
tion are fundamental, the temporal variable can be conceptualized ini-
tially in terms of abstract time, which then renders it open to eventual
marking, for particular purposes, in terms of specifically chosen punctua-
tions (Guyer 2007). An abstract model depicts the punctuations in terms
100

of a sequential series: time 1, time 2, time 3, and so on (written as T1, T2,

T3 … ). The simplest example for our purposes here would be the conclu-
sion of a financial commitment (at T2), under conditions of compound
interest, where the temporal space between T1 and T2 is then inserted to
calculate the principal-​plus-​interest by the termination date. So two years,
or five years, or any other temporal range is inserted into the abstract
model for situational prediction for particular cases, and for their combi-
nation into systemic prediction. Some possible contingent interventions
may also be turned into calculable probabilities and fed into the model.
In some contracts, for example, there are stated conditions for inflation
adjustment, if the currency value changes between T1 and T2. Or certain
substantive possibilities can be made explicit. For example, kinship is cen-
tral to the devolution of responsibility in case of the death of a party to
a contract between T1 and T2. All these models, however, are aimed at
exact calculability and predictability, through the assumption of abstract
time, which then makes them amenable to combination into the regimes
that are fundamental to “modernity.” Large numbers of temporal commit-
ments can be brought into tight consonance with each other when a debt
that is incurred now, at T1, is to be repaid at a defined interest rate, at a
defined T2, since this creates the conditions of the contract and the pre-
dictability of the encompassing regime. The individual is responsible only
in the moments of the contract enactment: T1 and T2. The interim is left
to his or her own management.
All possible contingencies, however, may not necessarily be calculable or
legally definable with respect to all parties, all commitments, and all time
frames. The “till death do us part” of the traditional marriage vows invokes
two moments in a lifelong commitment, but time 2 is not predictable, nor
is the temporal duration amenable to modeling the contract, and nor are
these contingencies happening in “empty, homogenous” time. They can
simply emerge from dynamics for which “modern,” “empty, homogeneous”
time hardly corresponds to the experiential time of relationships and of
life. And yet people do still make such commitments, mediate such contin-
gencies, and increasingly find that debt management—​incurred in “mod-
ern time”—​is deeply interwoven into the durational ethics of life (Guyer
2014). With respect to modernity, Augé has argued that the “anthropologi-
cal”—​that is, human and social—​qualities of memory, relationship, and
identity, which are grounded in “places,” do remain. The question is then,
in what form, with what implications, and for whom? “Modernity does not
obliterate them but pushes them into the background”: they “survive like
the words that express them,” “intertwined and tangled” together (Augé

[ 100 ]  Economy and Agency

  101

1995: 77, 107). Of the awkward incompatibilities that can arise from such
entanglement, the most obvious is between the homogeneous empty time
of modernity, and the temporal micro-​rhythms and indeterminacies of
life in biological and social mode. People gain and lose employment and
income, fall ill, age, organize and disorganize their social relationships,
and live with sudden shifts in general conditions (war, disease, disaster,
windfalls). And they may come together in existentially challenging ways.
None of them—​alone or in combination—​can easily be calculated as “risk”
and entered into debt contracts as such. It is in life, rather than debt law,
that causation, intention, assessment, responsibility, blame, the source
of ongoing momentum and the implications of failure to sustain it, work
out. Whereas in life, they all refer, in ways that vary with philosophical
principles, to a supple grid of ongoing interactions in textured time, to
which people commit themselves, without knowing what contingencies
may arise in life, and intersect with each other, in barely manageable ways.
The attempt to recalculate, and to live, these as calculable probabilities may
remain intractable to the abstract model, either by authorities, by experts,
or by people themselves. In many philosophies of life, it is ritual, of both
collective and intimate kinds, that mediates such temporalities and may
well distribute agency within the life processes between the abstract T1
and T2 of modern time.
The recent expansion of debt regimes, framed in empty homogeneous
time, meets these philosophies and exigencies in zones of tension. Debt
management is at least 5,000 years old, according to Graeber (2011), and
is fundamental to the financial regimes of present-​day capitalism. It can
therefore serve as an example of the modern effort to make the world of T1
and T2, and the intractable human realities and philosophical legacies, con-
sonant with each other in an ever-​expanding range of debt relationships.
For example, under American law, personal bankruptcy can be declared,
subject to the person’s submission to a “reorganization plan” mediated
by experts (United States Code, Chapter 13, Title 11). Such a law defines
agency and responsibility in exact, individuated, modeled terms; the con-
cept of “default” puts the individual at “fault,” even if life became contin-
gently intractable for them. My chapter suggests that this is one deeply
consequential aspect of the individuation of modernity proposed by Max
Weber in The Protestant Ethic and the Spirit of Capitalism:  an important,
although under-​examined, entailment of the entrepreneurial proposition
that Weber developed most strongly. Within debt practices, the older, intu-
itive, and culturally varied traditions of distributed agency face those of the
present formalized world.

Di s t r i b u t ed Ag e n c y a n d   De b t   [ 101 ]
102

BACKGROUND: TEMPORAL REGIMES AND DURATIONAL ETHICS

The social historical and anthropological records show that all temporal
regimes are comprised of multiple durations of varying length, running
concurrently. Most obviously, there are diurnal, lunar, seasonal/​annual,
and life cycles, nested within each other, with key moments of conver-
gence, whose ritual marking is often managed and celebrated in complex
calendrics that presume that people’s trajectories intersect, and often
apprehend and anticipate the possibility—​day to day, or eventually at the
end of time—​intervention by non–​human beings, whether spiritual enti-
ties or an all-​present God. Probably only a fundamentalist faith in sover-
eignty of some kind—​either “everything happens for a (divine) reason,”
or “the individual is master of his fate”—​could remove all cultural and
political framing of T1 and T2. Prospective imagination, in both cases, may
invoke fate: “what will be, will be,” and individuated responsibility. The pas-
sage of time itself, then, is so imbued with propositions of agency that its
study can show the full range of the beings that are accorded, or deprived
of, agency, and their mutuality of accompaniment, interruption, support.
and failure.
The anthropological study of how T1 and T2 are mutually entailed is
implicit in the ethnography of moments of convergence, such as the ritual-
ized performances of the life cycle through which a person, him/​herself,
is transformed. Certain entailments of personhood in a past stage of life
are abrogated, and new capacities are crafted into the enactment of social
and ethical life for the next phase. In western law, by contrast, a person
is defined as a child, a juvenile, or an adult on abstract calendrical time
rather than social and biological time. For the ethnographic examination
of agency, we first establish the parameters for personhood, and then turn
to their implications for the T1/​T2 mutual entailments among people.
General ethical precepts may be explicitly evoked to make things work and
to sanction people’s failures, but for us to examine the distributed agency
question closely, full immersion in the temporal process itself is required,
with respect to what I  have termed “durational ethics” (Guyer 2014).
Commitment. Promise. Responsibility. Perseverance, that is, Fortitude,
in classic Christian thought. Fortitude is one of the four cardinal virtues,
alongside Prudence, Justice, and Temperance, without which these other
three are impossible to realize.
By requiring a person to take on durational ethics, he/​she creates an
entailment between the T1 in which he or she is now living and the T2 when
the agency will be fulfilled. A promise usually refers only to the two moments
that are explicitly named, whereas a commitment or responsibility may be

[ 102 ]  Economy and Agency

╇ 103

in play all the time, between any two moments. Parental responsibility for a
child’s safety is continual, not intermittent. In popular practice, both nec-
essarily assume processes that continue alongside, and contribute to, real-
ization: the durational cycles of life, the knowledge that others have taken
on comparable commitments, under known material circumstances, and
that the gods are still in the heavens. I may be accountable individually, but
the place of others in “aiding and abetting,” or “confounding” is always in
the picture, in ways that vary by culture and philosophy. In a sense, distrib-
uted agency is the common-╉sense philosophy of most, if not all, durational
ethics. That I can repay my debts depends on you fulfilling your promises to
me, which you may neglect due to moral lapse or you may simply be unable,
due to illness or other unforeseen demands. Do I  make generous allow-
ances for your difficulties? Do I explain them to my own creditor? Where
does the responsibility rest: on those who seek for a narrow definition of
contractual justice, or on those whose behavior most affects social har-
mony at large? By being aware of philosophies of distributed agency with
respect to the temporality of debts, we may then focus on the history of
the modern measures to shear off popular legacies of distributed agency, in
order to produce a sparse definition and legal precision: individuals make
choices, in neutral time, for which they are solely responsible.

DEBT AS RELATIONSHIP AND REGIME

The topic in which philosophies of agency with respect to temporality and

agency, and to historical changes in such philosophies, can be most readily
studied anthropologically in the present, and with the clearest implications
for durational ethical and political regimes, is the topic of debt. A debt is
an explicitly defined promise to repay something given, usually in a defined
form and amount, at a defined date. In some systems, the initial giving ges-
ture of extending credit is accorded a positive moral aura, whereas being in
debt implies actual or potential culpability (Peebles 2010). The vast expan-
sion of personal and national sovereign debt in the world economy of the
past 30 years has created a system of mutual implication of debts among
themselves that is often referred to as a debt regime, which then puts mon-
etary debt and its moral underpinnings onto the intellectual radar screen
for many disciplines. Lazzarato (2012) succinctly reviews large-╉scale politi-
cal and philosophical approaches to debt, arguing that the agency question
is an ideological one. The proposition that “the temporality of debt [is pre-
sented as one of] possibility, choice and decision” (2012:44) invokes com-
pletely individuated capacities, in modern time. As a regime, as distinct

Di s t r i b u t ed Ag e n c y a n d   De b t â•… [â•›103â•›]
104

from a particular dyadic transaction, debt capacitates powerful centers of

control, precisely because agency in neutral time frames makes initial con-
ditions, and matters arising in the interim between T1 and T2, irrelevant to
the implementation of the terms through which the debts, rather than the
persons, are brought into substantive and temporal consonance. The indi-
vidual debtor’s agency is here shorn of the changing conditions that would
otherwise distribute accountability across multiple intersecting durations
and participants, accidents and life contingencies.
The new anthropological work on debt can be approached through two
texts that work in ways that are classic to the discipline. David Graeber
(2011) takes on the vast sweep of world social history over 5,000  years,
largely from the vast library of documentary sources (with hundreds of
footnotes and 37 pages of references). Clara Han (2012) studies a particu-
lar neighborhood of very poor people in Santiago, Chile, through classic
ethnographic methods attending in detail to particularity, sometime after
the fall of a dictatorship, so in the context of rapid change.
Graeber starts with the idea of “moral confusion” over debt. In theory
a debt is always an obligation to repay. In practice, failure to pay, in the
original terms and time frames, has been increasingly protected by state
power, bankruptcy laws, and other instruments, which work highly selec-
tively. Amato and Fantacci (2011) have made the same argument about the
capacity of states and large corporations to delay repayment, repeatedly.
The enactment of debt regimes at this macro level, they suggest, is engaged
in T1 (including under duress) and then imbued with differential power in
T2 when it comes to repayment. The preservation of any distributed con-
cept of agency in the passage of time is then differentially distributed, in
fact if not in explicit theory. The powerful can have recourse to the mitigat-
ing circumstances of sheer scale at T2, under the rubric of “too big to fail”
and all the complex fallout that is projected if they would end up in failure.
The powerless, by contrast—​as we see later—​have to submit to the debt
regimen of debt regimes that impose a modernist temporal model. Both of
these works, by both Graeber and by Amato and Fantacci, about the macro
level of debt, point out the centrality of payment for war in the creation
and validation of the commitments of the creditor (the taxpayer and finan-
cier in this case), while leaving the repayment by the state in non-​monetary
terms: eventual peace (a general reward), and favor in the postwar econ-
omy (specific benefits). In such a macro context, the small individuals may
become “delinquent” on debt repayment in ways that would otherwise be
wholly understandable in intimate contexts: they lost a job, someone died
and they had to make up the gap somehow, the economy tanked, and so
on. But they will be subject to a bankruptcy law that recuperates much of

[ 104 ]  Economy and Agency

  105

their assets and submits it to compulsory legal disciplines. Graeber’s argu-

ment thus rests on the historical evidence for monetary debt having been
the creation of states, along with markets and other monetary institutions.
Meanwhile he notes (2011:116) the ongoing difficulty of people combining
what they understand as sharing, implicitly in the distributed agency mode
of their own long-​term philosophies of life, with legally defined exchange
relations predicated on debt and repayment, over precise time frames
defined by the powerful.
Graeber’s book ends with allusion to the biblical institution of the
Jubilee, which is completely at variance with the modern-​ time debt
regime. On a strict cycle of seven times seven years, all debts would be
cancelled and all slaves freed. And social life would start over, rather than
allowing indebtedness to become cumulative and to turn into a regime of
ongoing life. In all the Abrahamic religions, the framing of lending on inter-
est, which necessarily enlarges the amount to be repaid at T2, was morally
prohibited among members of the same religious community (Leviticus
25:37), and failure to repay was not attributed to the singular agency of
the borrower. The lender is not to press the borrower for repayment if the
lender knows that the borrower does not have it (Exodus 22:24). There are
16 such framings for lending and borrowing among the classic mitzvot of
Judaism, including the injunction to lend to the poor and the destitute
(presumably beyond what is possible to supply through charity). The his-
tory and practice of “usury” is a long and vast one, but includes the implicit
assumption of distributed agency, in the sense that a debtor may not be
entirely responsible for his or her situation. The harvest failed; there was
cattle disease; a woman became a widow. The conversion of time into a pos-
sible monetary valuation was achieved by the School of Salamanca in the
late Middle Ages, and subsequently built on to create the mathematized
debt projections underlying the current debt regime. In his final paragraph,
Graeber takes in the import of this history and comparative anthropology,
to return us from the sophisticated intricacy of modern debt regimes, cre-
ated, he argues, through “math and violence,” to the “promise,” as a com-
mitment, implicitly in the multiple temporal frames of life of socially and
culturally based philosophies of distributed agency.
Clara Han’s book Life in Debt: Times of Care and Violence in Neoliberal Chile
(2012) takes up anthropology’s other classic method, participant observa-
tion within particular times, places, and communities. Following the end of
the Pinochet regime (1974–​1990), the new government is embracing both
a philosophy of moral indebtedness to the population for resilience and
resistance, and promoting a neoliberal policy that introduces new modes
of individuated credit and debt into the same people’s ordinary lives. Han’s

Di s t r i b u t ed Ag e n c y a n d   De b t   [ 105 ]
106

closest focus is on the process of living in this debt regime. In nuanced and
particular accounts, she shows how the temporality of debt increasingly
fails to inter-╉digitate smoothly with the other temporalities of life. Indeed,
its shaping power at T1, and its insistent threat of inflexibility at T2,
actively disorganize relationships that unfold on another durational ethic.
“The self is simultaneously enmeshed in different relations that entail dif-
ferent demands and desires” (Han 2012:20). “Monthly debt payments took
up half of Rodrigo’s income” (36), and bus fare to his job took up one-╉third
of his income (38). And “this ‘made time’ rubbed against the temporal-
ity of monthly debt payments and the uncertainty of unstable wages that
impinged on the home” (38). Regular commitments to children became dis-
rupted: “You need to buy gifts for children” (49); the children don’t under-
stand debt. In our own terms here, this would imply that accountability for
failure could be attributed by children to their parents, even when adults
would see the fault elsewhere. Clearly the conditions of an inflexible debt
version of T1 and T2 are incompatible with the temporalities of commit-
ment in the current economy and in other domains of durational ethics, as
understood by other participants in the nexus. The mutual implication of
intimate lives is a key theme of this ethnography of debt in a world where
different parties, different dimensions of life, different projections of what
counts as T1 and T2, are examined in detail.
These two works show clearly how the specific distribution of agency
is defined and shaped by different parties, living different temporalities
and different durational ethics, in relation to regimes of “modern time.”
Both call attention to temporal phenomena as the ground on which agency
and responsibility are built. As Augé argued for places, so also for tempo-
ralities:  “Modernity does not obliterate them but pushes them into the
background”: they “survive like the words that express them” (1995:77),
“intertwine[d]â•„and tangle[d]” together (1995:107). With respect to debt
regimes, the word “tangled” can be literally true rather than a meta-
phor: debts that are modeled in modernist time tie them in knots when
it comes to living with their own conditions and the durational ethics of
distributed agency and responsibility.

REFERENCES

Amato, Massimo, and Luca Fantacci. 2011. The End of Finance. Cambridge,
UK: Polity Press.
Augé, Marc. 1995. Non-╉Places: Introduction to an Anthropology of Supermodernity.
London: Verso.

[â•›106â•›]â•… Economy and Agency

  107

Benjamin, Walter. 1968. Illuminations. New York: Schocken Books.

Graeber, David. 2011. Debt: The First 5,000 Years. Brooklyn, NY: Melville House.
Guyer, Jane I. 2007. “Prophecy and the Near Future: Thoughts on Macroeconomic,
Evangelical and Punctuated Time.” American Ethnologist 34(3): 409–​421.
Guyer, Jane I. 2014. “Durational Ethics: Search, Finding, and Translation of
Fauconnet’s ‘Essay on Responsibility and Liberty.’” HAU: Journal of
Ethnographic Theory 4(1): 397–​340.
Han, Clara. 2012. Life in Debt: Times of Care and Violence in Neoliberal Chile.
Berkeley: University of California Press.
Lazzarato, Maurizio. 2012. The Making of the Indebted Man: An Essay on the Neo-​liberal
Condition. Amsterdam: Semiotext(e).
Peebles, Gustav. 2010. “The Anthropology of Credit and Debt.” Annual Review of
Anthropology 39: 225–​240.

Di s t r i b u t ed Ag e n c y a n d   De b t   [ 107 ]
  109

CHAPTER 12

Money as Token and Money as Record

in Distributed Accounts
BILL MAURER

M oney has always had agency, or, rather, has been an expression of
agency, or even is agency itself, or … well, this is precisely the prob-
lem of money.
People have certainly imagined money objects to possess agency. Who
among us has not thrown a coin into a fountain or picked one up imagin-
ing it would bring us luck, would change our destiny? Eighteenth-​century
English tracts on money, somewhere between didactic pamphlets and the
early serialized novel, imbued both notes and coins with the ability to wit-
ness the affairs of humans around them, and to remark on questions of
morality and social order as they were passed from hand to hand—​unless,
unfortunates, they ended up in someone’s safe, or misplaced somewhere.
Saved or lost money lost its powers of action and narration. Money in cir-
culation, even counterfeit, chronicled the world of men and women, boys
and girls, from every station of life and every corner of the world. And
seemed to enjoy doing so, too.
In telling their tales, these coins and notes reflected upon the economies
of which they were a part and which they were creating in their travels.
They served as a distributed, collective memory bank, to use Keith Hart’s
(2000) phrase, each note or coin testifying to its transit during human
exchange and, in the process, forming a collectivity having the potential
to record all of human intercourse—​or at least those aspects of it in which
money was quite literally brought to bear. Money was not inert but nor was
110

it completely fluid. It occupied a Goldilocks spot: just malleable enough to

be impressed with others’ stories, but not so malleable as to melt away. In
that respect, the memory bank made by these money tokens imagined in
these stories was not unlike the digital transaction records of contempo-
rary means of payment, from credit cards to PayPal or Bitcoin, electronic
means of recording transactions and human interrelation. They also bear a
family resemblance to the databases inscribed on clay tablets in the ancient
Sumerian temples and storehouses.
Like today, money in these narratives was brought to bear pretty much
everywhere. There was no escaping it. Almost every interaction involved it.
Money objects’ distributed agency recorded the stories of humanity, map-
ping relations in time, space, and value, and memorializing—​again, often
for pedagogical purposes—​the distributed action of humans, hapless or
otherwise.
What do these stories teach us today? What do they teach us as our mon-
etary relationships increasingly become an object of business concern—​
when Apple or Google wants to create and manage the record of all of our
interchange, to create a new hoard of our interpersonal transactional data?
Insofar as they remain didactic for contemporary readers, it may be
that these money narratives point up the fallacy that money is ever a
mere token, just a thing that represents something else or else is valuable
somehow in itself. It is about time we dispensed with this idea. The stuff
of money is a lot of fun, and does indeed matter, as will become clear. But
the stuff of money has also distracted monetary theorists and “money
nutters” (Maurer 2011) alike: both those who would seek to understand
money and those who would seek to solve any number of social, environ-
mental, or political problems by making money anew err in their attention
to its things absent its relations in time and space, the relations of debt
and credit that the token helps mark. Commodity money theorists, sound
money proponents, or survivalists preparing for the end of the world see
a money object essentially the same way as in the tales of its origination
in primitive barter (Graeber 2011). Money’s qualities—​rarity, portability,
malleability, divisibility—​and its qualities in themselves, not any agen-
cies they might possess, are for such people what makes it an object of
universal value.
Money’s matter matters in a different way, however.
When people write IOUs, relations of credibility extend outward from
person and paper, and can move across time and space often toward
unplanned or unexpected ends, outside whatever original intention may
have first animated them. When states issue currencies, they leverage their
own credibility in a financial and moral sense, but then let money go to

[ 110 ]  Economy and Agency

  111

do what it will, whatever people will do with it, from underground to dark
economies, to adornment and decoration and magic tricks. Private or pub-
lic credit, money is regardless a record of all manner of relationships of
credit and debt across time and space, and not just economic relationships
of credit and debt.
In arguing that money is a “means of collective memory,” a distributed
record of all human intercourse, Hart (2005) implicitly poses the question
of the relationship between those records and the infrastructures that pro-
duce and maintain them. It is not just cash circulating hand to hand that
constitutes this great database; it is also all the systems for transferring
money, recording those transfers, and creating great globally expansive
ledgers into which human collectivity finds a kind of rhetorical and mathe-
matical expression. This is the matter of money that matters, and the sense
in which that matter matters: it affords a carriage across space-​time and
between other agents. Hart wondered whether the infrastructures—​the
internet, the electronic payment card networks like Visa and MasterCard—​
could be seized for “human,” rather than simply corporate, purposes. Could
they be decentralized rather than held in walled gardens, shared with the
many whose everyday economic activity made them repositories of collec-
tive memory in the first place rather than enclosed for the profit of the few?
Early 21st-​century experiments with digital currencies like Bitcoin
respond to this challenge, despite the at times incoherent perspective on
money they evince (Maurer, Nelms, and Swartz 2013). The center of the
confusion is whether people understand, and then seek to create, money
as “cash-​based” or “ledger-​based.” For central banks, this is a primary
preoccupation, since money in circulation and money on the books must
be measured if the bank is to intervene in monetary affairs: cash money
has to be ledgered. Eduard de Jong (2014), a pioneer in the development
of digital currency systems—​and the toll road payment network for the
Netherlands!—​uses these terms to distinguish between moneys that are
emitted by a central source and then embedded in objects that leave no
record of their passage from one agent to another, and moneys that by
design record their passage in a database, the authority of which all accept.
For a toll road, you might want what de Jong calls a cash-​based system,
if you are worried about the prying eyes of the state into your travels, or
you just didn’t care about the massive amounts of data generated at the
toll booth about people’s comings and goings every day, day after day, on
their regular commutes. In such a system of value embedded in token,
however, people can pour too much power into the money object. Some
Bitcoin proponents thus evince a kind of digital metallism, imagining that
the value of the bits in Bitcoin derives wholly from their scarcity, a scarcity

M o n e y a s T ok e n a n d M o n e y a s   Re c or d  [ 111 ]
112

built into the design. With the latter, the ledger-​based system, people pon-
der whether the database must be centralized for it to have its authority.
Must it be maintained and verified by a trusted, or simply overwhelmingly
powerful, third party—​a temple, a corporation, a notary, a government?
Or can there be a democratically decentralized database, owned by none or
owned by all, without the intercession of any scribes, bookkeepers, banks,
or governments? Just how far can the distribution of agency go?
Bitcoin and other so-​called cryptocurrency experiments rely on the
existing network infrastructures of the internet and the electrical grid.
On top of those systems—​which are more or less centralized, depending
on how you look at it—​Bitcoin creates a database. This database has a spe-
cific structure. Imagine a great ledger book. Imagine further that every-
one—​everyone!—​has a copy of that ledger. Anyone can make an entry
of a debit or credit into that ledger. I credit you, while debiting myself.
But then everyone has the opportunity to verify that entry (though not
everyone does). Those who do the work of authenticating transactions
are rewarded with a new credit line in the ledger. They can also levy a fee,
in the form of an additional incremental credit, for their work of verifi-
cation. A verified set of transactions—​a complete ledger page—​is called
a block. The entire database is called the blockchain, a chain of groups
of verified transactions. A great database, distributed among all partici-
pants, public yet pseudonymous, written by the collective, collaborative
and competitive effort of the participants in the system. A Visa network
without Visa, banks, or the government. Just computers and people (and
electricity … and the internet). At least, this is the theory, anyway.
In one of the 18th-​century money narratives, money objects possess
distinct agencies, and multiple agencies within themselves, too, which the
money objects can send out of their bodies and into the world, and into
other bodies. In one story, a coin, an Indian gold rupee, describes itself
as having multiple spirits, though it names only three. They are Ductility,
Malleability, and Fusibility. “Men have foolishly called them qualities,” he
says (Scott 1782, in Bellamy 2012:36). But they are not qualities, in the
conventional sense of the term. They are different aspects or externaliza-
tions of its core self:  agents. They are also among the properties of pre-
cious metal that Enlightenment philosophers and, earlier, the Greeks had
imagined to be perfectly suited to serve as money. But the story suggests
that what matters is not their matter in itself but the agency the qualia
afford, to carry value across time and space and between persons. They are
also infrastructure, then—​something that transports or transmits other
things. Again, this is the sense in which the matter of money matters—​it

[ 112 ]  Economy and Agency

  113

bestows the ability to act over distances, but it does not possess by virtue
of those qualities any intrinsic value in itself.
The rupee is able to dispatch its spirits and send them into people’s
minds, or, more precisely, into their brains. There, the spirits can peer into
a person’s history by reading the literal inscriptions on a specific region
of the brain where memories are recorded. Or, not memories exactly: the
word our rupee uses is transactions:

One of my subordinate spirits immediately mounted his cella turcica [the struc-
ture in the skull holding the pituitary gland] by my command, from which spot
the brain above may be seen marked with impressions, like figures on a celestial
globe. These impressions are nothing but the scratches made by objects which
have been presented to the senses, and of which memory makes use in her oper-
ations. By reading these, we can discover all the transactions of any consequence
in which a man has been engaged. (Scott 1782, in Bellamy 2012:51)

We have qualities—​ductility, malleability, fusibility—​that are agents

that can read individual humans’ memories. Those memories are literally
written on the brain, like tally marks, at the seat of the human person’s own
agency, which is revealed by the money object to be itself a great database
of transactional records. This is a wondrous hall of mirrors: money objects,
distributed throughout all humanity, replicate at a higher level of scale the
transactional database in each person’s brain. Money’s agencies, embodied
in money’s qualities, intermediate between the great transactional data-
base of all human interaction and the personal archive of every individual’s
memories. This places money—​all of money, that is, all the money in the
world—​in the position of serving as the record of all of human history.
A memory bank, indeed!
It is difficult to think of a coin as a record-​keeping instrument. But this
is precisely what the rupee’s tale reveals.
The earliest coins were themselves a kind of receipt, the insignia on their
faces a constant reminder of the sovereign from which they were issued and
to which, through tribute or tax, they would return. Users of the first coins
sometimes stamped their own marks on them, announcing their presence
to the wider community of people among whom the coins circulated. It is as
if people still lived with a cuneiform mentality, the urge to record one’s own
symbol on coins so great that one numismatist, examining a collection in
the British Museum, noted that some examples were so covered with these
countermarks as to be losing their material integrity (Hill 1919).
In Sumer, ancient bureaucrats kept cuneiform tablets in central tem-
ples and palaces (Hudson 2000). Later, in Anatolia, coins were issued from

M o n e y a s T ok e n a n d M o n e y a s   Re c or d  [ 113 ]
114

mints and allowed to circulate. Agency centralized; agency dispersed. Yet

the latter was no less a giant record book than the former. Similarly, coin
and cash are no less a system of credits and debts and no less an infra-
structure or an archive for being materialized in what comes to our hands
as individual objects, individual tokens. Just a distributed archive. All the
money in circulation is a distributed database of all our credits and debts,
if we take the point of view of our Indian rupee.
If money’s qualities could extend outward from money objects to look
into personal trajectories, today electronic payment systems disaggre-
gate and distribute the agency of the person across spatial and temporal
scales, while opening up the person’s prior—​and, predictively, future!—​
transactions. If transactions in the 18th century were recorded as traces
on the brain, many of today’s transactions are recorded in other memory
banks: the computer servers that store the transactional data of these elec-
tronic systems.
With Bitcoin or with new electronic payment systems—​ ApplePay,
PayPal—​what falls outside of this archive are precisely the relations our
rupee was able to witness:  the hand-​to-​hand transfers of cash and coin.
From a digital computational point of view, these transactions are virtually
invisible, especially if there is no electronic receipt-​producing device like a
cash register or point-​of-​sale terminal linked to computers processing and
storing other data that may enter into the system. Corporate entities at
the time of this writing are actively seeking to enclose this “commons”—​a
term used by more than one payments industry professional in conversa-
tions with me.
Experts have predicted the obsolescence of physical instantiations of
money almost as soon as 19th-​and early 20th-​century governments stan-
dardized their issue. They have a point but miss their mark. Before there
was coin, there were centralized ledgers, records of transactions warrant-
ing other transactions and literally inscribing (in clay, stone, papyrus) the
distributed agencies of human interaction. Financial crisis and “disrup-
tion” in the payments industry—​the business of transferring value—​in
the early 21st century again raise the question of money-​as-​ledger, making
the era of cash and coin seem like a brief interregnum in the deep history
of value transfer. Asking after the infrastructures facilitating that transfer
leads to the role of accounting not as a record of monetary interaction, but
as that interaction itself. It is precisely a question of the distribution of
agency: who shall make entries into the great ledger of human transaction
and exchange? As the ledger pluralizes, who controls the cross-​referencing,
the gateways between newly dispersed accounts? Will it be the new

[ 114 ]  Economy and Agency

  115

corporate masters of multiple, independent or interdependent temples,

each holding its own record of our interactions (a twist on Peebles’s (2008)
story:  instead of the hoards migrating to the banks, they migrate to the
great server farms of our era)?
Or, as a joint commitment par excellence—​something we do together,
with it, over it, through it, because of those Goldilocks qualities—​can
money be made to redistribute agency, away from the Apples and Googles
and Facebooks, and toward another economy? Ledgers, after all, are vul-
nerable:  to fraud, to damage, to fire. Their intentional destruction can
have world-​changing consequences, from the burning of the Inkan khipu
by Spanish colonials to the burning of the tally sticks—​and the fire’s
spread through the House of Commons—​in early 19th-​century England.
If money’s agencies are increasingly being locked up now in corporate
ledgers, hoarded like a rupee in a safe, removed from the human inter-
action that once delighted it, can we help set them free so they can tell
their—​our—​stories?
Or shall they burn?

ACKNOWLEDGMENTS

I would like to thank Geoffrey Bowker for bringing the Indian rupee tale
to my attention. I would also like to thank Lana Swartz, Taylor C. Nelms,
and Scott Mainwaring for many productive hours and days of fieldwork,
conversation, writing, and thinking together. Mrinalini Tankha, Ursula
Dalinghaus, and Nathan Coben offered helpful criticism and advice, as did
Paul Kockelman. Research in the payments industry has been supported by
the US National Science Foundation (SES 0960423 and SES 1455859). The
opinions presented here are the author’s own and do not reflect those of
the National Science Foundation or any other organization.

REFERENCES

de Jong, Eduard. 2014. “Towards an Open E-​currency System.” Lecture given at

MoneyLab, Amsterdam, March, 23, 2014. Available at http://​networkcultures.
org/​moneylab/​2014/​03/​23/​edward-​de-​jong-​towards-​an-​open-​e-​currency-​
system/​, last accessed November 15, 2014.
Graeber, David. 2011. Debt: The First 5,000 Years. Brooklyn, NY: Melville House.
Hart, Keith. 2000. The Memory Bank: Money in an Unequal World. London: Profile.
Hart, Keith. 2005. “Notes toward an Anthropology of Money.” Kritikos: An
International and Interdisciplinary Journal of Postmodern Cultural Sound, Text and

M o n e y a s T ok e n a n d M o n e y a s   Re c or d  [ 115 ]
116

Image 2, available at http://​intertheory.org/​hart.htm, last accessed November

15, 2014.
Hill, G. F. 1919. “Notes on the Imperial Persian Coinage.” The Journal of Hellenic
Studies 39 116–​129.
Hudson, Michael. 2000. “Introduction: The Role of Accounting in Civilization’s
Economic Takeoff.” In Creating Economic Order: Record-​keeping, Standardization,
and the Development of Accounting in the Ancient Near East, Volume IV, edited by
Michael Hudson and Cornelia Wunsch, 1–​22. Bethesda, MD: CDL.
Maurer, Bill. July 2011. “Money Nutters: Economic Sociology_​The European Electronic
Newsletter 12(3): 5–​21.
Maurer, Bill, Taylor C. Nelms, and Lana Swartz. 2013. “When Perhaps the Real
Problem Is Money Itself!”: The Practical Materiality of Bitcoin.” Social Semiotics
23(2): 261–​277.
Peebles, Gustav. 2008. “Inverting the Panopticon: Money and the Nationalization of
the Future.” Public Culture 20(2): 233–​265.
Scott, Helenus. 2012. The Adventures of a Rupee. London: J. Murray, 1792. In The
British It-​Narratives, 1750–​1830, Volume 1: Money, edited by Liz Bellamy, 31–​
71. London: Pickering and Chatto.

[ 116 ]  Economy and Agency

  117

PART FIVE

Distributing Agency within Selves

and Species
  119

CHAPTER 13

Distribution of Agency across Body

and Self
RU TH H. PARRY

M any chapters in this book examine distribution and attribution of

agency across multiples of people, or people and artifacts. Here
though, I  look at attribution of agency within persons:  how agency, and
thus flexibility and accountability, can be distributed across body—​or body
parts—​and self. To do so, I draw on a corpus of video recordings of physi-
cal therapy treatment sessions, supplemented with brief examples from
recordings made in other settings—​so as to illustrate that the phenomena
I examine are not peculiar to physical therapy.
I demonstrate that people can construct three broad configurations:

1. The body (or body part) and the self are constructed as agents separate
to one another, disaggregated, capable of “independently”: acting one
upon the other.
2. The self and body (or body part) are constructed as connected: with the
self possessing the body or body part and having an associated control,
flexibility, and accountability for bodily actions.
3. The self and body are constructed as one and the same entity and
agent: an integrated, accountable, flexible individual.

I describe the seemingly innocuous and subtle interactional practices we

use to construct these quite radically different embodiments/​identities
120

(Kockelman 2007a). I will show that we can do so in reference to our own

self, or to other persons. The practices I describe and illustrate comprise
the following:

• Ways that body parts get referred to: whether, for example, we refer to
the thumb, or to your or my thumb.
• What we put in the subject and object position of our utterances:  for
example, whether I  say that I do something, or that my leg does
something.
• Eye gaze—​through which we can show ourselves to be talking to a part
of our interlocutor’s body—​and thus separating it out from a larger
whole. Conversely, we can show that we are talking to the “inner” or
whole person by looking him or her in the eye.
• Meanings (i.e., semantics) and metaphors that provide for different con-
figurations of the person, body, and self.

It’s worth briefly mentioning that the first practice is available in English
but not in all languages—​although it seems likely that in different lan-
guages, other practices are available to do the same kind of work. It is also
worth noticing that these configurations have a dualist character, some-
thing I will consider as I go along.
As well as describing how people manage to construct different body/​
self configurations, I examine when and why they do so. In the context
of this book, it should come as no surprise that this has to do with
how humans attribute and convey capacity and responsibility for their
actions, failings, and endeavors. In a final section, I outline some impli-
cations for how we understand and conceptualize distributed agency,
and consider how the dualist character of personhood so prevalent in
our talk makes sense in social terms. That is, whatever the material
discoveries of neuropsychology, and their echoes in linguistic under-
standings of inalienable possession (Kockelman 2007b), I  am going to
show you that when we talk and interact with one another, we do and
we can make self and body more, less, or completely—​for all practical
purposes—​separate.
First though, I need to attend to two terminology issues. Frustratingly
for me, English does not currently provide a word that signifies, exclu-
sively, the particular part or aspect of a person that is other than the per-
son’s body. Soul might once have been used, but has lost some of its value
in the context of current understandings that brain, consciousness, and
body are utterly intertwined in physical terms. In this piece though, I need

[ 120 ]  Distributing Agency within Selves and Species

╇ 121

to refer to this non-╉body aspect of a person, and to do so I use self. Also,

rather than the body or body part(s), hereon I use the shorthand body/╉part.

CONFIGURING BODY AND SELF AS UNIFIED, OR AS MORE

OR LESS SEPARATE

To put flesh on the bones of the three broad configurations, I examine a

series of cases drawn from a much larger set. What I say about the practices
and their patterning and functioning in these individual cases reflects anal-
ysis of 70 sequences of interaction, nearly all of which involved more than
one, often many more than one reference to body/╉part. Forty-╉five cases
were drawn from a dataset of 86 hours of video-╉recorded physiotherapy
treatment sessions, and 25 were collected from other recordings and tran-
scripts of everyday conversations and also other healthcare interactions.
All transcripts presented here are simplified versions of the more detailed
transcripts created during analysis.

Case 1: “Keeping you in that position”

In this example, a therapist (T) and a patient (P) who is receiving
rehabilitation for a stroke are talking about a particular goal of his
therapy: getting his back and pelvis area more flexible. In the course of
this, the therapist describes the physical problem:
Parry PD DT1_╉2.35
1 T: And what’s happening is your pelvis is tilting
2 forwards. So it’s keeping you in that position.
3 and then you’re quite rounded h[ere a]t the shoulders
4 P: [Mmm]

Using everyday, mundane language resources, the therapist moves rapidly

and seamlessly (lines 1–╉2) from (a) referring to your pelvis with its implication
of a possessive connection between patient’s self and body, to (b) formulating
the pelvis as a separate agent that is keeping the self of the patient in a partic-
ular posture. The second formulation entails referring to the pelvis with the
impersonal pronoun it, and doing so in the subject position, with you—╉the
self/╉patient, in the object position. This subject/╉object arrangement works
to convey the pelvis as an agent, and the self/╉patient as being acted upon by
it. In other words, the therapist constructs a part of the patient’s body as a
discrete, separate agent that acts upon the self of the patient.

Di s t r i b u t i o n of   A g e n c y a c r o s s   B o d y a n d   Se l f â•… [â•›121â•›]
122

As she continues her description (line 3), the therapist uses another
means to separate or distance the self of the patient and his pelvis. The
subject/​object arrangement is different, the patient—​you—​is now the sub-
ject and thus agent over the shoulders. The distancing here is done through
the pronoun the. In extracts to follow, I will demonstrate and examine how
people recurrently use impersonals such as the, this, or a when naming a
body part, and the way that this distances the body part from the self.
Why do people talk in ways that distance, disaggregate a self and their
body (or body parts)? Recurrently, as above, they do so when referring to
something that is a problem—​an undesirable physical state or movement.
By talking in a way that makes the body part a separate agent acting on the
self, or in a way that avoids using possessives such as my or your, one can
camouflage the connection—​sometimes referred to as “inalienable posses-
sion”—​between the person whose physical state is referred to and his or
her body/​part. One thereby implies that the person neither controls, nor
bears accountability, fault, and responsibility for the undesirable bodily
state or action. One can sever, so to speak, the connection—​physically as
much as psychologically.
The next case once again illustrates use of the rather than your, and also
introduces another way people can do this disaggregation: through their
gaze patterning.

Case 2: “Always the slow one”

A patient and therapist have been working together on a balance
exercise in which the patient kneels up on a mat, trying to keep herself
upright with straight hips and bottom tucked in. The patient repeatedly
sinks back and down because of hip weakness. The therapist is kneeling
behind the patient and encouraging her to squeeze her buttock muscles
to keep upright (as in line 1). As the patient begins to sink downward
once again (line 3), the therapist says “Go on an’ again this one especially.
There” while touching and tapping on the patient’s right hip and
buttock—​urging the patient to straighten up, and in particular to do so
by using her right buttock and hip.
Parry PD GT1_​26.49
1 T: That’s better. Go on squeeze squeeze lovely,
2 (4.4)
3 T: Good. Go on an’ again this one
4 especially. There ((gazing to P’s hips))

[ 122 ]  Distributing Agency within Selves and Species

  123

5 (1.5) ((T taps on P’s right hip, P continues to attempt

to keep upright, then shifts body position leaning
towards front of P))
6 T: This right one’s always the-​((gazing to P’s face))
7 P: This [one yes the sl]ow one
8 T:   [the slow one.]
9 T: Yea[h ]
10 P:    [This one is] (like) the slow one
11 T: Now see whether you can find your balance.

When giving instructions (lines 1, 3, 4), the therapist looks to the back of
the patient’s hips. She moves on to give a reason for her “this one especially”
instruction (lines 6, 8). By the time she starts this talk about a reason, she
has shifted her body enough to allow her to direct her gaze to the patient’s
face and make eye contact. Through this conduct, the therapist treats the
patient’s body and self as separate agents: the body as agent of the physical
action, and the self—​or consciousness—​as agent with regard to hearing
and understanding a reason for the action. This gaze patterning is recur-
rent in the physical therapy data (Parry 2013), with therapists using their
gaze and associated body positioning in a way that disaggregates patients’
body parts from their reasoning minds—​their selves. As noted, this case
also features use of an impersonal pronoun: the therapist talks of this right
one, not your right one (line 6). This brings us neatly to case 3, in which an
impersonal pronoun the, and the personal pronoun your, are used in rapid
succession.

Case 3: The thumb your thumb

This patient’s traumatic brain injury has left him with difficulties
controlling his left side. The therapist here has been moving and
stretching his affected left hand and thumb. As we join the session, the
therapist asks the patient to try to shake her hand with his left hand.
Parry PD KT3_​9.15
1 T: C’you shake my hand?
2 (3.7)
3 T: Pleased to meet you ((jokey tone))
4 = uh huh huh heh huh
5 (0.9) ((P lets go of T hand))

Di s t r i b u t i o n of   A g e n c y ac r o s s   B o d y a n d   Se l f   [ 123 ]
124

6 T: Now
7 (0.4)
8 T: ‘S not bad
9 (0.6)
10 T: ‘Part from, part from the thumb it’s fine Andrew
11 (0.2)
12 T: And can you take your hand away from my hand?
13 (1.9)
14 T: See if you can try again but making more
15 of a space for your thumb.
16 (2.4)
17 T: Yeah. That’s-​that feels-​that feels better to me.
18 P: I got the thumb splint.
19 (0.7)
20 T: Yeah, and whe-​how long are you wearing that for.

The therapist says the thumb (line 10)  when your thumb would be equally
grammatical, and in a context where it is perfectly clear which of his two
thumbs she is referring to. As in the other cases, she does this when referring
to a physical shortcoming. The thumb is, for all practical purposes, conveyed
as a separate entity at this point: the patient’s self and thumb are disaggre-
gated, and this reduces the implication that he is responsible or accountable
for the physical failure being talked of. This case also illustrates a counter-
part: when she moves into instructions the therapist says, can you take your
hand away and [make] more of a space for your thumb. By using your at this junc-
ture, she foregrounds the patient’s connection with—​in other words, his con-
trol over—​his hand and thumb. So now, the thumb is portrayed as under the
patient’s control.
In the physical therapy recordings, both therapists and patients fre-
quently use possessives in situations where controlled, purposeful action
is talked about and/​or encouraged. Using your or my (when alternatives are
possible) treats self and body as connected, and people do so when there is
good reason to emphasize someone’s control, responsibility, and physical
agency in relation to their body.
The rapidity with which one can move between different configurations of
the body/​self/​person relationship is particularly evident in this case, as is
the fact that in doing so, a speaker does not need to do anything extra as it
were—​one does not need to provide talk that, for instance, cancels out an
earlier meaning or gives some reason for the change.

[ 124 ]  Distributing Agency within Selves and Species

  125

Case 4: It collapsed; you were controlled

The final case from physical therapy illustrates particularly clearly how
people use different body/​self configurations to convey just who or what
is responsible for a physical success or a physical failure. The patient,
whose stroke has resulted in left-​sided weakness, is standing with feet
slightly apart. The therapist is encouraging her to shift her weight from
one foot to the other. A key point of the exercise is for the patient to try
to keep her weaker left leg and knee straight as she shifts her weight to
this weaker left side.
Parry PD QT2_​20.17
1 T: Okay that’s-​that was better that time
2 because the first time it just kind
3 of collapsed ((smiley, gentle tone))
4 (0.9)
5 T: That time you were more controlled.
6 (0.2)
7 T: Now, as you come back over towards me I want you to …

At line 2, the therapist’s it refers to the patient’s left knee and leg. Her talk
ascribes this as the entity that collapsed. When she describes the recent
successful controlled performance of the exercise, the patient and her body
are construed as a single unified agent: you. The therapist is thereby able to
praise the patient and treat her as responsible for the success.
A series of shorter cases now supplement the detailed analysis above by
demonstrating that not just therapists but also patients construct different
body/​self configurations via the same resources; also that the same kind
of phenomena arise in other healthcare settings and in informal conversa-
tions between friends.

Case 5: My brain doesn’t trust the left side

Near the start of a treatment session, a patient who suffered a stroke two
weeks previously talks to a therapist about his difficulty in standing up
from a sitting position:
Parry Doc HT3_​17.03
1 P: Still my brain doesn’t trust? the-​the er
2 t[he le]ft er [tha]t’s why-​that’s why
3 T:  [yeah] [yeah]

Di s t r i b u t i o n of   A g e n c y ac r o s s   B o d y a n d   Se l f   [ 125 ]
126

4 P: every time I stand up? I-​I put most of

5 the-​my weight on my le-​=on my right.
6 (0.2)
7 P: Because my brain doesn’t trust the left side

Here the patient makes my brain the subject of his talk, and his left side the
object. This subject/​object arrangement has parallels to case 1, in which
the therapist’s talk constructed the patient’s pelvis as an agent that was
making the patient rounded. There are differences though: rather than self
and body part, the patient disaggregates two body parts:  brain and left
side. Nevertheless, it is reasonable to hear him treating my brain as in some
sense his controlling self, and the left side as separate to this brain/​self.

Case 6: I’ve got a foot

This case is from a different healthcare setting: a consultation with a
primary care doctor. The extract comes from the start of a consultation.
Robinson and Heritage 2005 page 487: Extract 11: lines 1-​5 (simplified)
1 D: What you up to?
2 (.)
3 P: I’ve gotta bad foot that I can’t-​get well.
4 (0.2) ((Doc moves his chair close to foot))
5 D: Which part.

Here, the patient uses both an impersonal pronoun, a, and a verb phrase
that I can’t get well. These work to disaggregate her self, as an agent, from
her foot—​upon which she has been attempting to act.

Case 7: Those big toenails

With this case, we move to everyday conversations. Two friends, Emma
(E) and Lottie are talking on the telephone. We join the conversation
as Emma describes, in dramatic terms, a fungal infection affecting her
toenails.
NB IV:10R (original transcript line 1472-​6) (simplified)
1 E: Oh God it’s terrible Lottie my toenails
2 they’re jis look so sick those big toenails it jis
3 u-​makes me sick. Yihknow they’re diss (.) u-​dead.
4 Everything’s dead I d-​I sat out (.) tihday en
5 I said my God em I jis (.) DYING it’s (.) like
6 I’m ossified.

[ 126 ]  Distributing Agency within Selves and Species

  127

Emma switches from the personal my toenails (line 1), to the imper-
sonal those (line 2) in the context of escalating from “terrible” to nau-
seating. She also moves from describing a situation—​it’s terrible—​to
conveying what she sees as she looks at those big toenails. The imper-
sonal pronoun those adds to the conveyed sense of an external view. By
line 6, she reaches a further configuration: her whole self is conveyed
as ossified.

case 8: physically, not mentally

Another everyday conversation case comes from a telephone call
between friends. Ava (A) has called Bee (B), and at the point we
enter the call, Bee has been checking with Ava where she
is calling from.
TG Ava and Bee, page 15 Line 09 (simplified)
1 B: I finally said something right. (0.2)
2 You are home. Hmm
3 A: Yeh-​I believe so. [Physically anyway
4 B: [mmh
5 A: Yea –​a-​h Not mentally though ((smiley, slightly laughing
voice))

Within a jokey exchange, Ava makes a dualist claim—​spacially separat-

ing her mental and physical self via semantic resources (the meanings
of her words) as opposed to the grammatical or gaze resources we saw
above.

Case 9: My body didn’t need it

Similarly, in this next case, in an everyday conversation between Sherry
(S), Carol (C) and Ruth (R), we see semantics in use to configure the
physical body and self as separate.
SN-​4_​05_​lines 01-​06 (simplified)
1 S: Hi Carol
2 C: H[i]‌
3 R:   [Caro]l, Hi
4 S: You didn’t get an icecream sandwich
5 C: I know, I decided that my body didn’t need it
6 S: Yes but ours did heh-​heh-​heh

Di s t r i b u t i o n of   A g e n c y ac r o s s   B o d y a n d   Se l f   [ 127 ]
128

CASE 10: SOMETHING BEHIND HER

This final case comes from a telephone conversation between friends
Marsha and Erma. Erma (E) is describing her son’s girlfriend Ilene:
MDE:MTRAC:60-╉1:5 (simplified)
1 E: Cause she’s gotta very pretty face she ez gorgeous
2 blue eyes, and she really pays very little attention
3 to it and I kind’v am not suprized when Joey says
4 that she wants to go into Social Services, she
5 seems like that kind. Y’know w’t I mean she seems
6 like she has a lot (.) there’s something behind her.

Erma builds a sense of connection but also a distinction between the inner,
private Ilene (behind her—╉line 6)  and Ilene’s physical aspect—╉gorgeous
eyes, and a pretty face. This sense is built both through the design of her
description—╉Ilene has physical attributes, and the semantic and meta-
phoric use of prepositions to indicate this character of selfhood.

CONCLUSIONS

People can allocate to the body, or parts thereof, agency that is more or less
separate from the agency of the self. In our social lives at least, while the
body and self can be a unified and fused entity, the body and self can also
be disaggregated, fissioned, divided. That is, in our interactions, we commit
to a form of dualism. This is not a strict “here is body, there is mind” type
of dualism, but rather a continuum that can be cut in a variety of places.
We commit to this form of dualism as a way of understanding ourselves,
our personhood. This enables us to manipulate—╉to shape—╉the meaning
of our own, and other people’s, physical successes, failures, and features.
As I have shown above, these kinds of allocation of agency are accom-
plished via rather basic design features of body movement and talk, includ-
ing gaze, and grammatical and semantic resources. Through the meanings
we thereby build, we can encourage others to hear matters of fault, respon-
sibility, endeavor, and achievement with regard the physical as more or less
attributable to the body or body part, or to the “true,” “inner” self (the
soul, or even the brain). It is noticeable that the more distancing configu-
rations are noticeably charitable—╉we might even say compassionate; the
person talking can absolve the person/╉self being talked about. Being able
to allocate responsibility and even blame for physical problems to the body

[â•›128â•›]â•… Distributing Agency within Selves and Species

╇ 129

(and therefore as less—╉or even not at all—╉the actual person’s fault) helps
us to limit negative emotional and relationship consequences of criticizing
or blaming another’s physical state or behavior. Conversely, being able to
convey that positive performances and attributes are integral to the person
him-╉or herself allows us to emphasize that person’s personal competence,
and thus praise, motivate, and encourage him or her, facilitating align-
ment in terms of both local actions and the ongoing sense of relationship
between the people involved.
My analysis has shown that people locally constitute agency and rela-
tions between selves and bodies in a moment to moment way that is fitted
to social circumstances and also highly flexible. Previous research that has
shown that the way one refers to another person is not random; rather, it
does work—╉it performs social actions—╉such as conveying another person
as more or less close to the speaker (Enfield and Stivers 2007; Stivers 2007).
My analysis shows that the way we refer to our bodies and our selves is also
not a random or haphazard matter, but a socially shaped means of working
up distinct embodied identities fitted to interpersonal circumstances and
relationships. Finally, the evidence I have presented shows that in order to
understand fully the distribution of agency, we need not only to recognize
the ways and means through which it can be distributed, fissioned, and
fused “across and among individuals” (Enfield 2013:104), but also to add
the recognition that people also distribute agency within individuals.

REFERENCES

Enfield, N. J. 2013. Relationship Thinking: Enchrony, Agency, and Human Sociality. New
York: Oxford University Press.
Enfield, N. J., and T. Stivers. 2007. Person Reference in Interaction: Linguistic, Cultural,
and Social Perspectives. Cambridge: Cambridge University Press.
Kockelman, P., 2007a. “Agency.” Current Anthropology 48(3): 375–╉401.
Kockelman, P. 2007b. “Inalienable Possession and Personhood in a Q’eqchi’-╉Mayan
Community.” Language in Society 36(3): 343–╉369.
Parry, Ruth H. 2013. “Giving Reasons for Doing Something Now or at Some Other
Time.” Research on Language & Social Interaction 46(2): 105–╉124.
Stivers, T. 2007. “Alternative Recognitionals in Person Reference.” In Person Reference
in Interaction—╉Linguistic, Cultural and Social Perspectives, edited by N. J. Enfield
and T. Stivers, 73–╉96. Cambridge: Cambridge University Press.

Di s t r i b u t i o n of   A g e n c y a c r o s s   B o d y a n d   Se l f â•… [â•›129â•›]
  131

CHAPTER 14

Distributed Agency in Ants

PATRIZIA D ’ET TORRE

A gency in humans is characterized by a distributed nature; we have an

extraordinary capacity to share agency via multilevel cooperation but
also via coercion and different forms of social exploitation. Agency as expe-
rienced in human societies is unique—​or is it?
I argue that distributed agency is an inherent consequence of social life
and it is proportional to the degree of social complexity. Many animals live
in groups that may be structured at different levels, ranging from simple
aggregations to small family groups, up to cooperative breeding societies,
such as those of meerkats with clearly defined helper roles (Clutton-​Brock
et al. 2001). However, some invertebrates, the social insects, have reached
the apex of social organization: eusociality, characterized by reproductive
division of labor, cooperative care of young, and overlapping generations
(for a clear and complete definition of eusociality, see Boomsma 2013).
With this social structure, social insects (all ants and termites; some spe-
cies of bees and wasps) have become tremendously successful in colonizing
almost all terrestrial habitats. In tropical forests, social insects represent
around 30% of the total animal biomass and 80% of the insect biomass.
Ants, with more than 14,000 described species, have evolved a multi-
tude of life histories and adaptations to a variety of environments. Ant
societies show several similarities with human societies (Wilson 2012).
Ants build cities, maintain infant nurseries and cemeteries, and construct
elaborate nests, which can contain hundreds of rooms connected by a com-
plex network of galleries, such as those of harvester ants (Tschinkel 2004).
132

Ants and other social insects have evolved sophisticated techniques of

waste management (e.g., in leaf-​cutting ants; Bot et al. 2001) and practice
social prophylaxis by acquiring immunity via social contact with infected
individuals (e.g., in the garden ant; Ugelvig and Cremer 2007). Ants, like
humans, store food: for instance, harvester ants collect seed in the warm
season and put them in long-​term stores to be consumed later (Johnson
2001). Ants practice agriculture and animal husbandry. Some species of
ants cultivate fungi for food. These fungus-​growing ants have evolved an
obligatory association with fungal symbionts, which they provide with
growing substrate and protection from pathogens and parasites (Poulsen
and Boomsma 2005). Many ant species keep aphids as we keep cows and
regularly “milk” them for their honeydew secretion (Ivens et  al. 2012).
In exchange, ants protect the aphids against predators, such as ladybird
beetles, and give them shelter. Ants conduct wars and slave-​raiding. They
are aggressive against same-​species competitors and engage in prolonged
battles and tournaments (Hölldobler and Wilson 1990). Some ant species
are specialized social parasites: they raid other ant colonies, pillage their
brood, and bring it to their own colony. Upon reaching adulthood, these
“slaves” will work for the parasitic colony as if they were in their original
mother colony (d’Ettorre and Heinze 2001).
The highly specialized division of labor among individual ants, with the
evolution of morphological differentiations (queen and worker caste, and
in some cases different types of workers), has given rise to the analogy
between the ant colony and the body of a Metazoan animal. The ant colony
is considered as an individual of a higher order, a “superorganism” with
unique emergent properties (Hölldobler and Wilson 2008). Recently, the
term superorganism has been questioned based on the observation that
many other organisms originated as groups of genetically distinct mem-
bers, including eukaryotes, and are not called superorganisms. A broad defi-
nition of social organism has been proposed: “a unit of high cooperation and
very low conflict among its parts” (Strassmann and Queller 2010), which
includes eusocial insects. The central concept remains that of an entity
made of different specialized components and characterized by the cooper-
ative partitioning of work. Like prokaryotic cells joined together in a sym-
biotic union, giving rise to eukaryotes, solitary individuals started working
together into social units (colonies), which then evolved nonreproductive
castes, resulting in advanced division of labor.
Unlike most human societies, ants are perfectly able to live in a condition
of anarchistic socialism without the need of a leader (Wheeler 1910). In the
absence of any centralized order, the individual ant workers behave collec-
tively, following self-​organized processes to solve, for instance, a foraging

[ 132 ]  Distributing Agency within Selves and Species

  133

task. Therefore, the behavior of the ant colony is an emergent property

resulting from the dynamics of feedback loops between interacting ant
workers, each following simple decision rules. A typical example is the
selection of the shortest route to a food source. Ants are central-​place for-
agers and they come back to the nest once they have found a suitable food
source. In many ant species, nestmates are recruited to recently discovered
food sources by a trail pheromone, the Ariadne’s chemical thread between
the nest and the food. Small ants no longer than one or two centimeters
can cover considerable distance (a hundred meters) during their foraging
trips in an environment that is heterogeneous and full of obstacles. Indeed,
they have evolved strategies to select the shortest and most efficient path
from the nest to the food. Laboratory experiments with the Argentine ant,
Linepithema humile (formerly Iridomyrmex humilis), have been among the
first to elucidate the mechanisms underlying this collective strategy and to
model the ant foraging behavior (Gross et al. 1989). In one of these experi-
ments, the nest and the foraging arena are linked by a bridge with two
branches of different lengths in a way that each ant must choose between
the shorter or the longer branch two times when going from the nest to the
food, and back (Figure 14.1). Argentine ants mark the substrate with their

(a) (b)
Food Food

Nest Nest

Figure 14.1  A colony of Argentine ants selecting the branches of a bridge between the nest
and the food area: (a) four minutes after the placement of the bridge ants choose branches
randomly; (b)  eight minutes after the placement of the bridge ants choose the shortest
route between the nest and the food.
(Modified from Gross et al. 1989)

Di s t r i b u t ed Ag e n c y i n   A n t s   [ 133 ]
134

trail pheromone both when leaving and when returning to the nest. At the
beginning of the experiment, explorer ants use both branches randomly
(Figure 14.1a); however, after a few minutes, additional ants are recruited
to the food source and the shortest route is suddenly selected by the vast
majority of ants (Figure 14.1b).
How is this possible? Each ant leaves a pheromone mark both ways (from
and to the nest); ants that take the shortest route arrive earlier to the food
and come back earlier to the nest, compared to ants that take the longest
route. Therefore, it takes less time to the shortest branch to be marked by
an increasing amount of pheromone left by the ants. The chemical signal
on the shortest route will be amplified quicker and, as a consequence, more
and more ants will be recruited on this shortest route. It is a simple positive
feedback mechanism based on an indirect communication between nest-
mate ants (via the trail pheromone), which results in an intelligent collective
decision. This is a relatively flexible, context-​dependent, adaptive solution.
In such cooperative tasks based on self-​organized processes, the agency
is totally distributed and gives rise to an emergent phenomenon: the sud-
den collective choice of the shortest path from the nest to the food. In this
process, each individual ant has very little flexibility, particularly because the
response to pheromones is typically “hardwired” (so-​called innate). An ant
has no other choice than responding to the highest concentration of pher-
omone, and therefore she will follow the path that has been marked more
heavily by her nestmates. This lack of flexibility results in little accountabil-
ity. It might well be that there is individual variation among the various ants,
with some individuals being, for instance, less sensitive to the trail phero-
mone and thus not able to perceive the difference in concentration between
the two trails when this difference is still small. These individuals may choose
the long path when the majority of ants go for the short path. Nevertheless,
the ants on the short path will continue being faster in coming back to the
nest, and soon the pheromone trail will reach a concentration that will be
attractive to every ant. In a cooperative collective task, it does not matter
if few individuals do not participate: the system is automatically buffered.
Indeed, there are usually more ants coming to a food source than the number
required to bring that particular amount of food back to the nest.
In a social group characterized by extreme cooperation, agency is distrib-
uted beyond the individuals (the single units forming the group) and there
is a very low individual level of flexibility and accountability. Ants, and other
social insects, are a paramount example of distributed agency because they
pushed the division of labor to its extreme form: the partitioning of repro-
duction. In many ant species, such as the Argentine ant, workers are totally
sterile and the only way they can pass on their genes to the next generation

[ 134 ]  Distributing Agency within Selves and Species

  135

(new colonies) is by assisting their mother queen to reproduce. Therefore,

it is in the workers’ best interest to work for the good of the colony (even
when this involves committing suicide: e.g., a honey bee worker stinging a
vertebrate); workers do not have any other option.
Yet, ant societies are not always behaving as harmonious cooperative
units. In some species, workers—​which are all females—​retained the abil-
ity to produce eggs. As in other hymenoptera, ants have a haplodiploid
sex-​determination system:  males develop from unfertilized eggs and are
haploid, while females develop from fertilized eggs and are diploid. This
means that even if workers cannot mate, they can lay male eggs and thus
achieve direct reproduction. A  worker is always more related to her own
sons than to the sons of her mother (the worker’s brothers); therefore a
worker should prefer laying eggs instead of caring for her brothers. Indeed,
some ant workers do so. Worker reproduction reduces overall colony effi-
ciency because workers that reproduce do not work efficiently. This con-
flict over male production is resolved by a collective mechanism named
“worker policing”: workers preventing each other from reproducing by eat-
ing worker-​laid eggs or by aggressively suppressing reproductive workers.
Worker policing by egg eating was first observed in the honey bee (Ratnieks
1988) and then in several species of bees, ants, and wasps. We showed that
in colonies of the ant Pachycondyla inversa some workers specialize in polic-
ing behavior (van Zweden et al. 2007). Workers in this species present the
typical ant polyethism, with specialization in tasks such as foraging, nest
defense, and brood care, but these societies evolved also a worker role spe-
cialized in tackling social corruption: the policing force (Figure 14.2). These

(a) (b)
100
90
80
Observed
Number of workers

70
60
50
Police force
40 Expected
30
20
10
0
0 1 2 3 4 5 6 7 8 9 10 11 12 13 14
Number of aggressive policing acts per worker

Figure 14.2  (a) A worker of the ant species Pachycondyla inversa holding eggs.

(Photo: Mauro Patricelli)
(b) Frequency distribution of aggressive policing acts among workers. The observed sharing
of policing tasks is significantly more skewed than would be expected from random chance.
(Modified from van Zweden et al. 2007)

Di s t r i b u t ed Ag e n c y i n   A n t s   [ 135 ]
136

policing workers act for the collective interest of the workers as a whole
(the colony) and they are not moved by selfish benefits since none of the
policing workers showed developed ovaries, suggesting that they do not
kill other workers’ eggs or attack reproductive workers (so that they cannot
reproduce) to be able to reproduce themselves.
This ant example shows that when individual units are allowed some
degree of flexibility (e.g., the possibility to develop ovaries), they might
attempt to act in their best selfish interest (e.g., by reproducing), but they
are held accountable for their behavior and therefore punished by other
members of the community. When the risk of destabilization is high—╉if
many workers reproduce, the colony as a whole would collapse—╉societies
should evolve specialized units (e.g., the ant policing force) to restore and
maintain social cooperation. Coercion, defined as a social pressure in the
form of punishment of group members to prevent them from acting self-
ishly and thereby harming the group as a whole, has been described in a
variety of species, including humans (review in Ratnieks and Wenseelers
2008). Indeed, levels of cooperation across human societies are correlated
with the degree to which uncooperative individuals are punished (Henrich
et al. 2006), suggesting that coercion plays a key role in the evolution and
maintenance of cooperative systems.
In conclusion, I argue that when individual units of a social group are
prevented from acting selfishly by drastically reducing their flexibility,
agency becomes better distributed and individual units cannot be consid-
ered accountable for their behavior. On the other hand, when some flex-
ibility is retained in a cooperative system, individual units may pursue their
direct benefits and cause a “tragedy of the commons.” Cooperative systems
that have survived natural selection are likely those that have found effec-
tive strategies to make individuals accountable for their behavior, thus
enforcing cooperation. The more sophisticated the social organization, the
higher the level of cooperation, the wealthier the common good, and the
more specialized the retaliation system against free riders should be. Some
ant species have solved the problem by evolving individual units that are
totally sterile and act as cells in a body. As a consequence, flexibility was
obliterated, giving rise to completely distributed agency.

ACKNOWLEDGMENTS

Many thanks to the editors N.  J. Enfield and P.  Kockelman for having
invited me to join this project and for their insightful comments on a pre-
vious version of the manuscript.

[â•›136â•›]â•… Distributing Agency within Selves and Species

╇ 137

REFERENCES

Bot, A. N. M., C. R. Currie, A. G. Hart, and J. J. Boomsma. 2001. “Waste Management

in Leaf-╉Cutting Ants.” Ethology Ecology & Evolution 13: 225–╉237.
Boomsma, J. J. 2013. “Beyond promiscuity: mate-choice commitments in social
breeding.” Philosophical Transactions of the Royal Society B: Biological Sciences
368 (1613): 20120050.
Clutton-╉Brock, T. H., A. F. Russell, L. L. Sharpe, P. N. M. Brotherton, G. M.
McIlrath, S. White, and E. Z. Cameron. 2001. “Effects of Helpers on Juvenile
Development and Survival in Meerkats.” Science 293: 2446–╉2449.
d’Ettorre, P., and J. Heinze. 2001. “Sociobiology of Slave-╉Making Ant.” Acta Ethologica
3: 67–╉82.
Gross, S., S. Aron, J. L. Deneubourd, and J. M. Pasteels. 1989. “Self-╉Organized
Shortcuts in the Argentine Ant.” Naturwissenschaften 76: 579–╉581.
Henrich, J., R. McElreath, A. Barr, J. Ensimger, C. Barrett, A. Bolyanatz, J. C.
Cardenas, M. Gurven, E. Gwako, N. Henrich, C. Lesorogol, F. Marlowe, D.
Tracer, and J. Ziker. 2006. “Costly Punishment across Human Societies.”
Science 312: 1767–╉1770.
Hölldobler, B., and E. O. Wilson. 1990. The Ants. Cambridge, MA: Harvard
University Press.
Hölldobler, B., and E. O. Wilson. 2008. The Superorganism: The Beauty, Elegance, and
Strangeness of Insect Societies. New York: W. W. Norton.
Ivens, A. B. F., D. J. C. Kronauer, I. Pen, F. J. Weissing, and J. J. Boomsma. 2012. “Ants
Farm Subterranean Aphids Mostly in Clone Groups—╉An Example of Prudent
Husbandry for Carbohydrates and Proteins?” BMC Evolutionary Biology 12: 106.
Johnson, R. A. 2001. “Biogeography and Community Structure of North American
Seed Harvester Ants.” Annual Review of Entomology 46: 1–╉29.
Poulsen, M., and J. J. Boomsma. 2005. “Mutualistic Fungi Control Crop Diversity in
Fungus Growing Ants.” Science 307: 741–╉744.
Ratnieks, F. L. W. 1988. “Reproductive Harmony via Mutual Policing by Workers in
Eusocial Hymenoptera.” American Naturalist 132: 217–╉236.
Ratnieks, F. L. W., and T. Wenseleers. 2008. “Altruism in Insect Societies and
Beyond: Voluntary or Enforced?” Trends in Ecology and Evolution 23: 45–╉52.
Strassmann, J. E., and D. C. Queller. 2010. “The Social Organism: Congresses, Parties,
Committees.” Evolution 64: 605–╉616.
Tschinkel, W. R. 2004. “The Nest Architecture of the Florida Harvester Ant,
Pogonomyrmex badius.” Journal of Insect Science 4: 21.
Ugelvig, L. V., and S. Cremer. 2007. “Social Prophylaxis: Group Interaction Promotes
Collective Immunity in Ant Colonies.” Current Biology 17: 1967–╉1971.
van Zweden, J. S., M. A. Fürst, J. Heinze, and P. d’Ettorre. 2007. “Specialization in
Policing Behaviour among Workers of the Ant Pachycondyla inversa.” Proceedings
of the Royal Society of London B 274: 1421–╉1428.
Wheeler, W. M. 1910. Ants: Their Structure, Development and Behavior.
New York: Columbia University Press.
Wilson, E. O. 2012. The Social Conquest of Earth. New York: Liveright.

Di s t r i b u t ed Ag e n c y i n   A n t s â•… [â•›137â•›]
  139

PART SIX

Social Bonding Through

Embodied Agency
  141

CHAPTER 15

Group Exercise and Social Bonding

EMMA COHEN

INTRODUCTION

Group exercise, broadly construed, is a cultural universal. From the cere-

monial rituals and persistence hunting of foraging bands to the beach festi-
vals, team sports, and urban marathons of large-​scale communities around
the world, people everywhere get together to move together. The ubiquity
of coordinated, physical, exertive movement in everyday social activity
points to its significance not only for individual physical and mental health,
but for social well-​being also (Davis, Taylor and Cohen, 2015). Anecdote,
theory, and evidence indicate that when people coordinate in group move-
ment and exercise, powerful and unique psychological and physiological
effects are generated over and above those that occur when individuals
perform the same activities alone. For the theologian Michael Novak, team
spirit is no misnomer:  “For those who have participated on a team that
has known the click of communality, the experience is unforgettable, like
that of having attained, for a while at least, a higher level of existence”
(in Murphy and White 2011:15–​16). Such tightly coordinated teamwork,
characteristic of many sports, dissolves boundaries between individual
and social agency: “In rowing … it feels like you have at your command
the power of everybody else in the boat. You are exponentially magnified.
What was a strain before becomes easier. It is absolutely the ultimate team
sport.”1 What is this “click of communality”? Are there special psychological
142

and physiological effects of exercising with others as compared to doing it

alone? Is individual agency magnified in group exercise?
This chapter draws on research in evolutionary anthropology, sports
and exercise physiology, clinical neuroscience, and social psychology to
begin to address these broad questions and to develop an account of the
reciprocal causal links between group movement and social bonding. The
account comprises two interconnected claims: group movement and exer-
cise forges social bonds between participants; social bonds, in turn, alter,
and potentially heighten, participants’ motivations and performance abili-
ties. The claim that exercise increases bondedness among participants is
developed in the first two sections, which focus specifically on the social
effects of exercise highs (“social high”) and coordinated behaviour (“social
sync”). The third and final sections explore how social context influences
exercise participation and performance. Specifically, we consider uniquely
human motivations to collaborate in joint activities (“social drive”) and
capacities to benefit from social support in physically and psychologically
challenging conditions (“social boost”).

SOCIAL HIGH

Although prolonged aerobic exertive movement is typically associated

with high levels of physical stress and pain, people also frequently report
responses to exercise that include improved affect, sense of well-╉being, anx-
iety reduction, and post-╉exercise calm. Extremes of these exercise-╉related
changes are popularly known as “runner’s high.” The effect is not limited to
running or sports, however; anthropologists’ descriptions of ceremonial
rituals frequently describe dancers achieving similar euphoric states (e.g.,
Radcliffe-╉Brown 1948).
The neuropharmacological mechanisms that make these activities
pleasurable appear to involve two core neurobiological systems—╉the opi-
oidergic system and the endocannabinoid system. The first is our body’s
natural morphine; the second is its natural cannabis. In addition to their
role in pain modulation, these systems produce feelings of euphoria, tran-
scendence, and pleasure. Boecker et. al. (2008) first tested the exercise-╉
opioid link on endurance runners via a measure of endorphin levels in
the brain and a self-╉report mood measure. They found positive and corre-
lated effects for post-╉exercise endorphin activity and self-╉reported eupho-
ria. Similarly, recent tests of the endocannabinoid hypothesis in runners
and cyclists found significantly increased concentrations of anandamide,
an endocannabinoid previously associated with aerobic activity in other

[â•›142â•›]â•… Social Bonding Through Embodied Agency

  143

cursorial species, in participants’ blood after moderately intense exercise

(e.g., Raichlen et. al. 2013).
Our endogenous pain relief and reward systems, although ancient and
involved in many different bodily functions, were likely recruited in human
evolution to reduce pain associated with the physical exertion of run-
ning and to enhance motivation. Endurance running may have played an
important role in human evolution, facilitating the exploitation of protein-​
rich resources via hunting, chasing, and scavenging. Consistent with
this hypothesis, evidence suggests that endocannabinoid-​based pleasure
induced by exercise is an ancient trait that evolved to encourage aerobic
activity (e.g., Bramble and Lieberman 2004; Raichlen et. al. 2012). Insofar
as persistence hunting typically occurred within and benefited from a coop-
erative and cohesive group context, a tight association between endurance
aerobic exercise, analgesic and euphoric mental states induced by endocan-
nabinoid and endorphin activity, and social bonding may be hypothesized.
Evidence supporting these claims includes findings that, compared with
solo exercise, group exercise enhances activity in pain and reward systems.
A series of studies with rowers and runners using side-​by-​side ergometers
and treadmills found that participants tolerate significantly greater pain
following group synchronous workouts than after solo or group non-​syn-
chronous workouts of the same basic form, intensity, and duration (e.g.,
Cohen et. al. 2010; Sullivan et. al. 2014). This suggests that there is greater
endorphin and endocannabinoid activity among participants who train
together in synchrony than among those who train alone. A separate body
of research has implicated endorphins and endocannabinoids in social
interaction and bonding across a range of contexts (e.g., primate grooming,
social play, drug-​taking activity). Taken together, these findings support
the claim that forms of group exercise that effectively activate opioidergic
and endocannabinoid rewards may have particularly powerful effects on
individual mood, well-​being, and group camaraderie and cohesion via what
might be called an exercise-​induced “social high.”

SOCIAL SYNC

Sports and exercise in group settings typically entail the tight coordina-
tion of bodies and minds in time in space, both intra-​and interpersonally.
Coordinated, synchronous behavior with others in ritual, dance, singing,
and drill has long been thought to enhance social solidarity (e.g., Durkheim
2012), and similar effects may be integral to the most intimate human
activities (e.g., in mother-​infant interaction, lovemaking). Experimental

G r o u p E x e r c i s e a n d S o c i a l B o n di n g   [ 143 ]
144

evidence shows that synchronous marching, singing, and even simple fin-
ger tapping significantly increase cohesion, liking, affiliation, perceived
similarity, and generosity among participants. In a seminal study on syn-
chrony and cooperation, Wiltermuth and Heath (2009) found that people
who marched and sang with one another under controlled experimental
conditions cooperated more in subsequent economic exercises than those
who participated in similar, non-​synchronous activities.
This interpersonal synchrony effect is thought to derive in part from the
tight attentional union that occurs between individuals when they match
the content and timing of their actions. By performing the same action at
the same time, interpersonal similarity is enhanced and self-​other bound-
aries become blurred in action processing and recall. In a recent study by
Miles et al. (2010), pairs of participants were instructed to sit opposite one
another and perform arm-​curling movements while alternately pronounc-
ing aloud a series of country names that were played through personal
headphones. They were then individually asked to identify who had said
which country names. When participants had performed the same arm-​
curl movements in synchrony with one another, their recall accuracy for
who had said what was significantly poorer than when they had performed
the same movements in anti-​phase synchrony (i.e., when participants
raised and lowered arms in opposite alternating order). The self-​memory
advantage that strangers normally exhibit in such tasks was eliminated
via a simple and subtle behavioral synchrony intervention that appears to
have shifted participants’ attentional focus from self to other. This effect
on memory was similar to that commonly observed among people who
have long-​standing interpersonal connections.
Self-​other blurring may be central to the “magnifying” effect described
by the team rower in the anecdote at the beginning of this chapter. In expe-
rienced crews, all oars touch and exit the water at the same precise instant.
Multiple bodies—​legs, backs, arms, and wrists—​bend in perfect unison
to generate a singular force propelling the boat forward. In this moment,
crew members’ relative efforts and effects on the boat they all occupy are
individually indistinguishable. The resultant blurring of self-​other bound-
aries potentially generates a physical-​proprioceptive illusion that one’s
own effort and output are exponentially magnified. This sensation, in turn,
appears to give rise to unique psychological states variously described
by those who have experienced them as “pure pleasure” and “exultation”
(Brown 2013: 161, 269).
It is unlikely that all group exercise activities achieve these psycho-
logical effects to the same degree, or that they are associated with such
intense psychological reward. Indeed, long and hard hours of training are

[ 144 ]  Social Bonding Through Embodied Agency

  145

often required to perfect a team’s capacity for coordination. Across many

group movement contexts, however, the combination of attentional union
and self-​other blurring among participants, together with the magnifying
effects of jointly coordinated physical effort, may account in part for the
“click of communality” experienced by participants. Culture exploits this
basic package of features across many different social interaction contexts.
Dance, song, music, play, sports, and ritual have been the focus of consider-
able social scientific speculation about the possible role that synchronous
movement has in promoting social solidarity and cooperation. Mounting
evidence from the behavioral and psychological sciences now indicates a
powerful causal association between synchronous group activity and social
cohesion.

SOCIAL DRIVE

Motivations that underpin participation in sports and group exercise are

highly variable across activities, individuals, ages, and cultures. Evidence
reviewed below suggests that, irrespective of the many differences, the
common social context of group exercise activities is a compelling factor in
participants’ motivation and pleasure.
From an evolutionary perspective, selection favored those of our ances-
tors who were best able to coordinate with others—​that is, those who had
the skills and motivations to coordinate with others and to identify and
recruit other suitable coordinators. Evidence from developmental, social,
and comparative psychology suggests that species-​specific forms of human
sociality depend upon a complex set of social-​cognitive mechanisms and
processes. These include joint attention abilities as well as abilities to infer
others’ intentions and goals, share common knowledge, detect cheaters,
evaluate fairness, make decisions about punishment, and prefer in-​group
over out-​group members. Successful social interaction permits a wide
range of fitness-​enhancing behaviors such as group hunting, collective
child-​rearing, warfare, and exchange.
Our unique sociality is not all down to cognitive skill, however. Humans
have unique social motivations also. We are motivated to collaborate, coop-
erate, and coordinate together across many different interaction contexts.
The possibility that collaborative forms of sociality, including group exer-
cise, music, religious ritual, and sports, are intrinsically rewarding is con-
sistent with a selectionist account of human social and cognitive evolution.
Intrinsic motivations to collaborate and cooperate together may originate
in evolutionary advantages reaped not only in the material domain, such

G r o u p E x e r c i s e a n d S o c i a l B o n di n g   [ 145 ]
146

as subsistence and defense, but from successful reputation management

and strong social attachments in increasingly complex and interdependent
group settings. Observational and experimental studies have shown that
young children engage in coordinated, collaborative activity with others
even when this offers no immediate material advantage over acting alone.
In contrast, chimpanzees engage others in collaborative activities only
when there are obvious advantages to doing so (e.g., obtaining more food
than would otherwise be possible). This suggests that interspecific differ-
ences in sociality may to some extent reflect differences in social motiva-
tions rather than divergent social-╉cognitive skills (Melis and Tomasello
2013). These social motivations are exhibited at a basic level in a general
and pervasive attraction to connectedness and aversion to loneliness. For
humans, doing things together offers intrinsic psychological rewards—╉in
other words, the winning, or pleasure, is in the taking part.
The importance of social motivations and pleasure for sustaining exer-
cise motivation and performance is well established in sports psychol-
ogy research. Despite the cross-╉cultural ubiquity of exercise and its many
known benefits, physical inactivity is an important contributor to ill health
across the developed world. Many people take up exercise for health rea-
sons, but the bulk of evidence suggests that, when it comes to motivating
and sustaining participation, education and informed decision-╉making are
no match to intrinsic affective pleasure (Ekkekakis et al. 2011). Further, in
a meta-╉analysis that included 87 studies and almost 50,000 participants,
Carron and colleagues found that, from an adherence perspective, exercis-
ing with others was superior to exercising alone. More specifically, motiva-
tion and pleasure in exercise settings is significantly predicted by group
cohesion and intragroup similarity (Burke et  al. 2008). Uniquely human
social motivations potentially link these two findings—╉exercise in cohe-
sive social settings sustains pleasure and, therefore, participation. Cohen
et. al.’s (2010) study with rowers further suggests that social context can
also positively impact performance, reducing perceptions of pain and giv-
ing a higher high. In the following section, we consider in more detail how
the support of a cohesive group might come to have such powerful effects
on perceived effort and pain.

SOCIAL BOOST

Participants in sports and exercise often experience pain, stress, and physi-
cal discomfort in the pursuit of their goals. In many competitive sports, the
person or team who can go harder, faster, and longer, conquering fatigue

[â•›146â•›]â•… Social Bonding Through Embodied Agency

  147

and intense pain, ultimately wins the prize. Groundbreaking neuroscien-

tific and psychological research is currently illuminating our understanding
of the ways in which social context influences pain perception and activ-
ity in pain-​related neural circuitry. This work has potential relevance for
understanding how and why social context influences the stamina, endur-
ance, and performance outcomes of individuals and teams.
Studies on placebo analgesia—​pain relief through treatment that has
no active ingredient—​have revealed significant effects of belief alone on
reported intensity of pain and on neural processes that mediate pain.
Placebo analgesic effects are known to operate through expectations based
on, for example, beliefs about efficacy or previous experience of the treat-
ment. There is increasing evidence that these beliefs may be influenced
via both implicit and explicit cognitive processes. Humphrey and Skoyles
(2012) have suggested that evolved mechanisms governing conscious and
unconscious calculations about the costs and benefits of recovery under
variable environmental conditions may be responsible for the placebo
response in analgesia and in immune system defense more broadly.
Cognitive neuroscientific research into placebo analgesia has mostly
taken place within clinical contexts in which, for example, experts explicitly
inform patients of the efficacy of a particular medical intervention, thereby
influencing patients’ expectations about the possibility of relief. Bingel et.
al. (2011) investigated responses of 22 participants to pain stimuli (using
controlled heat) and pain relief (Remifentanil administered via controlled
infusion pump). At baseline, participants’ average reported pain intensity
was 66 out of 100 (where 0 means “no pain” and 100 means “unbearable
pain”). Ratings dropped to 55 when a controlled dose of Remifentanil was
administered without participants’ knowledge. Participants were then
informed that they would receive a dose of powerful pain reliever; the aver-
age pain rating fell further to 39. They were then told that the pain relief was
being withdrawn, though in reality it was not; scores returned to baseline
levels (64). Positive expectations therefore effectively doubled the effect of
Remifentanil, while negative expectations abolished it. These differences in
self-​reported pain ratings were reflected in changes in unpleasantness and
anxiety ratings and activation in core areas of the cerebral pain network.
Recent social neuroscience research outside of a strictly clinical context has
shown that merely holding a photo of an attachment figure (e.g., romantic
partner, as compared with photo of a stranger) can have similar effects on
pain perception and associated neural activity (Eisenberger et al. 2011).
In conditions of distress and pain, the relief, comfort, and feelings of
security that the presence of loved ones can bring is well known. The stud-
ies described above have established that social context influences not just

G r o u p E x e r c i s e a n d S o c i a l B o n di n g   [ 147 ]
148

the perception of pain, but its underlying neurochemistry and neurophysi-

ology, including opioidergic and endocannabinoid activity. This indicates a
possible role for such “social placebo” effects in group exercise. According
to the social placebo account, the presence of close others, or possibly
even mere cues to their presence (e.g., photographic images, team colors)
or to social cohesion (e.g., behavioral synchrony), can significantly influ-
ence these factors for success, reducing pain and pain perception, fatigue,
unpleasantness and anxiety, thereby enhancing performance and reward.
Future research on the role of implicit and explicit social placebo holds
great promise for understanding the links between social cohesion and
pain, performance, and motivation in group exercise. More generally, there
is enormous scope to investigate the range of social triggers for placebo
analgesia.

CONCLUSION

Humans’ combined physical, cognitive, and social capacities for coordi-

nated and exertive movement are unique. They are thought to have served
important survival functions in human evolution, for example, in hunting,
travel and communication, and defense. For most people today, however,
the substantive rewards of group exercise are less obvious and immediate.
Modern technologies have largely dissociated coordinated group exercise
from everyday subsistence and communications. Yet, despite the huge
emotional and energetic costs that group exercise often incurs, it remains
a pervasive feature of human culture. Why is sustained, coordinated aero-
bic activity such a recurrent feature of everyday life, cross-╉cutting cultural
domains as varied as religion, sport, music, and play?
The independent lines of research reviewed above suggest a powerful
role for physical exercise in individual well-╉being and social cohesion and,
in turn, a powerful role for social context in exercise psychology and per-
formance. Together, they substantiate sporting anecdote and social theory
alike, elucidating how physical exercise can promote camaraderie among
participants, how the perceived boundaries between the physical actions
of self and other can become blurred in perception and memory, how the
self’s power to act can be (and be perceived to be) thrillingly magnified by
social collaborators, context, and cues, and how the efforts and output of
a cohesive group of individuals can be significantly greater than the sum
of their parts. In “the ultimate team sport”—╉in other words, euphoric, in
sync, cooperative, and supportive—╉the normal physical, biological, and

[â•›148â•›]â•… Social Bonding Through Embodied Agency

╇ 149

psychological limits of individual agency are dramatically lifted. This, per-

haps, is the agency of the team’s spirit.

NOTE

1.
http://╉usatoday30.usatoday.com/╉educate/╉college/╉careers/╉hottopic7.htm,
accessed October 11, 2014.

REFERENCES

Bingel, U., V. Wanigasekera, K. Wiech, R. N. Mhuircheartaigh, M. C. Lee, M.

Ploner, and I. Tracey. 2011. “The Effect of Treatment Expectation on Drug
Efficacy: Imaging the Analgesic Benefit of the Opioid Remifentanil.” Science
Translational Medicine 3(70): 70ra14.
Boecker, H., T. Sprenger, M. E. Spilker, G. Henriksen, M. Koppenhoefer, K. J. Wagner,
M. Valet, A. Berthele, & T. R. Toll. 2008. “The Runner’s High: Opioidergic
Mechanisms in the Human Brain.” Cerebral Cortex 18(11): 2523–╉2531.
Bramble, D. M., and D. E. Lieberman. 2004. “Endurance Running and the Evolution of
Homo.” Nature 432(7015): 345–╉352.
Brown, D. J. 2013. The Boys in the Boat. London: Pan MacMillan.
Burke, S. M., A. V. Carron, and K. M. Shapcott. 2008. “Cohesion in Exercise
Groups: An Overview.” International Review of Sport and Exercise Psychology
1(2): 107–╉123.
Cohen, E. E. A., R. Ejsmond-╉Frey, N. Knight, and R. Dunbar. 2010. “Rowers’
High: Behavioural Synchrony Is Correlated with Elevated Pain Thresholds.”
Biology Letters 6(1): 106–╉108. doi: 10.1098/╉rsbl.2009.0670.
Davis, A., J. Taylor, and E. Cohen. 2015. “Social Bonds and Exercise: Evidence for a
Reciprocal Relationship.” PLoS ONE 10(8): e0136705. http://╉doi.org/╉10.1371/╉
journal.pone.0136705.s012.
Durkheim, E. (1912) 2012. The Elementary Forms of the Religious Life. Mineola,
NY: Courier Dover Publications.
Eisenberger, N. I., S. L. Master, T. K. Inagaki, S. E. Taylor, D. Shirinyan, M. D.
Lieberman, and B. D. Naliboff. 2011. “Attachment Figures Activate a
Safety Signal-╉Related Neural Region and Reduce Pain Experience.” PNAS
108(28): 11721–╉11726.
Ekkekakis, P., G. Parfitt, and S. Petruzzello, S. J. 2011. “The Pleasure and Displeasure
People Feel When They Exercise at Different Intensities.” Sports Medicine
41(8): 641–╉671.
Humphrey, N., and J. Skoyles. 2012. “The Evolutionary Psychology of
Healing: A Human Success Story.” Current Biology 22(17): R695–╉R698.
Melis, A. P., and M. Tomasello. 2013. “Chimpanzees’ (Pan troglodytes) Strategic
Helping in a Collaborative Task.” Biology Letters 9(2): 20130009.
Miles, L. K., L. K. Nind, Z. Henderson, and C. N. Macrae. 2010. “Moving
Memories: Behavioral Synchrony and Memory for Self and Others.” Journal of
Experimental Social Psychology 46(2): 457–╉460.

G r o u p E x e r c i s e a n d S o c i a l B o n di n g â•… [â•›149â•›]
150

Murphy, M., and R. A. White. 2011. In the Zone: Transcendent Experience in Sports.

New York: Open Road Media.
Radcliffe-​Brown, A. R. 1948. The Andaman Islanders. Glencoe, IL: Free Press.
Raichlen, D. A., A. D. Foster, G. L. Gerdeman, A. Seillier, and A. Giuffrida. 2012.
“Wired to Run: Exercise-​Induced Endocannabinoid Signaling in Humans and
Cursorial Mammals with Implications for the ‘Runner’s High.’” The Journal of
Experimental Biology 215(8): 1331–​1336.
Raichlen, D. A., A. D. Foster, A. Seillier, A. Giuffrida, and G. L. Gerdeman. 2013.
“Exercise-​Induced Endocannabinoid Signaling Is Modulated by Intensity.”
European Journal of Applied Physiology 113(4): 869–​875.
Sullivan, P. J., K. Rickers, and K. L. Gammage. 2014. “The Effect of Different Phases
of Synchrony on Pain Threshold.” Group Dynamics: Theory, Research, and Practice
18(2): 122–​128.
Wiltermuth, S., and C. Heath. 2009. “Synchrony and Cooperation.” Psychological
Science 20(1): 1–​5.

[ 150 ]  Social Bonding Through Embodied Agency

  151

CHAPTER 16

Social Bonding Through Dance

and ‘Musiking’
BRONW YN TARR

INTRODUCTION: SOCIAL GROUPS

From a biological perspective, successful interpersonal interactions are

critical in the establishment and maintenance of social groups. As a gregar-
ious species, much of our survival depends upon these social groups; more
cohesive and tightly bonded networks are well coordinated, and group-​
level selection predicts that members of these groups are more likely to
survive than those acting alone or in less bonded units (Dunbar and Shultz
2010). In humans, experimentally manipulated social support influences a
range of physiological measures such as blood pressure, cortisol levels, and
heart rate, and social isolation is correlated with reduced life expectancy
(for a review see Thorsteinsson and James 1999).
Primate social groups (particularly those of anthropoid primates) are
characterized by a large number of interpersonal relationships between
non-​kin and non-​sexual partners. These “social bonds” (feelings of, and
behaviors signaling, closeness and affiliation) are frequently termed “friend-
ships,” and form the backbone of our social networks. Large, complex, and
cohesive social groups place a high cognitive load on interacting members;
each agent needs to establish interpersonal connections and keep track
of ongoing interactive histories with other agents in their social network
(Dunbar 1998). These cognitive constraints associated with sociality con-
tribute toward a species-​specific upper limit to group size, a relationship
152

that is reflected by covariance between mean group size and neocortical

volume (relative to the remaining volume of the brain; Dunbar 1998).
We have particularly large social networks (approximately 150 individu-
als) compared to our primate cousins, who establish and maintain their
smaller networks predominantly via reciprocal allo-​grooming (reviewed
in Dunbar 1998). Grooming is subject to time constraints:  it is typically
a dyadic activity, and the number of individuals one can groom in a day
is limited by the amount of time that can be spared from other activities,
such as foraging and sleeping (Dunbar 1998). In the case of large human
groups, this time investment is unfeasible. Our extensive networks require
bonding behaviors that can involve multiple agents building affiliation
simultaneously, or “grooming at a distance” (Dunbar 2012).

DANCE AS A MEANS OF “GROOMING AT A DISTANCE”

A number of different behaviors are likely to have helped humans estab-

lish social bonds with multiple individuals simultaneously, and to maintain
those bonds with relatively little time investment. One hypothesis is that
rhythmic bodily movement and synchronization may have played (and
continue to play) an important role in contributing to group cohesion and
social bonding. This has been termed the “social bonding hypothesis” of
dance, although it is generalizable to other activities involving music (e.g.,
Koelsch 2014).
To our knowledge, there is no human culture that does not possess some
form of group music-​making and dance. Clearly the stylistic and socio-​
contextual details of these activities are by no means homogenous across
cultures, making definitions of these activities challenging. Nevertheless,
underpinning this diversity is the curious fact that the production of
organized, expressive sound (broadly “music”) and the coordination of
our movement to music of some form and with others (broadly “dance”),
are prominent features in ritual, religion, and human expression gener-
ally (Ehrenreich 2008; Tomlinson 2015). Arguably, these activities are
fundamentally cooperative in nature; not only do they demand complex
sensorimotor coordination, but we tend to do these activities with oth-
ers. Music-​based group activities demand some degree of coordination of
actions and cooperation in terms of establishing a joint goal and shared
intention (Koelsch 2014). The social bonding hypothesis suggests that
ritualized group dance and music activities foster social contact, social
cognition, homogenous emotions, communication, coordination, coopera-
tion, and group cohesion (Koelsch 2014). Furthermore, as we can engage in

[ 152 ]  Social Bonding Through Embodied Agency

╇ 153

these activities simultaneously with many individuals, music-╉based activi-

ties could provide an effective means of bonding on a large scale.
Evidence that social bonding arises from, and also is signaled by, various
music-╉based activities has been provided by, inter alia, ethnographers, his-
torians, ethnomusicologists, biomusicologists, and experimental research-
ers. Empirical studies have focused on the fact that music-╉based activities
generally involve participants synchronizing their movements to a rhyth-
mic beat (usually provided by music), and to one another.

SYNCHRONY AND SOCIAL BONDING

Experiments involving simple finger tapping, walking, and rocking in

chairs have demonstrated that engaging in coordinated, specifically syn-
chronized, activities increases interpersonal closeness. Compared to asyn-
chronous or solo conditions, performing movements in synchrony leads to
increased trust between co-╉actors, liking each other more, being more will-
ing to help each other, and feeling similar in personality (literature reviewed
in more detail in Tarr et al. 2014). These positive effects on social bonding
are evident when individuals perform synchronized movements in small
groups of three or four and when the synchronized action involves singing
and/╉or moving (Wiltermuth and Heath 2009) and dancing with others to
music (Tarr et al. 2015). Furthermore, the capacity for synchronized tasks
to increase interpersonal social bonding is evident from a young age: chil-
dren are more likely to display helping behavior after moving and singing
in synchrony than those in a non-╉musical control condition (Kirschner and
Tomasello 2010).
These studies demonstrate that performing similar movements in a pre-
dictable, rhythmic fashion leads to individuals feeling socially bonded. This
phenomenon is thought to be based on action-╉perception networks facili-
tating a blurring of the sense of “self” and “other” during synchronized
action (e.g., Demos et al. 2012), which leads to a social bond between actors.
Additionally, there is growing evidence that synchronized action is associ-
ated with various neuropharmacological mechanisms (see Â�chapter  15 by
Cohen in this volume). For example, synchronized rowing (see Â�chapter 15)
and dancing in synchrony (Tarr et al. 2015) elevates pain threshold, a com-
mon proxy measure for endorphins due to the close relationship between
opioids and perception of pain. Given the role of endorphins in social
bonding across a range of other mammals, it is plausible that opioids play
a role in underpinning the social closeness that arises during synchronized
bonding activities such as dance (Tarr et al. 2014).

S o c i a l B o n di n g T h r o u g h Da n c e a n d ‘ M u s i k i n g ’ â•… [â•›153â•›]
154

In addition to synchrony, a number of other important factors contrib-

ute toward increased interpersonal affiliation when dancing together. For
example, shared attention with co-╉actors, working toward a joint goal and
improved mood after successful co-╉engagement, all contribute to increas-
ing feelings of closeness. These elements do not arise exclusively in music-╉
based activities, and are clearly evident in a range of human behaviors such
as team sports (see Â�chapter 15). What sets dance apart from other simi-
larly complex, coordinative, and exertive human activities is the fact that it
involves responding directly and purposefully to music.

HOW MUSIC MOVES AND BONDS US

Music adds to the social bonding effects that can arise in situations of group
coordination in various ways. First, people spontaneously synchronize to
music (Janata et  al. 2012), and as a consequence music encourages and
facilitates occurrences of synchrony. People naturally entrain to a rhythmic
beat embedded in music, and are capable of doing so from a very young age
(discussed in more detail in Koelsch 2014). Music can therefore establish a
shared and predictable, rhythmic scaffolding, thereby facilitating synchro-
nization of timing between multiple individuals. In large groups, it is dif-
ficult to coordinate synchrony and observe the movements of all the other
participants simultaneously, arguably making “self-╉other” merging a less
likely prospect. Rhythmic music can facilitate synchrony for large numbers
of people, perhaps by providing a central “other” to which each participant
can direct their attention (e.g., drill marching), thereby leading to the social
closeness described above.
Indeed, synchrony with music itself appears to improve interpersonal
affiliation. Demos et al. (2012) found that participants who performed a
chair-╉rocking task to music reported feeling more connected to each other
than those who performed the task without background music. This inter-
personal connectedness was predicted by synchrony with the music, and
not by synchrony with one another. Accordingly this study suggests that
synchrony to music is sufficient to cause bonding effects, and synchrony
between people may be irrelevant when they are engaging with a shared
external source of rhythm as provided by the music.
But the effects of music go beyond its capacity for facilitating synchro-
nous movement. In addition to the predictability provided by a rhythmic
beat, listening to music induces strong emotional and pleasurable effects.
These arise via core brain regions generally involved in processing affect
(including areas that are responsible for releasing endorphins), and those

[â•›154â•›]â•… Social Bonding Through Embodied Agency

╇ 155

associated with reward and motivation (Koelsch 2014). When we engage

with shared music alongside others, this can homogenize participants’
emotional states (Koelsch 2014), which could theoretically play an impor-
tant role in increasing interpersonal cohesion and decreasing social conflict.
Aside from moving us emotionally, music literally moves us. According
to PET scans, intense pleasurable responses experienced when merely lis-
tening to music (and not moving physically) are associated with activation
in areas of the brain involved in movement control (Blood and Zatorre,
2001). More often than not, listening to music does induce some form of
actual physical movement, be it simple head bobbing, foot tapping, or fin-
ger tapping in time with the rhythmic beat. At a neurological level, certain
elements of music are particularly good at inducing a motor response, for
example certain tempos. At a higher level, the ability of music to induce
body movement and dance has been described as “groove” (e.g., Janata
et  al. 2012), and certain musical styles with particular musical elements
appear better at inducing groove. The close relationship between music
and our often spontaneous and unconscious urge to move is likely a con-
sequence of a tightly coupled and ancient relationship between movement
and music, originating from a time when our engagement with music was
directly linked to a performance of agency.

“MUSIKING”: AN EXPRESSION OF AGENCY

Our immense aptitude and appreciation for sound and movement as a

means to express, communicate, and connect has probably been part of
humanity for a very long time. The oldest identified instruments (bone
“flutes”) are dated to approximately 50,000–╉35,000 years ago (for refer-
ences see Kirschner and Tomasello 2010), but arguably our first instru-
ments were our voice and body. When music first originated, our bodies
likely acted as a moving, percussive instrument (Cummins 2009). Indeed,
at the most fundamental level, it is not possible to create sound without
vibration, and, by extension, the production of music requires movement.
Consequently, movement is involved when we use any instrument to pro-
duce sound, but also when we attach bells, shakers, heels, and so on to
the body to accentuate the production of sound associated with each step
or movement. When our bodily movements contribute to musical stimu-
lus (for example when we have bells attached to our stomping feet) we are
exercising agency over the production of the music. Clapping, snapping,
and stomping while dancing is evident in many cultures today, with both
audiences and performers using their bodies to create, or at least add to,

S o c i a l B o n di n g T h r o u g h Da n c e a n d ‘ M u s i k i n g ’ â•… [â•›155â•›]
156

the musical stimuli. Indeed, in many cultures, music and dance are consid-
ered indistinguishable and some languages do not have separate words to
describe them (Blacking 1995).
This close link between movement and music fits with Christopher
Small’s term “musiking,” defined as “[taking] part, in any capacity, in a
musical performance, whether by performing, by listening, by rehearsing
or practicing, by providing material for performance (what is called com-
posing), or by dancing” (Small 1998:9). Given that dance most likely co-​
evolved with music as a means of generating rhythm and sound (Cummins,
2009), it is understandable that music tends to make us want to move,
and specifically synchronize, with the musical beat. In matching our move-
ments in time to a musical rhythm, we are simulating the timing of move-
ments that could have conceivably contributed toward the production of
that rhythmic beat.
The ability to have agency over the production of musical sounds
through movement has been shown to have significant physiological and
psychological effects. For example, varying the degree of musical agency for
people exercising in a group can influence their ability to perform strenu-
ous exercise and their mood. In a high agency condition, exercise machines
were linked to musical output software such that a musical sound resulted
from each movement, and consequently individuals “created” music as
they exerted themselves (Fritz et al. 2013). This experiment (and others
by the same authors) demonstrated that when movement (during group
exercise) results in musical feedback, participants perceived exertion to be
lower, reported enhanced mood, and felt a greater desire to exert them-
selves further, in comparison to when they were exercising while listening
(passively) to independently provided music not connected to their move-
ments (Fritz et al. 2013). According to these results, musical agency (i.e.,
executing purposeful movement that results in musical sounds) feels good
and improves the capacity to withstand strenuous exercise.
Although the exercise scenario described above is clearly a contrived case
of musical agency, the findings likely reflect the fact that we are wired to
engage with music in a context in which we do in fact have agency over the
production of sound. In reality, unless we are a member of a band or play
a musical instrument, we more frequently engage passively with music,
which rarely involves a context in which our movements directly contrib-
ute to the music. We frequently enjoy prerecorded music, which involves
a decoupling of our engagement with that music from the social context
and movement evident at its source. Under these modern circumstances,
and in the absence of having any of our music-​induced movements actually
result in some direct contribution to the music, perhaps when we move

[ 156 ]  Social Bonding Through Embodied Agency

╇ 157

to music we are attempting to pretend we possess some agency over the

production of that music. The more tightly we couple our movement with
the music, for example, the better we synchronize our movement with the
beats, the more convincing is this simulated agency. In this manner, even
if we are not actually making the music to which we dance, our synchroni-
zation with the musical beat simulates movement required to explain, or
physically justify, that musical beat.

CONCLUSION

Humans’ social agency occurs in the context of large, interconnected

social networks that are established and maintained by a variety of group
activities, such as dance. Our tendency to move in time to music, and the
ubiquity of music-╉based activities such as dance are important aspects of
what makes us human. Dance and synchronizing to music facilitates inter-
personal closeness, providing a means to build and maintain large social
groups. Furthermore, the way we engage with music is likely embedded
in an ancient history, in which movement, sound, and agency were inter-
twined. Based on this account, music-╉making and dance, which are con-
ceptualized here as a single, coupled concept—╉“musiking”—╉provides an
invaluable, naturalistic empirical and theoretical platform for understand-
ing foundations of our social agency.

REFERENCES

Blacking, J. 1995. Music, Culture, & Experience: Selected Papers of John Blacking.

Chicago/╉London: University of Chicago Press.
Blood, A. J., and R. J. Zatorre. 2001. “Intensely Pleasurable Responses to Music
Correlate with Activity in Brain Regions Implicated in Reward and Emotion.”
Proceedings of the National Academy of Sciences of the United States of America
98: 11818–╉11823. doi:10.1073/╉pnas.191355898.
Cummins, F. 2009. “Rhythm as an Affordance for the Entrainment of Movement.”
Phonetica 66: 15–╉28. doi:10.1159/╉000208928.
Demos, A. P., R. Chaffin, K. T. Begosh, J. R. Daniels, and K. L. Marsh. 2012. “Rocking
to the Beat: Effects of Music and Partner’s Movements on Spontaneous
Interpersonal Coordination.” Journal of Experimental Psychology: General
141: 49–╉53. doi:10.1037/╉a0023843.
Dunbar, R. I. M. 1998. “The Social Brain Hypothesis.” Evolutionary Anthropology
6: 178–╉190. doi:10.1080/╉03014460902960289.
Dunbar, R. I. M. 2012. “On the Evolutionary Function of Song and Dance.” In Music,
Language and Human Evolution, edited by N. Bannan. 201–╉214. Oxford: Oxford
University Press.

S o c i a l B o n di n g T h r o u g h Da n c e a n d ‘ M u s i k i n g ’ â•… [â•›157â•›]
158

Dunbar, R. I. M., and S. Shultz. 2010. “Bondedness and Sociality.” Behaviour

147: 775–​803. doi:10.1163/​000579510X501151.
Ehrenreich, B. 2008. Dancing in the Streets: A History of Collective Joy.
New York: Metropolitan.
Fritz, T. H., J. Halfpaap, S. Grahl, A. Kirkland, and A. Villringer. 2013. “Musical
Feedback during Exercise Machine Workout Enhances Mood.” Frontiers in
Psychology 4: 921. doi:10.3389/​fpsyg.2013.00921.
Janata, P., S. T. Tomic, and J. M. Haberman. 2012. “Sensorimotor Coupling in Music
and the Psychology of the Groove.” Journal of Experimental Psychology 141: 54–​
75. doi:10.1037/​a0024208.
Kirschner, S., and M. Tomasello. 2010. “Joint Music Making Promotes Prosocial
Behavior in 4-​Year-​Old Children.” Evolution and Human Behavior 31: 354–​364.
doi:10.1016/​j.evolhumbehav.2010.04.004.
Koelsch, S. 2014. “Brain Correlates of Music-​Evoked Emotions.” Nature Reviews
Neuroscience 15: 170–​180. doi:10.1038/​nrn3666.
Small, C. 1998. Musicking: The Meaning of Performing and Listening. Hanover,
NH: Wesleyan/​University Press of New England.
Tarr, B., J. Launay, and R. I. M. Dunbar. 2014. “Music and Social Bonding: ‘Self-​Other’
Merging and Neurohormonal Mechanisms.” Frontiers in Psychology: Auditory
Cognitive Neuroscience 5: 1–​10. doi:10.3389/​fpsyg.2014.01096.
Tarr, B., J. Launay, E. Cohen, and R. I. M. Dunbar. 2015. “Synchrony and Exertion
during Dance Independently Raise Pain Threshold and Encourage Social
Bonding.” Biology Letters 11: 1–​3. doi:10.1098/​rsbl.2015.0767.
Thorsteinsson, E. B., and J. E. James. 1999. “A Meta-​analysis of the Effects of
Experimental Manipulations of Social Support during Laboratory Stress.”
Psychology & Health 14: 869–​886. doi:10.1080/​08870449908407353.
Tomlinson, Gary. 2015. A Million Years of Music: The Emergence of Human Modernity.
Cambridge, MA: MIT Press.
Wiltermuth, S. S., and C. Heath. 2009. “Synchrony and Cooperation.” Psychological
Science 20: 1–​5.

[ 158 ]  Social Bonding Through Embodied Agency

  159

PART SEVEN

Agency and Infancy

  161

CHAPTER 17

Timescales for Understanding
the Agency of Infants and Caregivers
JOANNA R ĄCZA SZEK-​L EONARDI

THE PROBLEM

Consider an everyday event: an infant looks at an object, his mother fol-

lows the gaze and, if the object is movable, picks it up and moves closer,
wiggles it in front of the baby, or, if it is safe, and the child can already
grasp, hands the object to the baby. The question is, who moved the object?
One intuitively obvious answer is, the mother. From the gaze of the baby
she has read the direction, selected an object, assessed its movability and
safety and then, on the basis of some assumptions about the infant’s needs
and intentions, decided to give it to the baby.
But perhaps this slice-​in-​time explanation packs too much into the head
of the mother, losing many aspects of the sense of this episode. Much in a
cognitive psychological way, it focuses on (a very important) “how” ques-
tion, neglecting “why” the whole episode is structured in this way. I  will
argue that such events do not happen only here and now, but rather are
instantiations of processes operating on several timescales, and involving
several different systems. If we take this into account, the story (what hap-
pens, who is the main protagonist) becomes more complex but perhaps
more sensible.
In what follows, I will try to show this with a microanalysis of an inter-
action between a baby and his caregiver, which is an instance of an every-
day, mundane activity. The goal is to identify the sources of forces behind
162

current, online behaviors by looking at them as points on several trajecto-

ries at the same time. Some causes for behavior will indeed, as cognitive
psychology wants, lie in the heads of individuals at a given moment. Yet for
others we will have to look at systemic levels, for example, dyadic, group,
or population, and toward slower timescales at which relevant processes
unfold.

TIMESCALES IN THE EXPLANATION OF HUMAN BEHAVIOR

The importance of timescales and multi-​systemic nature of human cog-

nitive and behavioral phenomena is increasingly often acknowledged. It
helps to identify important factors to be included in the explanations of
current behaviors (Rączaszek-​Leonardi 2003, 2010; MacWhinney 2005;
Enfield 2014), and to tackle questions about the origins of cognitive skills
(Dumas, Kelso, and Nadel 2014) and about outcomes of individual and col-
lective behaviors (MacWhinney 2005; Smith, Brighton, and Kirby 2003;
Rączaszek-​Leonardi 2009).
To give just one example of how such explanations work and how they differ
from the explanations by individual mental mechanisms, let’s look at Smith et
al.’s (2003) study of the sources for language compositionality. Traditionally,
the main explanation of this feature of language was based on the individual’s
universal grammatical device, which accounted for structuring of language in
the face of the “poverty of stimulus” argument (e.g., Chomsky 1980). Smith
and his colleagues used elegant computer simulations to show that, given the
iterative processes of communication and the nature of intergenerational
learning (a “bottleneck in transmission”), which is necessarily based only
on a subset of possible language expressions, the compositional systems are
simply the only systems that are stable. Thus, while acknowledging the main
constraint on the system (the poverty of stimulus), the authors do not see
the necessity for building explanation for compositionality by referring to
individual mental constructs. Instead, the vital elements of the explanatory
framework consist of events happening on two different timescales: ontoge-
netic processes and cultural evolution of a linguistic system. In later studies
(e.g., Kirby, Cornish, and Smith 2008) the importance of those slower times-
cale processes were confirmed in experiments with real participants in the lab.
One of the main messages from such a systemic approach is that moving
a level up, or down, focusing and refocusing on faster or slower processes
gives access to qualitatively new explanatory variables and shows that the
concepts of emergence, self-​organization, and the relation of reciprocal cau-
sality between the levels and timescales can be very helpful in constructing

[ 162 ]  Agency and Infancy

  163

viable theories of complex phenomena. This, of course, is important not

only for understanding, but also for finding points of leverage through
which it is possible to change a given dynamic system. By adapting tools
for the analysis of dynamic systems to social and behavioral sciences, we
have ways to study and to quantify the effects of action and interaction
of those different forces, and to relate processes on different timescales
(Kelso 1995).
However, since scientific explanations strive for simplicity, the multi-​
systemic and multi-​timescale approach may seem to blur the picture rather
than provide a useful analysis. The strategy is thus to focus only on certain
timescales and subsystems, chosen for their relevance on the basis of previ-
ous work and theoretical hunches, while remembering that they are only a
subset of what might become relevant. For the question of agency exam-
ined in this chapter we focus mainly on four timescales (with a fifth, that
of biological adaptation, lurking behind): cultural, ontogenetic, history of
interaction between particular individuals (“social”), and the online time
scale of current behavior.

TIMESCALES FOR AGENCY

So let us look at an episode like the one mentioned above more closely.
Here I describe in detail an interaction from the corpus of video data gath-
ered by Iris Nomikou and Katharina Rohlfing in Bielefeld (Nomikou and
Rohlfing 2011). This useful data set consists of a defined event (changing
the diaper) videotaped in 17 mother-​infant pairs in natural settings (home)
longitudinally, at five time points between the third and eighth months
of the baby’s life (for more detailed analysis of these data from the point
of view of tuning attention for coaction and development of routines, see
Rączaszek-​Leonardi, Nomikou, and Rohlfing 2013).
The interaction chosen for illustration involves a three-​month-​old baby
and his mother. The baby is lying on his back and the mother is going to
change his diaper. At one point the baby turns the head to the right, where,
approximately 7 centimeters from the baby’s face (probably much too close
for him to focus on), there is a colorful soft toy cube. The mother imme-
diately (within 1 second)1 shifts her gaze to the cube, at the same time
asking, “What is there?” Then, within the next second, she looks into the
baby’s eyes, repeating the question, but this time leaning over the baby. In
the meantime she is taking the diaper off, which eventually requires her to
move back to see what she is doing. However, with her words, in a steady
rhythm, she keeps the “topic” of the cube in focus, addressing the baby

T i m e s c a l e s i n A g e n c y of I n fa n t s a n d Ca r e g i v e r s   [ 163 ]
164

“Little mouse?” (1 second) “Hmm?” (1 second) “What is there” (1 second)

“is this a cube?” At this point, the cube is very close beside the head of the
baby and the baby seems to be looking slightly above it. After the mother
has disposed of the diaper, she asks again “Is this a cube?” and this time
picks up the cube and moves it away from the baby and higher, on the bor-
der of the changing table, saying “Look!” (1 second) “Yes, Look!”
There is nothing special about this interaction. There are thousands
just like this one between caretakers and babies every day. And here is the
point: an infant after thousands such interactions is just not the same as
before. Rich cultural structuring of coaction with a conspecific, enacted by
a caretaker, affects the perception and movement dispositions of a baby
by “coercing” him into taking a part in a holistic event. In other words, the
baby is being tuned for a meaningful, situated interaction. This requires
that the baby start to perceive own actions as enabling others’ behavior
and start perceiving others’ actions as affordances for own actions (for
more detailed analysis of the social aspect of affordances, see Costall 1995;
Rączaszek-​Leonardi et  al., 2013). The tuning happens in a “movement
first” way: the movement performed by the baby is embedded in coaction
and—​by this very coaction—​is given meaning as a part of a social event.
Let’s examine “frame by frame” how and why it happens and deter-
mine the mechanisms and causes behind the subsequent moves in this
interaction. A look to the side (most probably accidental) is reacted to by
the mother by a shift of her gaze to the nearest salient object. The same
behavior would be expected as a response to any look to the side by a con-
specific:  the biological propensity to focus on eyes (e.g., Johnson et  al.
1991)  shaped by interactions within culture is responsible for it. It does
not require any conscious reflection on the mother’s side. Due to biological
and cultural constraints, this is what she does automatically when seeing
the gaze of others. In a sense the baby’s look to the side makes the mother
look in the same direction.
A normal, in other words, culturally stabilized, course of events would be
to read the intention from the gaze of a conspecific and act accordingly (or
on the contrary, depending on the situation) toward the object. Since the
mother is in interaction with a baby who probably cannot be credited at this
age with intending to indicate something with his gaze, she shows the baby,
in an exaggerated way, how such an episode usually works. The focus here
is on the ontogenetic timescale: if you see, a thousand times, what other
people do after you look at the object, you will learn the power of your gaze.
The embedding of the infant’s behavior within the caretaker’s own actions
makes the infant’s behavior acquire meaning in the course of development.

[ 164 ]  Agency and Infancy

  165

Thus this interactive event is a re-​enactment (exaggerated, designed for

the audience this young) of a cultural script that has been stabilized in cul-
tural evolution. By drawing the baby’s behavior into the meaningful event,
as a vital part of it, the mother is letting the baby experience the causal
power of his gaze. Hence “What is there?”—​communicates her interest in
what he is looking at. A second later, she leans over the baby, bringing her
own face and eyes within his visual field—​this might be to show the impor-
tance of his gaze, or to check what the baby is really looking at. Since she
cannot immediately act on the object (having both hands busy with taking
the diaper off), she keeps up the topic of gaze with rhythmically spaced
utterances (that the baby, at this moment, presumably has no understand-
ing of, but with time will), and, finally, performs the eventual action toward
the object of the gaze, involving the baby (putting the cube directly on the
gaze line).
Here, in this single interaction, three timescales meet:  the behavior
here and now is an enactment of a culturally shaped event, which—​enacted
repetitively—​has its developmental effect. Due to the lack of space we have
neglected a fourth timescale: that of shaping and regulation of a particu-
lar relationship between this mother and this infant. Apart from enact-
ing cultural scripts for making the movements of the baby meaningful, the
mother is also regulating a social relationship with her own baby, by giving
her own, systematic, variation to the enactments. Thus the effect is not
only a biologically and culturally adequate ontogeny (learning to interact
like my conspecifics within my culture) but also a socially adequate one
(learning to interact like my mother’s son).
The above analysis may sound like a just-​so story. However, it seems
to be helpful in organizing data and posing testable hypotheses. In
Rączaszek-​Leonardi, Nomikou, and Rohlfing (2013) we document numer-
ous instances in which a baby’s movements are made to be part of cultur-
ally structured events, tailored appropriately to the age of the baby. In
an attempt to distinguish categories of behaviors, we noticed frequent
attempts of “catching the eye” of a baby, making a mother’s own move-
ments as though elicited by the infant’s moves, making pauses for elicit-
ing infant’s behaviors relevant for an event, treating random behaviors
as planned to fulfill the infant’s part, or even actually physically moving
a baby to perform a desired move. Very little, in this process, has to be
attributed to mental operations on representations, depending on elabo-
rate theories of mind, planning and/​or conscious education: the baby and
the mother are physically there to serve as their “best models” (Brooks
1990) in interaction.

T i m e s c a l e s i n A g e n c y of I n fa n t s a n d Ca r e g i v e r s   [ 165 ]
166

CONCLUSION: SO WHO DID MOVE THAT?

In this small example I wanted to convey the intricacy of the mesh of indi-
vidual, social, and cultural factors and forces bearing on a simple movement
in an interaction. Let us return to the original question: Who moved the
object? The infant? The mother? An infant-╉mother interactive system? An
encultured and socialized infant-╉mother interactive system? The mother
did perform the decisive move, but her actions were evoked by a gaze of
a child and constrained by the ontogenetic process of tuning for coaction,
which, in turn, involves re-╉enacting cultural scripts. The re-╉enactment of
the cultural scripts does not necessitate planning and conscious action on
her part. Constrained by living in a particular culture, slowly and appropri-
ately for the infant’s age she brings up the child, through co-╉movement,
basing on biological adaptations, into a culturally established coaction.
Connected to the question of agency and equally important for this
volume is the question of accountability. The dynamic shifts in account-
ability for behavior as developmental and socialization processes progress
is a particularly interesting issue. It requires philosophical, moral, and cul-
tural level analyses that are far beyond the scope of this short contribution.
Yet the identification of the threads of agency, unfolding within multiple
systems and on multiple timescales, which help recognize the sources of
agency in a particular action, seem to be an important step in the analysis
of accountability.
Even though most cognitive psychology in the last five decades has
tended to neglect the slower scales, and systems larger than an individual,
obviously sciences such as sociology or anthropology never did. For them,
the points raised here may be rather banal: it is obvious that in every deed
we are constrained by our culture, shaped in development, and driven
by our social relations. Yet the notion of timescales and the methods of
dynamic systems give us tools and a framework within which studying
mutual dependencies among the systems and reciprocal causality between
levels can become easier.

NOTE

1. The times given here are approximate. It is sufficient for the illustrative purposes
of this paper. However in the investigations of the emergence of coaction
patterns within dynamical framework, recording the exact timing is crucial.

[â•›166â•›]â•… Agency and Infancy

╇ 167

REFERENCES

Brooks, R., 1990. Elephants don’t play chess. Robotics and Autonomous Systems 6, 3–15.
Chomsky, N. 1980. Rules and Representations. Oxford, UK: Basil Blackwell.
Costall, A. 1995. “Socializing Affordances.” Theory and Psychology 5: 467–╉481.
Dumas, G., J. A. S. Kelso, and J. Nadel. 2014. “Tackling the Social Cognition Paradox
Through Multi-╉scale Approaches. Frontiers in Psychology 5: 882. doi: 10.3389/╉
fpsyg.2014.00882.
Enfield, N. J. 2014. Natural Causes of Language: Frames, Biases, and Cultural
Transmission. Berlin: Language Science Press.
Johnson, M. H., S. Dziurawiec, H. Ellis, H. and J. Morton. 1991. “Newborns’
Preferential Tracking of Face-╉like Stimuli and Its Subsequent Decline.”
Cognition 40: 1–╉19.
Kelso, J. A. S. 1995. Dynamic Patterns: The Self-╉Organization of Brain and Behavior.
Cambridge, MA: MIT Press.
Kirby, S., H. Cornish, and K. Smith. 2008. “Cumulative Cultural Evolution in
the Laboratory: An Experimental Approach to the Origins of Structure
in Human Language.” Proceedings of the National Academy of Sciences
105(31):10681–╉10686.
MacWhinney, B. 2005. “The Emergence of Linguistic Form in Time.” Connection
Science 17(3–╉4): 191–╉211.
Nomikou, I., and K. J. Rohlfing. 2011. “Language Does Something. Body Action
and Language in Maternal Input to Three-╉Month-╉Olds.” IEEE Transactions on
Autonomous Mental Development 3(2): 113–╉128.
Rączaszek-╉Leonardi, J. 2003. “The Interrelation of Time Scales in a Description of
Language.” Views & Voices 1(2): 93–╉108.
Rączaszek-╉Leonardi, J. 2009. “Symbols as Constraints: The Structuring Role of
Dynamics and Self-╉Organization in Natural Language.” Pragmatics and
Cognition 17(3), 653–╉676.
Rączaszek-╉Leonardi, J. 2010. “Multiple Time-╉scales of Language Dynamics: An
Example from Psycholinguistics.” Ecological Psychology 22(4): 269–╉285.
Rączaszek-╉Leonardi, J., I. Nomikou, K. J. Rohlfing. 2013. “Young Children’s Dialogical
Actions: The Beginnings of Purposeful Intersubjectivity.” IEEE Transactions in
Autonomous Mental Development 5(3): 210–╉221.
Smith, K., H. Brighton, and S. Kirby. 2003. “Complex Systems in Language
Evolution: The Cultural Emergence of Compositional Structure.” Advances in
Complex Systems 6(4): 537–╉558.

T i m e s c a l e s i n A g e n c y of I n fa n t s a n d Ca r e g i v e r s â•… [â•›167â•›]
╇ 169

CHAPTER 18

Movement Synchrony, Joint Actions,

and Collective Agency in Infancy
BAHAR TUNÇGENÇ

INTRODUCTION

Social encounters begin with infants’ interactions with their caretakers in

give-╉and-╉take relationships and continue all throughout toddlerhood and
beyond, as children explore and learn from their environment and play
with their peers. Much of this social interaction involves infants getting
together with others in various ways to achieve common goals. As natural
as they may seem, the ubiquity and human specificity of such joint actions,
as well as the early development of social, cognitive, and motor skills
required to carry them out are quite remarkable from a broader, evolution-
ary perspective.
One important feature of joint actions is their implications for agency.
For instance, when a child kicks a ball, it is clearly the child who is the
“agent” of this kicking action and it is the ball that is the “recipient” of
the action. Yet, the distinction between an agent and a recipient starts
to get blurred in, for example, a football game. Although agents can still
be identified for individual kicking actions, there emerges an overarching
collective agency, where individuals can be counted as agents even for the
actions that they have not fully executed. Now, it is us who play the football
game, rather than me kicking the ball and you kicking it back at me. In other
words, by joining actions, we also join agency.
170

The emergence of this “we-╉ness” relies initially on infants’ understand-

ing of others’ goals and intentions, which then contributes to engaging in
joint actions as infants start sharing attention with others. A key element
of both goal understanding and attention sharing is physical coordination
of movements among the interactants. Moving in well-╉coordinated ways
with others may have significant long-╉term effects on social development.
Thus, the development of each one of these aspects—╉goal understanding,
attention sharing, and performing synchronous, coordinated movements
in joint actions—╉is crucial for understanding the emergence of collective
agency.

GOAL UNDERSTANDING AND AGENCY ATTRIBUTIONS

IN INFANCY

Imagine a person reaching for a water bottle; as an observer, you might

automatically make the inference that the person wants to drink some
water. Viewing actions as goal-╉directed and attributing agency to actors
is a prominent feature of human cognition. Categorizing the world in
terms of “actors with goals” and “things acted upon” assumes two dif-
ferent roles in an interaction: agents and recipients. What kinds of cues
do infants make use of while identifying the agents and recipients in an
interaction?
A number of behavioral cues can guide infants’ attribution of goal-╉
directedness and agency. To assess this, in one seminal study, nine-╉
month-╉old infants were shown a small circle making contact with a large
circle by jumping over an obstacle that separated them (Csibra, Gergely,
Biro, Koos, and Brockbank 1999). Next, infants were shown the same
objects moving in irrational ways, namely “jumping over” even when
the obstacle was no longer there. The infants’ surprise reactions to this
irrational action were taken as evidence that they perceive the objects’
previous actions as goal-╉directed. Importantly, however, the small circle’s
movements had certain properties:  the circle initiated its own move-
ments (self-╉propulsion), got bigger/╉smaller in response to the large
circle’s size changes (contingency), pursued contact with the large circle
even when their positions were different (equifinal variation), and fol-
lowed alternative paths in order to reach its goal (flexibility in actions).
Under these conditions, nine-╉month-╉olds showed surprise, suggesting
that they interpreted this evidently inanimate and non-╉agentive small
circle’s actions as goal-╉directed. Evidently, infants have a tendency to

[â•›170â•›]â•… Agency and Infancy

  171

over-​attribute goal-​directedness and agency even to inanimate objects,

provided the objects’ motions meet certain criteria. Yet, they are not
alone:  adults also over-​attribute agency frequently, for instance, when
they see faces in the clouds or exclaim their frustrations at a malfunc-
tioning computer or car engine.
In addition to behavioral cues, infants also make use of social cues
while attributing agency to animate and inanimate objects. The social
cues may be certain physical properties such as the presence of eyes and
a face or communicative behaviors that the actors perform. Arguably,
these social cues become particularly relevant when there is uncertainty
about the actors or the conditions under which an event takes place. For
instance, when an inanimate object engages in a brief “small talk” with an
adult prior to performing some actions, 12-​month-​olds who observe this
inanimate object are more likely to attribute agency to the object than if
the object did not provide communicative cues (Johnson, Shimizu, and
Ok 2007). Another example is when infants observe only the end-​state
of an action (an object appearing on the scene) without observing the
initial state (an actor bringing the object into the scene). When asked to
guess the actor for such actions whose initial states are not available to the
infants, 10-​month-​olds assume that an animate (i.e., human hand) rather
than an inanimate (i.e., toy train) actor must have been the agent that
brought the object into the scene (Saxe, Tzelnic, and Carey 2007). Similar
to the previously mentioned studies, infants’ agency attribution in these
studies is evident by increased surprise reactions that they show upon
viewing the inanimate rather than the animate actor perform the action.
Indeed, infants are quite good at picking up sociality even from minimal
cues; they readily attribute agency to toy puppets that have facial features,
for example (Saxe et al. 2007). This indicates the importance of social cues
as well as how flexibly infants integrate information from various sources
as they consider agency.
Understanding the goals behind others’ actions and identifying the
agents and recipients of an action are important first steps on the way
to collective agency. Infants seem to develop an understanding of goal-​
directedness within the second half of the first year of life. Gradually, they
start including more intricate features of the actors to evaluate agency in
more complex circumstances. Following on from there, another milestone
on the way to collective agency is the ability to understand what others
intend to do and to align one’s own intentions with those of others. The
next section will deal with how understanding the intentions of others
facilitates attention sharing.

S y n c h r o n y, J o i n t A c t i o n s , a n d Ag e n c y i n I n fa n c y   [ 171 ]
172

INTENTION UNDERSTANDING AND ATTENTION SHARING

IN INFANCY

Recall the example of a person reaching for the water bottle and the com-
mon inference that the person would be reaching for the bottle to drink
some water. There are, indeed, many other sub-╉goals involved in that
action: holding the water bottle, bringing it closer to one’s mouth, pouring
water into the mouth, so on and so forth. It seems, however, that the human
mind often ignores these middle steps in reading the behavior. Depending
on what that particular circumstance brings, we identify an ultimate inten-
tion behind the action being performed—╉for example, quenching thirst—╉
and focus on that instead. How, then, do infants understand the intentions
of agents? When and how does this understanding transform into infants’
sharing attention and engaging in joint actions with others?
Reading the intentions of others might be tricky, partly because differ-
ent intentions might motivate similar action-╉outcome sequences. Still, an
adult would hardly ever miss the difference between a person purposefully
throwing a water bottle on the floor and a person dropping it accidentally.
Similar comparisons were made to investigate infants’ understanding of
intentional and accidental actions. In one study, nine-╉month-╉old infants
interacted with an adult who was not giving them the toy that they wanted
(Behne, Carpenter, Call, and Tomasello 2005). Crucially, the adult was
either teasing the infants and was unwilling to give them the toy (inten-
tional action) or actually wanted to give them the toy but was unable to
do so due to clumsiness or other physical obstacles (accidental action).
Although the adult’s actions looked very similar in both conditions, infants
reacted with more frustration and impatience when the adult was unwill-
ing as opposed to unable to give them the toy. This indicates that at nine
months, infants understand the intentions of agents, even if the observed
actions and outcomes may look similar. This understanding influences the
infants’ emotional state and how they respond to the actor of the inten-
tional/╉accidental actions. Understanding others’ intentions is a crucial step
in starting to share intentions with others. Sharing intentions on objects
of mutual interest is what makes joint actions possible. Hence, a closer look
at infants’ developing abilities to share intention with others is important.
A well-╉studied method for sharing attention and intention with others
is using the pointing gesture. As soon as someone points to an object with
an index finger, most people automatically turn their heads and look at the
object being pointed out to. Although young infants may use the pointing
gesture merely to attract attention, a more advanced use of it is communi-
cative and involves the intention to share attention with others. At least

[â•›172â•›]â•… Agency and Infancy

╇ 173

from 12 months of age, infants can use the pointing gesture for the lat-
ter purpose, for example, to inform adults of the locations of objects that
they have been looking for (Liszkowski, Carpenter, and Tomasello 2008).
A  more detailed analysis further reveals a bidirectional developmental
pattern between infants’ comprehension and production of the pointing
gesture. From 8.5 months of age onward, those infants who comprehend
others’ pointing gestures as directed at objects also point themselves to
direct others’ attention to objects they find interesting (Woodward and
Guajardo 2002). This indicates that infants already start sharing atten-
tion before the age of one. This can be considered as a preliminary form
of joint action, where the infant and the adult act jointly, by looking in the
same direction and sharing the intention to attend to this object of mutual
interest.
More elaborate forms of joint actions are also observed within the first
year. They usually take place in infants’ interactions with their mothers as
they play together or engage in give-╉and-╉take type relations, such as when
a mother feeds her baby. One notable feature of these joint actions is the
high coordination observed among the interactants. Coordinating the tim-
ing of actions can ease attention sharing, action prediction, and thereby
expand the scope of joint actions the infants can perform (Reddy, Markova,
and Wallot 2013). In the following section, we will examine this issue more
closely to highlight the role of highly coordinated, synchronous move-
ments in infants’ joint actions.

MOVEMENT COORDINATION AND JOINT ACTIONS IN INFANCY

Adding some more detail to our water bottle scenario, let us assume that
there is another individual, a baby, in the storyline, and that the person
reaching for the bottle is a mother who wanted to feed her baby. In this case,
the baby’s behaviors would suddenly be pivotal as well. It would be crucial
that the mother’s and the baby’s movements are well coordinated: the baby
should open his mouth as the mother reaches toward him, and the mother
should start the pouring action at the right time to avoid any accidents.
Although such precise coordination in movements might require highly
developed skills, infants and mothers do nevertheless coordinate their
behaviors with each other in various forms from very early on. Coordinated
interactions feed into children’s later abilities to work together with oth-
ers on a shared goal, communicate in socially acceptable ways, and develop
a sense of what is entailed in collective actions. Due to these significant
consequences on social development, coordinated interactions offer a

S y n c h r o n y, J o i n t A c t i o n s , a n d Ag e n c y i n I n f a n c y â•… [â•›173â•›]
174

promising context in which to explain the development of abilities under-

lying collective agency.
From as early as at six weeks of age, infants begin to display instances
of coordinated movements in their interactions with others. For instance,
infants time their eye movements with their mothers’ vocalizations by
shifting the direction of their gaze toward and away from the mother’s face
according to the time when the mothers start and stop talking (Crown,
Feldstein, Jasnow, Beebe, and Jaffe 2002). By nine months, infants
start comprehending the structure of adultlike communication and take
turns in their vocalizations as they interact with their mothers (Jasnow
and Feldstein, 1986). Notably, this is also the time when infants initiate
attempts to re-​engage their play partners in a joint game, by using com-
municative tools such as frowning, pointing to the toys, or vocalizing to
express their frustration when their play partners disengage from the
game (Ross and Lollis 1987). Through coordinating the timing of their
actions, then, infants learn the appropriate means to communicate with
others. Such coordinated interactions also have important implications for
other aspects of social development. For instance, the frequency of bidirec-
tional coordination observed between mother-​infant pairs at four months
has been shown to predict infants’ secure attachment to their mothers and
unfamiliar others, as well as infants’ cognitive skills, communicative com-
petency, and socio-​emotional adaptation, at 12 months of age (Jaffe et al.
2001). These findings point to the importance of coordinated movements in
infants’ social lives. Thus, a simple aspect of development—​namely, match-
ing the timing of one’s actions to those of others—​may have important
implications for the development of joint actions and collective agency.
Recently, researchers have started to investigate the specific role of
highly coordinated, synchronous movements in infants’ social cognition.
Despite being in an infancy of its own, this line of research has begun to
show quite promising results. One study has shown that 12-​month-​old
infants, but not nine-​month-​old infants, distinguish between synchro-
nous movements that take place in social and non-​social settings, and have
a preference for synchronously moving partners only in the former case
(Tunçgenç, Cohen, and Fawcett 2015). Taken within the broader frame-
work of coordinated interactions in early infancy, these results indicate
that precisely timing actions to one another might facilitate social bonds
among interaction partners. Furthermore, rather than being innate, these
facilitatory effects emerge toward the end of the first year of life, possi-
bly as a factor of infants’ increasing exposure to synchronous, coordinated
movements in joint action settings with their caretakers. This social mean-
ing embedded in coordinated movements might similarly imply “we-​ness”

[ 174 ]  Agency and Infancy

╇ 175

in more elaborate joint action settings. One study has revealed that fol-
lowing a face-╉to-╉face bouncing phase with an adult, 14-╉month-╉old infants
helped the adult more if they had bounced to the same rhythm (synchro-
nously) than if they had bounced to different rhythms (asynchronously)
with the adult (Cirelli, Einarson, and Trainor 2014). Importantly, moving
synchronously with others does not induce a general state of helpfulness.
Instead, infants were shown to help only those individuals who have moved
synchronously with them and those who were close affiliates of their syn-
chronizing partners (Cirelli, Wan, and Trainor, 2016), but not to neutral
observers (Cirelli, Wan, and Trainor 2014). This suggests that performing
synchronous movements might serve as a cue for social bonding and coop-
erative behavior among interactants.
To date, how synchronous, coordinated movements produce these social
outcomes is largely unknown. It might be that moving in similar ways with
another person gives rise to the inference that the interactants are simi-
lar to one another in other respects as well. The contingency involved in
coordinated movements might ease predicting the future actions of oth-
ers. This, in return, may imply a potential for being good cooperation part-
ners in future joint action settings. Overall, increase in perceived similarity
and closeness toward one’s interaction partner might blur the boundaries
between self and other, ultimately reinforcing the emergence and mainte-
nance of collective agency.

CONCLUSION

Adapting to a world filled with agents who continuously share goals and
perform collective actions is not trivial. Accurate interpretation of a simple
action, such as reaching for a water bottle, requires the development of a
suite of social, cognitive, and motor skills. Still, infants in the first year of
life already demonstrate an understanding of others’ goals and intentions
and point to objects to share attention with other people. Elaborate forms
of joint actions can also be observed in infants’ interactions with their
mothers from birth onward. One prominent feature of these actions is the
high coordination of movements performed by the interactants.
The effects of performing synchronous, coordinated movements on
children’s social development have recently attracted attention from
researchers. Evidence is accumulating on the importance of action timing
for social cognition and joint action development. Although far from being
conclusive, these new findings suggest that the similarity and contingency
of movements involved in synchronous, coordinated interactions might

S y n c h r o n y, J o i n t A c t i o n s , a n d Ag e n c y i n I n fa n c y â•… [â•›175â•›]
176

create an enhanced feeling of “we-╉ness” among the interactants. This way,

synchronous, coordinated movements can facilitate social bonding start-
ing from early years of life and pave the way from individual to collective
agency.

REFERENCES

Behne, T., M. Carpenter, J. Call, and M. Tomasello. 2005. “Unwilling versus

Unable: Infants’ Understanding of Intentional Action. Developmental
Psychology 41(2): 328–╉337. doi:10.1037/╉0012-╉1649.41.2.328.
Cirelli, L. K., K. M. Einarson, and L. J. Trainor. 2014. “Interpersonal Synchrony
Increases Prosocial Behavior in Infants.” Developmental Science 17(6):
1003–╉1011. doi:10.1111/╉desc.12193.
Cirelli, L. K., S. J. Wan, and L. J. Trainor. 2014. “Fourteen-╉Month-╉Old Infants
Use Interpersonal Synchrony as a Cue to Direct Helpfulness.” Philosophical
Transactions of the Royal Society of London. Series B, Biological Sciences
369(1658). doi:10.1098/╉rstb.2013.0400.
Cirelli, L. K., S. J. Wan, and L. J. Trainor. 2016. “Social Effects of Movement
Synchrony: Increased Infant Helpfulness only Transfers to Affiliates of
Synchronously Moving Partners.” Infancy. doi:10.1111/╉infa.12140.
Crown, C. L., S. Feldstein, M. D. Jasnow, B. Beebe, and J. Jaffe. 2002. “The Cross-╉
Modal Coordination of Interpersonal Timing: Six-╉Week-╉Olds Infants’ Gaze
with Adults’ Vocal Behavior.” Journal of Psycholinguistic Research 31(1): 1–╉23.
doi:10.1023/╉A:1014301303616.
Csibra, G., G. Gergely, S. Biro, O. Koos, and M. Brockbank. 1999. “Goal Attribution
Without Agency Cues: The Perception of ‘Pure Reason’ in Infancy.” Cognition
72: 237–╉267.
Jaffe, J., B. Beebe, S. Feldstein, C. L. Crown, M. D. Jasnow, P. Rochat, and D. N. Stern.
2001. “Rhythms of Dialogue in Infancy: Coordinated Timing in Development.”
Monographs of the Society for Research in Child Development 66(2): 1–╉132.
doi:10.1111/╉1540-╉5834.00137.
Jasnow, M., and S. Feldstein. 1986. “Adult-╉like Temporal Characteristics of
Mother-╉Infant Vocal Interactions.” Child Development 57(3): 754–╉761.
Johnson, S. C., Y. A. Shimizu, and S-╉J. Ok. 2007. “Actors and Actions: The Role
of Agent Behavior in Infants’ Attribution of Goals.” Cognitive Development
22(3): 310–╉322. doi:10.1016/╉j.cogdev.2007.01.002.
Liszkowski, U., M. Carpenter, and M. Tomasello. 2008. “Twelve-╉Month-╉Olds
Communicate Helpfully and Appropriately for Knowledgeable and Ignorant
Partners.” Cognition 108(3): 732–╉739. doi:10.1016/╉j.cognition.2008.06.013.
Luo, Y. 2011. “Three-╉Month-╉Old Infants Attribute Goals to a Non-╉human Agent.”
Developmental Science 14(2): 453–╉460. doi:10.1111/╉j.1467-╉7687.2010.00995.x.
Reddy, V., G. Markova, and S. Wallot. 2013. “Anticipatory Adjustments to Being
Picked Up in Infancy.” PLoS One 8(6): e65289. doi:10.1371/╉journal.
pone.0065289.
Ross, H. S., and S. P. Lollis. 1987. “Communication Within Infant Social Games.”
Developmental Psychology 23(2): 241–╉248.

[â•›176â•›]â•… Agency and Infancy

  177

Saxe, R., T. Tzelnic, and S. Carey. 2007. “Knowing Who Dunnit: Infants Identify
the Causal Agent in an Unseen Causal Interaction.” Developmental Psychology
43: 149–​158. doi:10.1037/​0012-​1649.43.1.149.
Tunçgenç, B., E. Cohen, and C. Fawcett. 2015. “Rock with Me: The Role of Movement
Synchrony in Infants’ Social and Non-​social Preferences.” Child Development
86(3): 976–​984. doi:10.1111/​cdev.12354.
Woodward, A. L., and J. J. Guajardo. 2002. “Infants’ Understanding of the Point
Gesture as an Object-​Directed Action.” Cognitive Development 17(1):
1061–​1084. doi:10.1016/​S0885-​2014(02)00074-​6.

S y n c h r o n y, J o i n t A c t i o n s , a n d Ag e n c y i n I n fa n c y   [ 177 ]
  179

PART EIGHT

The Agency of Materiality

  181

CHAPTER 19

The Agency of the Dead

ZOË CROSSL AND

I n this chapter I consider how human remains are thematized as agentive

within the field of forensic anthropology. The flourishing popular literature
on forensic science commonly summons the image of the corpse that
speaks or testifies. Here I explore how the corpse’s agency is imagined in
these texts, and how the forensic anthropologist’s agency is positioned
relative to the corpse. How might these common claims of corpse agency
prompt us to think more critically about the agentive aspects of the dead?
As an ontologically ambiguous entity that appears simultaneously as an
object and as a person, the dead body provides a productive site for think-
ing about how agency works in relation to people and things. Particularly
in moments when the corpse is found outside of its proper place in the
tomb, it seems to harbor some kind of problematic animacy that may be
expressed as hauntings, dreams, or, as I show here, in ways that are chan-
neled through a language of science and evidence.
Mass-​market texts about osteoarchaeology, forensic anthropology, and
forensic pathology are insistent in their use of the figure of the speaking
corpse. “Dead Men Do Tell Tales” assert the front covers of memoirs by
forensic anthropologists William Maples (Maples and Browning 1994)
and William Bass (Bass and Jefferson 2003). Peggy Thomas’s book Talking
Bones: The Science of Forensic Anthropology (Thomas 1995) locates the speak-
ing corpse within a rhetoric of science, while Witnesses from the Grave: The
Stories Bones Tell by Christopher Joyce and Eric Stover (1991), situates the
words of the dead within the performance of testimony, gesturing toward
182

the important forensic dimensions of the image. The corpse’s testifying

role is reiterated in Silent Witness by Roxana Ferllini (2002), which also
points to the paradoxical dimensions of the figure: these are dead bod-
ies that speak while manifestly remaining silent. The speaking corpse is
a complex image that reveals a folk theory of agency at the heart of the
assertively scientific fields of forensics. Clearly, these statements are meant
metaphorically, yet it is productive to explore the modes through which the
corpse speaks, and to ask what this might reveal about where the corpse’s
apparent agency lies. To think through the corpse’s treatment as agent,
and to explore its entailments, I take a semiotic approach that draws on
the work of Charles S. Peirce (1931–​1935, 1958), and on Paul Kockelman’s
recent theorization of agency in semiotic terms (2007).
How do the forensic dead speak? When we look at what the dead say
within forensic texts, it becomes clear that they always speak truthfully,
challenging lies and misinformation about their lives and deaths. Forensic
anthropologist Clyde Snow remarks that “bones make good witnesses.
Although they speak softly, they never lie and they never forget” (Joyce and
Stover 1991:144). These are personal, autobiographical narratives, specific
to the individuals in question. “Bones tell stories of identity, trauma, and
postmortem mutilation,” explains forensic anthropologist Kathy Reichs
(Owen 2000:8). The forensic corpse tells of childhood accidents, visits to
the dentist, and, most importantly, of how it died. Anything that has left
a visible trace upon the body can be expressed, but the vast range of lived
experience is left unspoken. The forensic dead may never lie, but theirs is a
constrained and descriptive field of speech that does not deal with feelings,
wishes, or desires. This restricted speech channels the apparent animacy of
the dead into a medico-​legal language of science and of evidence, in which
only certain kinds of statement are recognizable.
The speaking corpse shares in the broader empiricist image of facts
that speak for themselves, offering a kind of poetics of empiricism, a
way of imagining the status and efficacy of fact or evidence. Dead bodies
act as a metaphor for fact, and the investigative work that takes place
around the corpse provides a field for imagining the relationship of fact
to interpretation. The empiricist view of the dead body as self-​evident
fact is complicated however by the imagery of testimony that is com-
monly attached to the figure of the speaking corpse. Dead bodies are said
to act as “witnesses” and “testify” in the service of truth. Yet witnesses
are notoriously unreliable, their testimony fraught with uncertainty,
whether from the frailties of memory or from deliberate obfuscation.
In bringing the language of witnessing into the realm of fact and obser-
vation, the subjectivity of the witness is inserted into the apparently

[ 182 ]  The Agency of Materiality

  183

uncontestable realm of fact. This is a hazardous move, potentially

throwing the special truth-​telling abilities of the corpse into doubt, and
putting claims to objectivity at risk. The willingness to take such meta-
phorical risks suggests that the image of the witness offers something
important that the notion of fact on its own does not. To suggest that
facts are witnesses is another way of stating that facts act as evidence.
Indeed these two terms—​fact and evidence—​are often conflated in the
popular literature on forensics.
In the context of law courts, evidence has traditionally been solicited
from the speech or documents of human witnesses. Yet, within the realm
of medical fact, evidence is usually conceptualized in terms of the visual.
Indeed, the very term evidence encompasses a claim for the primacy of
sight, grounded in the Latin verb “to see,” videre, and echoed in the emer-
gence of the modern practice of autopsy (to see with one’s own eyes). Here
the emphasis is primarily on the researcher as active agent, as someone
who interprets the passive body that lies ready to be dissected (as discussed
by Sawday 1995). In contrast, the testifying corpse seems to offer the rec-
ognition of a relationship between those who speak and those who listen.
The dual transitive and intransitive nature of the verb to speak seems to be
important here: I may speak regardless of who is present, or I may speak
directly to someone. This allows for an image of fact as something that
simultaneously stands alone, and acts as evidence for or to someone. The
corpse’s speech exists on its own terms, whether witnessed or not, but it is
also directed to a potential or anticipated audience: the expert who is able
to hear and decode this speech.
The figure of the expert adds another complication to the corpse’s speech,
for the speech of the dead needs to be heard and interpreted. The forensic
specialist must move between the facts of the case and the argument that
he or she composes from such facts. This is a troublesome relationship. In
the context of forensic evidence, it is problematic to lay too much emphasis
on the work of interpretation, and this is because the theory that under-
lies forensic work is one that asserts a straightforward separation between
fact and value. On the one hand are the data, on the other the necessary
but fraught work of their interpretation and presentation by the forensic
witness. Within this frame it is understood that those witnesses who are
most expert will not pollute the facts with their own values, but instead act
as intermediaries, decrypting rather than translating the language of the
corpse. This points to another effect of the figure of the speaking corpse.
It not only constrains the agency of the dead but also denies some agency
to the analyst:  the expert does not speak, the evidence does. This is to
transfer agency to the dead, and along the way to acknowledge some of the

T h e Ag e n c y of t h e   Dead  [ 183 ]
184

conflicted emotions that work with the dead entails—​emotions that are
foreclosed to scientists working within a legal context.
The image of the speaking corpse folds together different evidential
fields, including law, science, medicine, history, and biography. Each field
has its own determinants of what counts as evidence, and how it should
be assessed. Evidence is here situated within different regimes that gov-
ern what is allowed to enter in as material fact and how it is evaluated.
Drawing on Peirce, we can conceptualize evidence not as something that
stands alone, but as a semiotic relation. So in this sense evidence is always
for something—​for prosecution, for historical narrative, for medical train-
ing. Yet at the same time, it is also evidence of something. Evidence for
prosecution might be evidence of violent events in the past. Evidence
for medical training might be of pathological changes in bodily tissues.
In thinking about the ways in which the dead are presented as evidence,
we need to attend both to how the different relationships between these
elements are composed and to the evidential regimes that govern how,
when, and for whom or what something acts as evidence. These may be
more or less explicitly formulated. Moving away from the problematic fact-​
value distinction, Peirce’s triadic sign relation offers a means to unpack
the evidential relation as specified under different regimes, and hence
to explore the axes along which the corpse might be understood to act.
Below I briefly sketch out three different evidential configurations, before
returning to consider the image of the speaking corpse. To do this I map
the material evidence of the dead onto the Sign that stands at the heart
of Peirce’s sign relation. The sign of the body stands for a number of dif-
ferent possibilities, events and processes (“Objects” in Peirce’s terms). So
bodily traces may be narrowly construed within a forensic sphere as signs
of injury or harm, or understood in ways that are less closely specified. The
body may act as a sign of possibility, of things that may have happened, or
indeed as a sign of things that should have happened, as I explore below.
What is this relationship for, however? For whom or for what does the
body stand as a forensic trace of violence, or as a sign of potential? What
is evidence if it is not marshaled in the service of a truth-​finding project?
Peirce directs us to the insight that evidence is ineluctably triadic. A third
term (Peirce’s “Interpretant”) is necessary to complete the evidential rela-
tion, so that material traces can be understood as both signs of something
and signs for different possibilities, goals, and projects.
For relatives of the dead, the corpse presents ambivalently as their fam-
ily member, but also as the necessary evidence of death. It is common in
the popular literature on forensic exhumation to hear reported claims that
friends and families cannot find a measure of peace without the presence

[ 184 ]  The Agency of Materiality

  185

of the body. The recovery of the dead shifts the status of those who are
missing from possibility to actuality. The here-​and-​now-​ness (or haecceity)
of the body acts as a concrete sign of death. What is important for families
in this context is recognition. Are these remains recognizable? Is this the
sought-​for family member? This recognition is often based on a relation of
similitude or iconicity that might not stand up in a law court. For example,
in the search for those killed from Srebrenica, in the former Yugoslavia,
photographs of clothing and possessions found with the dead were com-
piled by human rights organizations. These were meant for display to rela-
tives in order to mediate their experience with the traces of the dead, and
to aid identification. However, DNA analysis showed that some identifica-
tions carried out this way were mistaken (Huffine et al. 2001:274; Wagner
2008: 99–​100). Here we see the weaknesses of a purely iconic relation for
forensic inquiry. Ewa Klonowski (2007:167) describes how, in early exhu-
mations before DNA testing was introduced, she observed traumatic situ-
ations “in which two different families claimed the same remains because
of recognition of familiar clothing or shoes.” For a legal case, much hinges
on the ability to demonstrate a specific causal relationship between past
events and present traces. Certainly, iconic elements are important in
these claims. For example, if a jacket can be shown to resemble one worn
by a missing individual, this iconicity has a powerful visual force, useful
in a law court for evoking a feeling of recognition and convincing a jury.
However, to be accepted as evidence within science and law, an indexical
relationship has to be demonstrated; it is not enough that the traces are
the same. Even when the relation of resemblance is idiosyncratic and spe-
cific to the individual—​the torn pocket that was never fixed; the button
sewn back on with mismatched thread—​the connection with the miss-
ing must be demonstrated. Clothing acts as a powerful potential marker
of identity because of its indexical relation with the person who wore it.
Clothes take on something of the person who inhabited them and in cases
such as these can come to stand metonymically for the missing. Such
indexical traces may link crime scene and perpetrator, or human remains
and the missing, but the chain of evidence must always be preserved so
that the nature of this indexical trace can be maintained and assessed.
Sarah Wagner (2008:99) recounts how men who fled from the violence
at Srebrenica swapped clothing or were given identity cards and photos
by the wounded. These actions led to confusion over presumptive identi-
fications made on the basis of clothing. Here the sequence of events was
unclear and the nature of the indexical sign was misjudged: the clothes
and personal items did indeed point to their owners, but they had become
separated from them at some point. The particularities of a forensic sign

T h e Ag e n c y of t h e   Dead  [ 185 ]
186

relation are fundamental to its efficacy and authority. However, the spe-
cific indexical or causal link between the trace and the thing that made it
has to be carefully documented and preserved. In the legal context this
translates into a concern with the chain of evidence, which itself testi-
fies to the indexicality of the trace, and to the claim that no tampering
has taken place. It is only on the basis of an undisturbed and confirm-
able indexical trace, as interpreted as a fact, that a forensic argument can
be made about what happened at a specific moment and locale. Family
members may be relatively unconcerned with the evidential strictures of
science and law. Rather, recognition often seems to take place in an affec-
tive and relatively unexamined way. Yet this form of immediate perceptual
judgment certainly has its place and is operational within scientific and
legal inference. The difference lies not so much in the feeling of recogni-
tion, as the way in which these semiotic domains assess such judgments
and specify the nature of the relation of material evidence to its referent
(or Object).
Finally, moving to wider historical narratives, another evidential dimen-
sion comes into view. Whether constructing stories of nationalism and
reconstruction, or narrating anthropologists’ work around the dead body,
corpses come to stand for a range of meanings established through conven-
tion. The dead are often used to evoke larger entities: a group, an event,
the body politic, or indeed facts themselves. These conventions are often
embedded in perceived similarities and connections between the corpse
and the kind of narrative being told. For example, the work of exhuma-
tion, of revelation, offers a powerful and commonly used metaphor for the
recovery of history in situations of political repression, a symbolic rela-
tion that also draws upon iconic and indexical elements for its potency. In
such narratives the individual often drops out of view, subsumed within
a broader accounting. The conventional or symbolic relation between the
sign of the dead and the thing that it stands for brings me back to the
speaking corpse as a complex metaphor for fact. Although, before return-
ing to this theme, I should note that the above discussion presents a highly
schematized sketch of the field of shifting evidential relationships in foren-
sic work. Claims that center on the signs of the dead draw on all of these
dimensions, but may at different moments privilege one element of the
sign relation more than another.
Turning back to the speaking corpse, we can see that it provides a useful
way of imagining the different dimensions of forensic inquiry, and of think-
ing about forensic evidence. In its restricted speech the image acknowl-
edges the way in which the dead body can constrain particular readings
within certain evidential regimes. The figure of the speaking corpse also

[ 186 ]  The Agency of Materiality

  187

recognizes the disturbing animacy of the dead in a context where it can-

not normally be recognized. Finally, the image allows the analyst to act
as mediator while effacing his or her agency. And this is at the heart of
the question of where evidence sits in the relationship between fact and
value. The concept of evidence is fundamentally troubling for any attempt
to separate the world into passive fact and agentive interpretation, pre-
cisely because of the way in which it is simultaneously motivated (evidence
for something) and yet understood to stand apart from that motivation (as
evidence of something else).
What does the figure of the corpse that speaks offer for thinking about
the agentive dimensions of the corpse? It is clear that the indexical traces of
violence can push back against narratives that assert that a person simply
left home and was living elsewhere, or had died accidentally, or in battle.
Those responsible for such traces may lie about what took place, but the signs
of the dead can correct them, and so restrict their semiotic agency. Drawing
upon Paul Kockelman’s (2007) terminology, we could say that the semiotic
flexibility or ability to control the expression of forensic signs is highly con-
strained. In the former Yugoslavia, the dead were even dug up and reburied
in an attempt to hide what had taken place (Skinner et al. 2002). But as the
French criminalist Edmond Locard famously noted at the start of the twen-
tieth century “Every contact leaves a trace”—​even in changing the loca-
tion of the dead, new iconic-​indexical traces were left that could be tracked
and used as evidence. This illustrates too the limited degree of influence
that one has over how the indexical relation is composed, speaking again to
the constraints placed on the semiotic freedom of the guilty. Perhaps even
more significant is the ability of the signs of the dead to elicit accountabil-
ity—​forensic traces can force the culpable to account against their wishes.
This is not experienced as a purely passive kind of constraint. Although
evidence is specified under particular regimes, and the indexical traces
of forensic evidence seem to have the most agentive ability to intervene
in the lives of others, it is also important to take account of how evi-
dence itself works to constitute the context. When the sign of a corpse
is revealed, action must be taken, and a certain kind of action is required.
The discovery of human remains immediately points to something dis-
turbing that happened in the past, and in so doing demands a particular
response. The unexpected appearance of the dead petitions for investiga-
tion and for their status to be established. This starts with what Peirce
called an abduction—​at its most rudimentary, a feeling of unease when
confronted with something out of place, and at its most developed, the
hypothesis made by a forensic expert in seeing clues and, Sherlock Holmes-​
like, inferring the solution. There is something unfinished in this kind of

T h e Ag e n c y of t h e   Dead  [ 187 ]
188

evidence, but it is not simply waiting to be activated, it also shapes the ways
in which it is acted upon.
This is to move away from a view of agency as inhering only in the liv-
ing, and instead to view it as a collaborative semiotic project that is also
shaped and constrained by the dead. The goals of prosecution and justice
cannot be carried out without them. The image of the testifying corpse
recognizes these agentive dimensions. But where does this leave the idea
of fact as something that stands alone and speaks for itself? When facts
speak for themselves they act as Signs that have been conflated with their
Objects and prescinded from the Interpretant relation. This is to fetishize
the power of the indexical relation and to underplay the different interpre-
tive possibilities to which it gives rise. To move away from a folk under-
standing of the agency of the dead and to fully engage with the ability of
the dead to intervene in the lives of the living needs a theory of agency
that is fundamentally relational, while also able to recognize the ways in
which that relationality can be hewn and circumscribed. Bruno Latour has
approached this question by considering how facts are treated as stable
“black boxes” when they are enrolled successfully into scientific work. At
moments of breakdown or challenge such facts are revealed to be composed
of relational networks in which many different actors are involved (Latour
1999). However, in considering the agency of the forensic dead, our atten-
tion is drawn not only to the relational constitution of the corpse as fact,
but also to the different regimes under which certain evidential relations
are brought into view or removed from the realm of possibility. Here it
seems most helpful to think of agency as an adjective rather than a noun,
reconceptualizing it in terms of agentive constellations that permit differ-
ent forms of action (Crossland 2014:100–​137). This opens the way to more
fully decenter agency as well as to acknowledge that it is always a project
that is grounded in particular material conditions, and which unfolds over
time relative to certain goals.

REFERENCES

Bass, William, and Jon Jefferson. 2003. Death’s Acre: Inside the Legendary Forensic Lab
the Body Farm, Where the Dead Do Tell Tales. New York: Penguin.
Crossland, Zoë. 2014. Ancestral Encounters in Highland Madagascar: Material Signs and
Traces of the Dead. Cambridge: Cambridge University Press.
Ferllini, Roxana. 2002. Silent Witness: How Forensic Anthropology Is Used to Solve the
World’s Toughest Crimes. Willowdale, Ontario: Firefly Books.
Huffine, Edwin, John Crews, Brenda Kennedy, Kathryn Bomberger, and Asta Zinbo.
2001. “Mass Identification of Persons Missing from the Break-​up of the Former

[ 188 ]  The Agency of Materiality

  189

Yugoslavia: Structure, Function, and the Role of the International Commission

on Missing Persons.” Croatian Medical Journal 42(3): 271–​275.
Joyce, Christopher, and Eric Stover. 1991. Witnesses from the Grave: The Stories Bones
Tell. Boston: Little, Brown and Company.
Klonowski, Eva-​Elvira. 2007. “Forensic Anthropology in Bosnia and
Herzegovina: Theory and Practice amidst Politics and Egos.” In Forensic
Archaeology and Human Rights Violations, edited by Roxana Ferllini, 148–​169.
Springfield: Charles C. Thomas.
Kockelman, Paul. 2007. “Agency: The Relation Between Meaning, Power and
Knowledge.” Current Anthropology 48(3): 375–​401.
Latour, Bruno. 1999. Pandora’s Hope: Essays on the Reality of Science Studies.
Cambridge, MA: Harvard University Press.
Maples, William R., and Michael Browning. 1994. Dead Men Do Tell Tales: The Strange
and Fascinating Cases of a Forensic Anthropologist. New York: Doubleday.
Owen, David. 2000. Hidden Evidence. Forty True Crimes and How Forensic Science
Helped Solve Them. Buffalo, NY; Willdowdale, ON: Quintet Publishing;
Firefly Books.
Peirce, C. S. 1931–​1935. Collected Papers. Vols. 1–​6. Cambridge, MA: Harvard
University Press.
Peirce, C. S. 1958. Collected Papers. Vols. 7–​8. Cambridge, MA: Harvard
University Press
Sawday, Jonathan. 1995. The Body Emblazoned: Dissection and the Human Body in
Renaissance Culture. London/​New York: Routledge.
Skinner, Mark F., Heather P. York, and Melissa A. Connor. 2002. “Postburial
Disturbance of Graves in Bosnia-​Herzegovina.” In Advances in Forensic
Taphonomy: Method, Theory, and Archaeological Perspectives, edited by William D.
Haglund and Marcella H. Sorg, 293–​308. Boca Raton, FL: CRC Press.
Thomas, Peggy. 1995. Talking Bones: The Science of Forensic Anthropology.
New York: Facts on File.
Wagner, Sarah E. 2008. To Know Where He Lies: DNA Technology and the Search for
Srebrenica’s Missing. Berkeley and Los Angeles: University of California Press.

T h e Ag e n c y of t h e   Dead  [ 189 ]
  191

CHAPTER 20

Distributed Agency in Play

BENJAMIN SMI TH

INTRODUCTION

When Alberto and his brothers go out to play marbles, they can’t help
but do other things. In the Peruvian, Aymara-​speaking Andes, they must
take their family’s alpacas and sheep out to far-​flung grazing areas. And,
once there, Alberto—​the oldest brother—​occasionally keeps an eye on the
animals as they drift toward uneaten pastorage. The area where they play
marbles, then, is part and parcel of a landscape that is quite unlike, for
example, the radically transformed, angular spaces where games like bas-
ketball and football (ideally) get played. Although treeless and shrubless,
it is a landscape that is dense with rocks small and large, thick clumps of
grass, short twigs, holes, eroded gulleys, and inclines and declines. In such
a case, marbles play must go where alpacas and sheep go, and virtue must
be made of this necessity. And, frankly, it is. When Alberto and his brothers
play marbles, they must shoot them through this thicket of things, aiming
for a series of small holes dug into the ground. Much of the intrigue of
the game consists, then, of whether and how these material things serve
as agents in marbles, and whether and how boys contend with the way in
which these things act. Does a rock divert a marble’s path? Does it do so
across several turns of play? What does a boy do in response? From the
perspective of an Aymara boyhood imagination, the stakes here are quite
high. Is a marbles player who cannot contend with a rock really a boy? How
might rocks make and unmake men?
192

The Aymara game of marbles (t’inka) invites a concern with a form of

agency that has rarely been object of sustained analytic attention:  the
agency of material things in childhood play and games. In his classic study
of marbles, for example, Jean Piaget gives just a passing description of
marbles and their material context before moving on to his account of the
development of moral judgment in children (1965). The object of his con-
cern are the rules of the game as well as the changing attitudes that chil-
dren have with respect to those rules. Do children think of the rules of the
game as inviolable and even sacred? Do they think of them as subject to
negotiation and compromise? In posing these questions, Piaget privileges
two dimensions of childhood games: the sense in which they are regulated
by social norms (or rules) and the sense in which they are the object of
psychological processes (in this case, morally valenced attitudes). The end
result of such an analysis, inevitably, is to foreground children themselves
as agents in play. They are the ones whose behavior gets evaluated relative
to the rules of the game, and they are the ones who interpret rules in flex-
ible ways. This is an account, then, that precludes a serious concern with
the distribution of agency across children and material things in play and
games. And, in doing so, it is a token of a type. With only a handful of
notable exceptions,1 the study of play and games takes the child as its hero.
In what follows, I give an account of the game of marbles among Aymara
boys, showing especially how material things wield an agency within it.
To do so, I focus on a specific moment in the regular sequence of marbles
play: the task of engineering or designing the marbles playing field. This is
a moment that will show how the marbles playing field gets made into a
specifically semiotic agent, an agent that, genuinely, interprets its material
context and is made to have regular consequences on its surrounding world
(including, especially, its surrounding social and symbolic world). In other
words, it is not just the case that children play with toys and the material
contexts of those toys. Material things, also, play with children.

ENGINEERING A FIELD, EMBEDDING A WORLD

When boys head out to herd animals and play marbles, they must, at
some point, transform some patch of ground into a playing field. This is a
moment that, however brief, most clearly reveals the connection between
the semiotic agency of material things and social categories like masculine,
human, and nonhuman. When carving out a playing field, boys must first
of all create a space that can accommodate the set of physical relations pro-
jected by the rules of the game. In this local variety of marbles, there must

[â•›192â•›]â•… The Agency of Materiality

  193

be four spots that serve as the goals or targets for a boy striking a marble.
These are, typically, four holes that are created to have a diameter just a
bit wider than the width of a marble. Three of these holes must conform
to a line that is more or less straight, and the fourth sits off to one side
of this line. In order to play, a boy must first advance a marble through
these holes, ending in the fourth one off to the side. Advancing a mar-
ble means striking it with one’s pointer finger and having it successfully
land inside of the targeted hole. A boy has two marbles and may continue
advancing them until he misses a target with each marble. Once a marble
has reached the fourth hole, it attains a new status. It now has venom (or
wininu). Once it has venom, it can, upon striking another marble, kill it
(literally, jiwayaña, “to kill”). Once all of one’s opponent’s marbles are dead,
this player—​the one who has at least one marble still alive—​wins, and the
game is over. Although a game of marbles technically ends when all but one
or two marbles are dead, games more frequently than not end before this
moment arrives. This happens for a number of reasons: alpacas and sheep
drift towards other places, it’s time to bring everyone home, or it’s time to
eat, etc.
The act of creating an actual playing field happens quickly, and it is man-
aged by the oldest boy in a group (in most cases, an oldest brother).2 There
is a balance that must be achieved. On the one hand, there are plenty of
sites where it would be essentially impossible to play a game of marbles
(e.g., steep hillsides, boulder-​strewn areas, areas too thick with grass). On
the other hand, there are at least a couple of areas where it would be too
easy (e.g., the flat road that runs through the community). To put it in a
semiotic idiom, finding and making an appropriate marbles playing field
is an act of phatic labor, to use Elyachar’s (2010) term.3 It is a task of find-
ing a field in which there are number of possible channels through which
the sort of thing that is a marble can be struck towards a series of holes
dug about four feet apart from one another. It must be a field in which
not all possible channels are available, but certainly a majority of possible
ones. More often than not, the challenge of creating a playing field—​at
least, for the first three of the holes—​requires a marbles player to remove
certain things that would gum up too many of the possible channels. In
the moments before a game begins, for example, an oldest boy will direct
a younger one to dig the four holes, and he himself will start to groom the
playing field. If there is a rock too large, a twig too thick and long, a piece
of trash too large (e.g., a two liter plastic bottle), then he will pick it up
and throw it aside. There is an implicit logic to this kind of act: these are
things that would likely enclose or surround a marble struck toward them
(a large rock, a long, thick stick, etc.), and these are the things that must be

Di s t r i b u t ed Ag e n c y i n   P l ay   [ 193 ]
194

removed. To the extent that some number of channels are made more fully
available, then the communicative value of the flicking of the marble—​i.e.,
its value as a strategic act within the game—​is more likely to be legible to
marbles players, onlookers, and anthropologist.
In these moments of designing a field, players are just as invested in
making sure that at least some possible channels are closed off, made less
fully available or, at least, made more challenging to traverse. Although this
is true with respect to the entire playing field, it becomes a matter of spe-
cial concern with respect to the fourth and final hole, the one that gives
a marble its venom. In the moments before play, for example, an oldest
boy—​the engineer of the field, if you will—​takes special care to locate this
last hole in an area that is more fully dense with debris or higher or lower
up on a slope, among other possibilities. If the already existing landscape
does not provide such possibilities, he will toss some rocks, twigs or trash
into the area of the fourth hole (sometimes drawing on things collected
from the other areas of the playing field). These moments are ones in which
it becomes especially clear that a playing field—​in other words, this net-
work of possible channels—​is something that is at least partly engineered
or designed. And, it is designed as a kind of sieve, in Kockelman’s sense
(2013). Depending on the specific area within the playing field, it is an
artifact that is designed to more or less easily let certain things through
(i.e., a marble) or to have those things either get stuck or be diverted. In
other words, it has been engineered to select for a certain kind of sign
(again, a marble),4 and it has been made to bring about certain kinds of
consequences—​in other words, interpretants, in Peirce’s (1955) sense—​
for that kind of thing (diverting marbles, getting them stuck, letting them
pass through). It is in this sense that the playing field counts as a complex,
internally differentiated semiotic agent, albeit one whose peculiar form of
agency has been partly embedded in it through processes of engineering
or design. In acting as such an agent, a skillfully designed marbles playing
field helps to reveal marbles shots that are more or less skillful and more
or less strategic.
The marbles field, then, is designed to act and to signify, but it also gets
understood, symbolically or ideologically, as an agent. This is implicit, even,
to the narrative arc that it instantiates. This is most fully apparent in the
design of the field that surrounds the fourth hole. The region around the
fourth hole, after all, is designed to make available fewer and more chal-
lenging channels for marbles to cross. It is strewn with debris or located on
a slope. However, it is also serves, narratively, as a moment of culmination
or, even, as a rite of institution (Bourdieu 2001). Once a marble lands in
this fourth target, after all, it attains a new status: it acquires venom, and

[ 194 ]  The Agency of Materiality

╇ 195

it can thereby kill other marbles. In this case, then, the semiotic agency of
the playing field—╉that is, its design as a sieve—╉comes to be understood
relative to a game sequence that is framed, narratively or symbolically, as a
kind of quest. A boy and his marble encounter their greatest challenge (that
is, the dense sieve that surrounds the fourth hole), and a marble is thereby
rendered a more potent and indeed venomous instrument. There is, in
other words, an analogy that is made across the kind of embedded semiotic
agency that the playing field exerts and the symbolic, narrative-╉like render-
ing of that agency. In this area of the playing field, the sieve is dense and
the challenge great. There are additional reasons for this interpretation.
When boys add rocks, twigs, and trash to the area around the fourth hole,
they often describe these things as qhincha, or “bad luck.” These things
are explicitly framed, in other words, as challenges for players:5 qhincha is,
after all, something that, according to Urton, is a “principal cause of the
emergence of a state of imbalance and disequilibrium” (1997:147). These
observations suggest a more sweeping analogy or relation of iconicity. The
playing field is not just a sieve for marbles. It is a sieve for men who con-
front the challenge of qhincha. The interpretive (or, better:  interpreting)
world that is embedded in the marbles play field is fully “commensurable”
with (Latour 2005:74)—╉or, more strongly, compatible with and necessary
for—╉the sociocultural meaningfulness of the game. Or, to capture the
insight in an aphorism: only the skillful marbles shot shall pass, and only
the tough man shall win.

CONCLUSION

Marbles is one among many games that implicate material things as agents.
A  quick perusal of Avedon and Sutton-╉Smith’s edited compendium on
games (1971) reveals a bewildering number of examples. In Brewster’s arti-
cle on games in Shakespeare, for example, he describes games that involve
seeds, dice, a leather belt, knives, an object “thrown onto the ground” (39),
a “wooden image (usually of a Turk or a Saracen) mounted on a pivot” (42),
cards, plates, a table, a shilling, marbles, and a game that consisted of “roll-
ing small balls into holes at one end of the game-╉board” (47). In a more
anthropological contribution, Erasmus describes forms of play that require
shews, bowls, beans, and small sticks, a series of “twenty holes separated
in the middle by a wider space called a river” (119), and, more familiarly,
boards that provide a flat surface. Few of these material things exert an
agency that is like the one described here:  that is, as a field of material
things that acts as a sievelike catcher of marbles and men. This does not

Di s t r i b u t ed Ag e n c y i n   Pl ay â•… [â•›195â•›]
196

contravene the analysis I’ve provided, but it does suggest that the problem
at stake is a tremendously larger one. How can we theorize the full variety
of ways in which material things (are designed to [Murphy 2013]) select
and act in games and play?
How much do these “missing masses” (Latour 1992)—╉in other words,
the material things that are not visible in many accounts of play—╉matter?
This is a question that has as many answers as there are approaches to the
study of play and games in childhood. In the case of Aymara marbles, for
example, an analysis of these missing masses is necessary for the anthro-
pological project of understanding the form of masculinity that is at stake
for Aymara boys. A number of other, more general possibilities stand out.
How do material things provide an anchor and model for forms of “make-╉
believe” or “fantasy play” in childhood? To what extent are processes of
“design” and “engineering” a part of play, and how might such processes
mediate the kinds of sociocultural worlds and cognitive processes that are
at stake in play? What kind of object is an object-╉for-╉play, and how might
its artifactuality differ from other kinds of objects and technologies? What
role do material things and other nonhuman agents play in producing the
category “human”? Piaget might have also asked: how might the relative
adaptability and design-╉ability of material things in play help to socialize
the kind of negotiability that Piaget—╉and, increasingly, anthropologists—╉
see as being central to our moral and social lives?

NOTES

1. I should mention here some of the work inspired by Vygotskyan or Latourian

frameworks, especially authors like Goodwin (2006) and Brembeck (2011).
2. See Smith (2014) for an account of the caretaking relationships presupposed in
this example.
3. Elyachar draws upon Jakobson’s refinement of the concept of phatic function.
These are acts that serve to “establish, prolong, or discontinue communication”
(Jakobson 1960:350).
4. One complexity I have had to gloss over here are the “inter-╉material” relations
between a marble and its playing field. For instance, the agency of the playing
field makes sense only relative to the sort of thing that a marble is and the sort
of action that a marble undertakes (i.e., after being struck).
5. See Smith (2010) for a longer account of bad luck in Aymara marbles play.

[â•›196â•›]â•… The Agency of Materiality

╇ 197

REFERENCES

Avedon, Elliott M., and Brian Sutton-╉Smith. 1971. The Study of Games.
New York: Wiley.
Bourdieu, Pierre. 2001. Masculine Domination. Stanford, CA: Stanford
University Press.
Brembeck, Helene. 2011. “Toys Matter. McDonald’s Toys and Nordic Childhoods.”
Revue D’Histoire Nordique 12: 161–╉178.
Brewster, Paul. 1971. “Games and Sports in Shakespeare.” In The Study of Games,
edited by Elliott M. Avedon and Brian Sutton-╉Smith, 27–╉47. New York: Wiley.
Elyachar, Julia. 2010. “Phatic Labor, Infrastructure, and the Question of
Empowerment in Cairo.” American Ethnologist 37(3): 452–╉464.
Erasmus, Charles J. 1971. “Patolli, Pachisi, and the Limitations of Possibilities.” In The
Study of Games, edited by Elliott M. Avedon and Brian Sutton-╉Smith, 109–╉129.
New York: Wiley.
Goodwin, Marjorie H. 2006. The Hidden Life of Girls: Games of Stance, Status, and
Exclusion. Malden, MA: Blackwell.
Jakobson, Roman. 1960. “Closing Time: Linguistics and Poetics.” In Style in Language,
edited by Thomas Sebeok, 350–╉359. Cambridge, MA: MIT Press.
Kockelman, Paul. 2013. Agent, Person, Subject, Self: A Theory of Ontology, Interaction,
and Infrastructure. Oxford: Oxford University Press.
Latour, Bruno. 1992. “Where Are the Missing Masses? Sociology of a Few Mundane
Objects.” In Shaping Technology, Building Society: Studies in Sociotechnical
Change, edited by Wiebe Bijker and John Law, 225–╉258. Cambridge,
MA: MIT Press.
Latour, Bruno. 2005. Reassembling the Social: An Introduction to Actor-╉Network Theory.
Oxford: Oxford University Press.
Murphy, Keith M. 2013. “A Cultural Geometry: Designing Political Things in Sweden.”
American Ethnologist 40(1): 118–╉131.
Peirce, Charles S. 1955. “Logic as Semiotic: The Theory of Signs.” In Philosophical
Writings of Peirce, edited by Justus Buchler, 98–╉119. New York: Dover.
Piaget, Jean. (1932) 1965. The Moral Judgement of the Child. New York: Free Press.
Smith, Benjamin. 2010. “Of Marbles and (Little) Men: Bad Luck and Masculine
Identification in Aymara Boyhood.” Journal of Linguistic Anthropology
20(1): 225–╉239.
Smith, Benjamin. 2014. “Metacultural Positioning in Language
Socialization: Inhabiting Authority in Informal Teaching among Peruvian
Aymara Siblings.” Linguistics and Education 25: 108–╉118.
Urton, Gary. 1997. The Social Life of Numbers: A Quechua Ontology of Numbers and
Philosophy of Arithmetic. Austin: University of Texas Press.

Di s t r i b u t ed Ag e n c y i n   Pl ay â•… [â•›197â•›]
  199

CHAPTER 21

Contingency and the Semiotic

Mediation of Distributed Agency
EI TAN WILF

I n this chapter I  theorize the role of nonhuman entities not simply in

extending human agency but rather in providing humans with the train-
ing ground to cultivate basic dimensions of human agency understood to
be lacking in the specific ethnographic contexts I  examine. In doing so,
I aim to provide a window into a relatively under-​theorized dimension of
the idea that agency is distributed across the human/​nonhuman divide.
I draw on fieldwork I conducted in a number of ethnographic contexts: a US
academic jazz music program, and two US research institutes in which jazz-​
improvising computerized algorithms are developed. In these ethnographic
contexts human creativity constitutes a problem—​a scarce resource—​
which people try to address and resolve in different, though conceptually
relatable, ways. Broadly speaking, a recurring problem in all of these con-
texts is that of the predictability of people’s purposes or intentions, which
prevents them from being creative, in other words, from following a cul-
turally specific normative ideal—​a Romantic modern one—​that opposes
standardization of purposes. In one of these contexts, an additional prob-
lem is that of people’s inability to reliably execute via appropriate behavior
whatever purposes or intentions they have, whether creative or not. In all
of these contexts such problems are understood as lack of creative agency,
and a prevalent solution is the use of contextually meaningful contingency
produced by and harvested from different nonhuman entities (Wilf 2013).
200

Sign Interpretant

correspondence

Object

Figure 21.1  A Peircean Semiotic Framework.

I will use a Peircean semiotic framework to theorize distributed agency

as a property that is semiotically mediated across the human/​nonhuman
divide (Peirce 1955). In such a framework, semiosis consists of three ele-
ments (Figure 21.1): a sign (whatever stands for something else); an object
(whatever a sign stands for); and an interpretant (whatever a sign creates
insofar as it stands for an object). In a semiotic process, a sign stands for its
object, on the one hand, and its interpretant, on the other hand, in such a
way as to make the interpretant stand in relation to the object in a way that
corresponds to its—​in other words, the sign’s—​own relation to the object
(Kockelman 2005).
In the first ethnographic site I examine, the crisis of creative agency con-
cerns the failure to align private purposes with public actions. The site is
a US academic jazz music program, Berklee College of Music in Boston,
which is part of the growing academization and institutionalized profes-
sionalization of jazz training that have culminated in the establishment
of hundreds of academic programs that grant various degrees in jazz per-
formance (Wilf 2014). As an improvised form of music, jazz consists of
the real-​time generation and execution of “musical ideas,” as it were. To
function well, players must be able to flawlessly execute their constantly
emerging ideas with their instruments and respond to the musical cues
provided by their band members in the course of performance. Canonical
definitions of “creative thinking” in jazz focus precisely on the need for
mature jazz improvisers to “attain a perfect unity of conception with the body.
The artist becomes intensely focused on thoughts in the language of jazz,
and as they come—​one upon the other—​they are articulated as instantly
as conceived. No lead time separates conception from expression, and the gap
between intention and realization disappears” (Berliner 1994:217; emphasis
added). At stake, then, in this particular sort of framing, with its particular
ethnopsychology of intention/​action, is the skill to execute one’s real-​time
musical ideas with one’s instrument, to match private auditory representa-
tions with public action.

[ 200 ]  The Agency of Materiality

  201

Sign Interpretant
(“false” playing) (not knowing if
sign and “true”
object align)

False Object False Object

“True” Object
(internal auditory representations)

Figure 21.2  Determining Sign-​Object Relations: Problems of Execution.

To cultivate this skill, players need to condition their playing bodies

through intensive listening and playing in real-​time performance situa-
tions rather than through the mediation of the written score. However,
because of specific transformations in jazz’s conditions of existence, it
has become more difficult for students to train themselves in live perfor-
mance situations. Specifically, the dwindling demand for jazz since the
1950s has resulted in the gradual disappearance of performance venues
in which neophyte musicians can inform their competence with listen-
ing and apprentice with experienced practitioners in real-​time perfor-
mance situations. The educators and students I worked with argued that
although students try to compensate for this shift by orchestrating inde-
pendent sessions, they do so less than before because they must compete
over limited rehearsal space or hold sessions in their apartments, which
poses significant practical limitations (Wilf 2014:202–​209). Even more
important, these sessions, when they do take place, do not make up for the
loss of opportunities to apprentice with seasoned musicians. Such appren-
ticeship has gradually given way to abstract, formalized, and rationalized
training in the hundreds of academic programs that have come to function
as the pillars of the jazz world over the last few decades. The educators
I worked with argued time and again that as a result of these institutional
transformations—​the shift of jazz training from the clubs to the classroom
where standardized, printed pedagogical aids have become a key mode of
acquiring knowledge—​students fail to cultivate the skill that would allow
them to execute their real-​time, emergent musical ideas with their instru-
ments (Wilf 2014:189–​199).
At stake here is a form of reduced agency that concerns one’s reduced flex-
ibility in determining sign-​object relations (Figure 21.2). If we understand a
note played by a musician as a sign whose mediated object for an intended
listener is supposed to be a specific internal auditory representation, then

C o n t i n g e n c y a n d TH E Se m i o t i c Medi at i o n of DISTRIBUT E D Ag e n c y   [ 201 ]

202

Sign Interpretant
(playing) (knowing if sign
and object align)

Object
(from internal auditory representations to
publicly available random sonic input)

Figure 21.3  Determining Sign-​Object Relations: Addressing Execution.

students have reduced flexibility in terms of their ability to determine this

kind of sign-​object relations. They have a hard time mediating to others via
signs—​in other words, via their playing—​objects, that is, specific musical
ideas. They often end up playing other (that is, “false”) notes than those they
intended to play, thereby mediating to their listeners false objects (that is,
false intentions). This is a problem of execution (Enfield 2013:104–​112).
Think of it in terms of an agent not being able to engage in purposeful
behavior.
To enhance this specific dimension of flexibility, the students I worked
with often turned to contingency produced by nonhuman entities. They
engaged in gamelike interactions in which they competed with one another
over who was better able to match with his or her instrument random envi-
ronmental sonic stimuli. During my fieldwork it was not uncommon to see
students hanging out in class until a sound such as a car horn would infil-
trate the classroom from the street and immediately prompt them to try
to match it with their instruments, and then to engage in debates about
who managed to match it first and whether it counted as an accurate and
satisfying match. Indeed, one evening, while approaching Inman Square in
Cambridge, Sara, a vibraphone student, and I heard a car horn next to us
and immediately tried to match it by whistling. Because each of our whis-
tles represented a different pitch, Sara ordered me to hurry and ask the
driver sitting in the car to honk again so we could establish who was right.
(She was.)
I suggest that two related things happen here. First, these random sonic
stimuli come to function as public proxies for internal auditory representa-
tions whose execution students can practice (Figure 21.3). In other words,
the environment becomes a training ground for cultivating one’s ability to
determine sign-​object relations. The relatively contingent nature of these
stimuli is crucial because it simulates the unpredictable dimension of jazz

[ 202 ]  The Agency of Materiality

  203

improvisation as an emergent communicative event that requires musi-

cians to respond to and execute unpredictable musical input. Note, then,
the turn to nonhuman entities not for the purpose of extending human
agency—​which is underdeveloped here—​but for cultivating it, to begin
with, in the sense that the replication of an external pitch is a form of prac-
tice in executing an internal auditory representation on one’s instrument,
in other words, in being able to engage in a specific form of purposeful
behavior.
Second, the public nature of these random sonic stimuli makes the
assessment of another person’s ability to determine sign-​object relations
publicly available for everyone present. Whereas normally it would be diffi-
cult for me to know if another person has the skill to determine sign-​object
relations because my ability to access the objects in this case—​in other
words, his or her specific internal auditory representations—​and com-
pare them to the signs this person gives off—​that is, his or her playing—​
depends on this person’s skill to mediate these objects for me, to begin
with, now it is easy for me to ascertain whether he or she has this skill or not
because these representations have become public: their place is now taken
by publicly available sonic stimuli, which allows me to assess if there is an
indexically iconic relation between these stimuli and the person’s playing.
An indexically iconic relation between a publicly available random sound
and a note played by another person immediately after in an attempt to
match it—​this relation itself now becomes a sign that mediates for me this
person’s ability to determine sign-​object relations. These gamelike interac-
tions are thus metasemiotic: they concern the semiotic mediation of one’s
ability to semiotically mediate, in addition to being a training ground for
cultivating this ability.
A second dimension of students’ reduced flexibility and agency, which
I recorded during my fieldwork, is that of planning: deciding which sign-​
object relations are to be created in the first place (see Enfield 2013:104–​
112). Because of their training, which is heavily mediated by textual
artifacts of standardized patterns of improvisation, many students find it
hard to come up with creative musical ideas to begin with, regardless of their
ability to execute them with their instruments in the real time of perfor-
mance (Figure 21.4). Here the focus of attention is not execution, but what
to execute. Teachers often described their students as mere “button push-
ers” and “pattern players,” using tropes that connote the mechanical valves
of an engine to describe this lack of creative agency (Wilf 2014:167–​169).
To address the problem of standardized ideas in many jazz pedagogical
settings, James, a computer scientist, who is also a semiprofessional jazz
musician, developed computerized algorithms that purport to improvise

C o n t i n g e n c y a n d TH E Se m i o t i c Medi at i o n of DISTRIBUT E D Ag e n c y   [ 203 ]

204

Sign Interpretant
(playing) (disappointment)

Object
(standardized ideas)

Figure 21.4  Determining Sign-​Object Relations: Problems of Planning.

jazz in response to the real-​time improvisation of a human player. The

music generated by the algorithms is played by a humanoid robot marimba
player. It is an interactive system, that is, “sociable,” in that it produces
music in response to the music produced by the human musicians interact-
ing with it in real time.
In an interview I conducted with James in his lab in a major US institute
of technology, he attributed the added value of his system to the fact that it
generates musical ideas that humans could never generate by themselves.
Different kinds of algorithms—​mainly Markov models, genetic algorithms,
and fractal-​based algorithms—​generate music using complicated calcula-
tions that incorporate both stochastic and deterministic (nonlinear) pro-
cesses into their logic. Contextually meaningful contingency is a key part
of their logic, what allows them to introduce the novelty that human play-
ers presently lack. By contextually meaningful, I mean that the contingency
generated by the algorithms and played by the robot is not meaningless.
Rather, the music makes sense because it is generated in the context of spe-
cific tunes and their harmonic progressions, as well as in response to the
musical ideas of the people playing with the robot. It is patterned meaning-
fulness in which current notes make sense given past notes and relative to
genre-​based conventions. James’s hope is that by playing with this system,
which generates new creative ideas, people will be inspired to modify their
own repertoire of hitherto standardized intentions (Figure 21.5). Playing
with this system is thus supposed to help people to cultivate their ability to
plan more interesting or creative sign-​object relations, and thus to increase
their flexibility and agency.1 Motivated by the widespread discontent about
standardization in the field of jazz education, as well as the significant mar-
ket for jazz pedagogical aids, some of the scientists I worked with discussed
the commercial possibilities of integrating the algorithms they developed
in different jazz pedagogical technologies that would allow students to
improvise with these technologies and receive real-​time musical input

[ 204 ]  The Agency of Materiality

  205

Sign Interpretant
(playing) (inspiration)

Object
(from standardized to novel ideas
inspired by computerized algorithms)

Figure 21.5  Determining Sign-​Object Relations: Addressing Planning.

based on the “falsifiability” of who they are in terms of their improvisa-

tion style, in other words, to receive musical input that would not align
with their standardized style of improvisation. As in the case of the turn to
publicly available random sonic stimuli, the motivation for turning to com-
puterized algorithms is that they are nonhuman entities that can function
as sources of contextually meaningful contingency. This contingency is not
used to extend human agency—​which is lacking, here in the dimension of
planning rather than execution—​but to provide the training ground for its
cultivation, to begin with.
To add another dimension to my argument about the role of contin-
gency in the semiotic mediation of distributed agency across the human/​
nonhuman divide, note that there are numerous ethnographic contexts in
which contingency produced by and harvested from nonhuman entities is
mobilized as a resource for increasing one’s flexibility and agency by reduc-
ing those of other humans. Consider some forms of mechanical divina-
tion, a practice that has long been studied by anthropologists. Mechanical
divination—​in other words, the use of aleatory processes to divine hid-
den events and causes—​is frequently used to reveal group members’ mali-
cious intentions presumed to be responsible for different kinds of harmful
incidents. In many contexts in which mechanical divination is performed,
people often attribute to other people’s ill intentions what many mod-
ern subjects would attribute to coincidence. Danger emanates from the
increased agency of one’s group members to surreptitiously orchestrate
harmful incidents while maintaining the appearance of innocence.
To reduce the flexibility of one’s group members, which can be articu-
lated as flexibility in a specific dimension I will soon explain, people turn to
contingency produced by nonhuman entities, and they do so in the form of
mechanical divination. For example, the Azande divinatory practice called
the “poison oracle,” famously analyzed by E.  E. Evans-​Pritchard (1991),

C o n t i n g e n c y a n d TH E Se m i o t i c Medi at i o n of DISTRIBUT E D Ag e n c y   [ 205 ]

206

has been discussed as an “aleatory” mechanism individuals use to find out

who has bewitched them or intends to do so (Du Bois 1993:58). In this
practice, poison is administered to a fowl after asking a yes-​or-​no question.
The fate of the fowl—​whether it lives or dies—​provides the answer to the
question. The advantage of this mechanism is that “the meanings arrived
at are determined by something other than a volitional, human act” (Du
Bois 1993:54). This feature allows the poison oracle to function as a reli-
able or unbiased “method of revealing what is hidden” (Evans-​Pritchard
1991:120), namely, the malicious intentions of one’s group members, with
which this agonistic cultural context is overflowing.2 Although individuals
do not explicitly treat the poison oracle as a randomizer but rather assign
it mystical properties (Evans-​Pritchard 1991:147–​151), the administration
of poison to fowls does, in fact, amount to the production of a randomizer.
To ensure that it acts as such, “as soon as the poison is brought back from
its forest home it is tested to discover whether some fowls will live and
others die under its influence” (Evans-​Pritchard 1991:158). A poison that
systematically kills all fowls or has no effect on any of them is tantamount
to a die that falls on one face time and again: it is not a proper random-
izer in that, all things being equal, it is not governed by or involving equal
chances for each of the actual or hypothetical members of a population; its
outcomes are known in advance. Here, then, as in the contexts I discussed
before, contingency produced by and harvested from nonhuman entities
is mobilized as a cultural resource, only this time for reducing, rather than
increasing, agency.
I suggest that this form of mechanical divination is supposed to have its
effects in a specific dimension of another person’s flexibility and agency,
namely, “subprehending,” in other words, the ability of a person to effec-
tively anticipate the interpretants of his or her behavior (see Enfield
2013:104–​112). This is because ill intentions in these ethnographic con-
texts are often (though not always) masked, dissimulated, and hidden
behind behavioral signs that mediate for other people benign or even good
intentions (Figure 21.6). In other words, the flexibility and heightened
agency of a person who ill intends against other people often find expres-
sion in his or her ability to have one purpose but to deliberately give off
signs that mediate another purpose for other people. Such a person can
effectively anticipate and orchestrate the interpretants his or her publicly
available behavior will create and cause other people to experience by effec-
tively concealing his or her true purpose and mediating a false one. Such a
person has mastered execution, planning, and subprehension.
Mechanical divination, in turn, undermines this person’s ability to sub-
prehend because it reveals the true purpose—​in other words, the true

[ 206 ]  The Agency of Materiality

  207

Sign Interpretant
(smile) (being fooled)

“False” Object “True” Object

(good intensions) (ill intensions)

Figure 21.6  Anticipating Interpretants: Strategically Subprehending in Witchcraft.

Sign Interpretant
(smile) (revealing
another’s “true”
intensions via
mechanical
divination)

“False” Object “True” Object

(good intensions) (ill intensions)

Figure 21.7  Diagnosing Interpretants: Undermining Subprehending in Witchcraft.

object—​that lies behind his or her publicly available behavior—​that is,

behind the signs he or she publicly gives off (Figure 21.7). Mechanical divi-
nation allows concerned people to experience interpretants of the publicly
available behavior of an ill-​intending person, which are different from the
interpretants this ill-​intending person intends his or her publicly available
behavior to create. It is a form of diagnostics that determines actual inten-
tions from evinced intentions, allowing people to gain theoretical agency
(knowledge) over that which reduced their practical agency (actions) and
thus to try to ameliorate their situation.
To conclude, it is typically understood that humans are more agen-
tive than nonhuman entities because they are more flexible in the vari-
ous dimensions of agency discussed above. However, as I have argued in
this chapter, in some ethnographic contexts humans face the problem of
reduced flexibility in terms of their ability to compose their behavior, in the
sense of both execution and planning. They might consequently try to capi-
talize on the capacity of some nonhuman entities to function as sources of
contingency that can be used to cultivate flexibility in these two dimen-
sions. Although my ethnographic focus is a very specific form of human
agency, namely, artistic creation as understood by a specific group of

C o n t i n g e n c y a n d TH E Se m i o t i c Medi at i o n of DISTRIBUT E D Ag e n c y   [ 207 ]

208

artists, it can be extended to other, often more mundane, forms of agency.

For example, one can think of the use of the “random shuffle” function on
an iPod as a case in which contextually meaningful contingency is produced
by and harvested from a nonhuman entity (a computerized shuffling algo-
rithm) as a way of avoiding standardization of planning—╉here the choice
about the order of songs one is going to listen to. And one can think of the
“random oscillation” function of a tennis ball machine as a case in which
contextually meaningful contingency is produced by and harvested from a
nonhuman entity as a way of improving one’s ability to execute one’s pur-
poses in a real-╉time situation.3
At stake here is not the extension of human agency by nonhuman enti-
ties, as when I use a bike to cover a greater distance, but the use of non-
human entities for cultivating and enabling basic dimensions of human
agency, to begin with, in particular, the generation of novel purposes and
the means to execute them. Such a use can have different orientations
depending on culturally specific factors. Humans can use the contingency
produced by nonhuman entities to increase their own agency and flexibil-
ity, as in the cases of the jazz school and the jazz-╉improvising computer-
ized algorithms, or to increase their own agency and flexibility by reducing
those of others, as in the case of mechanical divination. In all of these
examples, semiotic mediation emerges as a key mechanism underlying the
distribution of agency across the human/╉nonhuman divide, and as a condi-
tion of possibility for its proper analysis.

NOTES

1. At the time of my fieldwork these algorithms were a work in progress and not
always successful in achieving their goal. I remember many instances of playing
with the system they animated and being uninspired by its responses to my own
(frequently uninspiring) improvisations. Lack of inspiration typically resulted
when the system generated responses that were either too predictable or too
unpredictable (Wilf 2013:614–╉615).
2. At the same time, “A Zande does not readily accept an oracular verdict which
conflicts seriously with his interests,” and “apart from criminal cases, there can
be no doubt that a man takes advantage of every loop-╉hole the oracle allows him
to obtain what he wants or to refrain from doing what he does not want to do”
(Evans-╉Pritchard 1991:163).
3. As one tennis training machine manufacturer puts it, “Random oscillation is
where the tennis ball machine replicates a match situation by shooting balls
randomly around the court. Rather than groove a particular shot, this allows a
player to … practice shots in a more realistic environment.” See http://╉www.
spinshot-╉sports.com/╉Knowledge.html, accessed April 5, 2015.

[â•›208â•›]â•… The Agency of Materiality

╇ 209

REFERENCES

Berliner, Paul. 1994. Thinking in Jazz: The Infinite Art of Improvisation.

Chicago: University of Chicago Press.
Du Bois, John W. 1993. “Meaning Without Intention: Lessons from Divination.” In
Responsibility and Evidence in Oral Discourse, edited by Jane H. Hill and Judith
T. Irvine, 48–╉71. Cambridge: Cambridge University Press.
Enfield, N. J. 2013. Relationship Thinking: Agency, Enchrony, and Human Sociality.
Oxford: Oxford University Press.
Evans-╉Pritchard, Edward E. (1937) 1991. Witchcraft, Oracles, and Magic among the
Azande. London: Clarendon Press.
Kockelman, Paul. 2005. “The Semiotic Stance.” Semiotica (1–╉4): 233–╉304.
Peirce, Charles S. 1955. “Logic as Semiotic: The Theory of Signs.” In Philosophical
Writings of Peirce, edited by J. Buchler, 98–╉119. London: Dover.
Wilf, Eitan. 2013. “Sociable Robots, Jazz Music, and Divination: Contingency as
a Cultural Resource for Negotiating Problems of Intentionality.” American
Ethnologist 40(4): 605–╉618.
Wilf, Eitan. 2014. School for Cool: The Academic Jazz Program and the Paradox of
Institutionalized Creativity. Chicago: University of Chicago Press.

C o n t i n g e n c y a n d TH E Se m i o t i c Medi at i o n of D ISTRIBUT E D Ag e n c y â•… [â•›209â•›]

  211

PART NINE

The Place of Agency

  213

CHAPTER 22

Place and Extended Agency

PAUL C. ADAMS

T his chapter explores a few of the ways in which place participates in

distributed agency. Drawing on geographic theories, literature and
philosophy, I offer emplaced action as action that transcends the taken-​for-​
granted boundaries of mind and body. This perspective breaks from the
natural attitude (Husserl 1982:5) since we commonly think of ourselves
as autonomous actors. According to a long tradition, we think of place as
a simple vessel or site for agency while neglecting the ways in which place
profoundly affects what we are able to do.
According to Aristotle, place is a container with interior boundaries
larger than what is contained. Newton understood place as a subdivision
of space and also as a location. Descartes developed the notion that every-
thing can be situated relative to a coordinate plane. These models share a
focus on location, since an area includes all locations that are contained
within given boundaries, and locations can be easily defined as positions in
an X-​Y grid or a three-​dimensional volume. Place is still understood quite
often in these classical terms as a geometrical phenomenon, namely a loca-
tion or a bounded space, but when we pause to reflect on lived place, we
find much more than a container, area, or point. We discover a manifesta-
tion of self that seems profoundly internal, yet perplexingly “out there”
(Tuan 1977).
The elements of place associated with distributed agency are revealed
by the phrase to take place—​a phrase suggesting that place is temporar-
ily seized and claimed by action rather than being separable from action.
214

We also say that actions are undertaken or carried out, implying that places
are fields for the externalization of knowledge, expectation, understand-
ing, and other elements of self. Combining these phrases, action seizes and
appropriates places while necessarily externalizing phenomena and quali-
ties originally perceived as internal.

FOUNDATIONS OF EMPLACED AGENCY

As Robert Sack has argued, the intimate relationship between person and
place is reciprocal: “Places need the actions of people or selves to exist and
have effect. The opposite is equally true—╉selves cannot be formed and
sustained or have effect without place” (1997:88). To offer a few concrete
examples:  an orator is subject to the visual and acoustic properties of a
lecture hall; a soldier fights on a battlefield structured in terms of sight
lines, slopes, and cover; a chemist needs a laboratory that permits the iso-
lation and recombination of precise amounts of various elements; an art-
ist’s studio provides tools and space for making objects of aesthetic value.
I offer the term emplaced action to designate action involving place in an
integral fashion. Emplaced action integrates the concrete and the abstract,
the subjective and the objective, the local and the global. All actions involv-
ing people are emplaced, although not all emplaced actions involve people.
Geographers increasingly follow Doreen Massey (1993) in conceptualiz-
ing places as constellations of flows rather than bounded areas. In analogous
fashion, people can be understood as extensible beings who both sense the
world at a distance and act on distant places, dynamically extending them-
selves outward through space-╉time (Adams 1995, 2005). Thus, emplaced
action is not contained action. Rather it links near and far aspects of the
world in particular ways that fluctuate from moment to moment. The ora-
tor speaking in the lecture hall incites a protest in a public square, against
a war fought in a distant country, according to orders signed in a legislative
chamber, involving a chemical weapon developed in a laboratory, inspiring
an artist cloistered in a studio, whose painting will circulate from gallery
to gallery. Every phase of action is linked to other actions through places,
and places connect actions to other actions in particular topologies and
configurations. Some of this action is physical, like fighting and painting,
while some of it is verbal, like orating and legislating. In either case, human
action becomes extensible through space by virtue of its active appropria-
tion of place.
To think of place in this way requires us to “get back into” place (Casey
1993), to recognize ourselves as fundamentally geographical in nature.

[â•›214â•›]â•… The Place of Agency

╇ 215

Scholarship is needed that breaks down the methodological, epistemo-

logical, and ontological boundaries, splitting apart the various domains of
emplaced action—╉physical/╉material, social/╉ethical, cognitive/╉discursive
(Sack 1997). Such theorizations of place-╉based agency are indebted to
Martin Heidegger whose concept of Dasein (being-╉in-╉the-╉world) posited
that the objects encountered in daily life are understood with regard to their
usefulness in relation to particular purposes and actions. On this account,
a place is “something to which our direct intervention gives rise”—╉a phe-
nomenon in which we are involved instrumentally (Casey 1997: 247, 250).
Dasein indicates a realm of possibility where things are inconspicuous and
unobtrusive as long as they can be perceived in terms of their uses in each
particular place. Moved to a new place, an object changes with regard to its
meaning whether we consider the transmogrification of Marcel Duchamp’s
infamous “Fountain” (a urinal turned 90 degrees, signed then shown in a
gallery) or the more subtle shift of a steel coffee can when it is emptied,
removed from the kitchen, and reused in the office as a pencil holder. The
dependency of action on place is exposed when our expectations about
things clash with preexisting notions of a place-╉based material/╉social/╉
symbolic order. Such dislocations reveal the classic, location-╉based formu-
lations of Aristotle, Newton, and Descartes as merely “the shrunken resi-
due of place” (Casey 1997:251).

PLACE AND FLEXIBILITY

Places bring flexibility to agency regardless of their scale. Consider the

workshop of a man who sells handmade furniture and takes temporary
jobs in construction, who is also rebuilding an old sailboat and struggling
to keep his vintage BMW motorcycle running. His workshop contains
drills, sanders, clamps, and levels; grease for the motorcycle; epoxy for the
boat; and many other tools and substances that enable various projects. Or
consider the studio of a sculptor. Her time there is spent cutting, chiseling,
sanding, brushing, polishing, and so on—╉but always with help from the
worktables, tools, materials, and light in the studio. Both workshop and
studio function as complex translation machines, turning bodily capaci-
ties into forces far more energetic and diverse than the human body could
produce on its own. This emplacement in turn lets people not only act on
things, but also become the selves they imagine or wish themselves to be.
Similar arguments can be made about other places. Houses, for exam-
ple, foster different types of subjectivity. Geographer Yi-╉Fu Tuan calls the
house “an architectural embodiment of social structure and values” because

P l a c e a n d E x t e n ded   Ag e n c y â•… [â•›215â•›]

216

it “contains, concentrates, and focuses the human psyche” (1982:52). In

Segmented Worlds and Self (1982) he explains that the typical European
home from the Middle Ages was a hall-​like structure, lacking clear-​cut dis-
tinctions between public and private space, whereas in contrast a modern
European or American home has many gradations between public and pri-
vate, including places to sequester oneself as well as various semi-​public
places for entertaining and socializing. Modernization involved increas-
ing segmentation of both the self and the domestic environment, since
elements of self derive from domestic routines, the potent symbolism of
the home, and the spatial coordination of life within the dwelling place
(Bachelard 1964). Each dwelling functions as a particular kind of “mega-
body” (Casey 1997:291), transforming, magnifying, and blending the
agency of particular selves.
Moving to a larger scale, the city can be seen as a machine where peo-
ple are “all being shot backwards and forwards … to make some pattern”
(Woolf 2002:24). The spatial organization of the city combines with vari-
able demands on people’s time, levels of access to transportation, and
rules about who can be where at what time. Hägerstrand’s time-​geography
(1970) frames these observations in terms of coupling constraints, capa-
bility constraints, and authority constraints. People inhabit particular
space-​time “prisms” (Carlstein 1982) that limit where they can go and what
projects they can join.
Emplaced urban action includes people who have legal or conventional
claims to the places they occupy, as well as immigrants, vagrants, tourists,
and others whose rights to place are disputed. Therefore urban emplace-
ment includes ideas about who belongs in any given place, and conversely
who is “out of place” as members of certain groups are symbolically linked
to threats such as pollution, disease, indecency, abnormality, criminality,
and disorder (Cresswell 1996). Such labels suggest that some people are
able to act with the urban environment while others are mainly acted on
by that environment. Material and symbolic boundaries work together
to shape the collective agency of the city, forming what geographers refer
to as the socio-​spatial dialectic (Soja 1980) and suggesting that places are
complex tools for producing social stratification. Virginia Woolf’s essay A
Room of One’s Own (2002) demonstrated that the creative output of women
remained hidden from the world precisely because women had been denied
private places to call their own. Luce Irigaray (1993) provided an interest-
ing complement, showing how women’s bodies have been appropriated
as places, through externally controlled intrusions and appropriations.
“Intellectual freedom depends upon material things” (Woolf 2002:88), and
these things include both places and emplaced bodies.

[ 216 ]  The Place of Agency

╇ 217

In short, place is implicated in action because it offers not only material

resources such as a workbench or drawing table, but also social scaling tools
ranging from bedrooms to houses to city streets, and symbolic resources
such as white picket fences, welcome mats, and “No Loitering” signs. These
processes enhance the flexibility of agency across a wide range of scales,
though not to the same degree for all occupants, suggesting the need to
consider broader implications of agency distributed in place.

PLACE AND ACCOUNTABILITY

This brings us to the second key ingredient of distributed agency—╉

accountability. To broach this discussion we need to consider what geog-
raphers call “sense of place,” a feeling of “identification with a place
engendered by living in it” (Agnew and Duncan, 1989:2). A strong sense of
place is an affective bond, which Tuan evocatively calls topophilia (1974).
Environmental psychologists carefully parse the differences between place
meaning and place attachment (Manzo and Devine-╉Wright 2013)  but
geographers tend to view these as complementary aspects of the same phe-
nomenon. In any case, we are dealing with issues including attachment to
one’s home and attitudes of nationalism, as well as the surprisingly inti-
mate connections between these variously scaled senses of place (Morley
2000; Smith 2012).
When we love a place we are inclined to justify actions in terms of the
place’s (perceived) meaning, and this framing of accountability radically
alters the calculus of right and wrong. For the love of my home I may set
mousetraps and install a security system. These actions may require adjust-
ments of my ethical system if I like mice (just not the ones in my walls)
and if I believe the destitute deserve public assistance (just not from my
refrigerator). The house needs to be maintained, I may argue, and it is home
maintenance (not greed) that justifies the use of locks and mousetraps. The
same logic can be inflated to justify what Mike Davis calls the “security-╉
driven logic of urban enclavization” (1990:  223, 244), which militarizes
city life with gated communities, “bum-╉proof benches,” closed-╉circuit sur-
veillance systems, and so on (Davis 1990: 235, 248). Such acts of exclusion
by the urban environment are treated as a form of care—╉for the city—╉
rather than as a calculating refusal to care—╉for people.
Moving up the scale even further, we can see the most extreme expres-
sions of territorial defensiveness in patriotism, militarism, and the harden-
ing of state borders. Uncomfortable moral questions about the innocence
of individuals can be avoided by situating guilt at the level of a foreign state,

P l a c e a n d E x t e n ded   Ag e n c y â•… [â•›217â•›]

218

then asserting the right of one’s own state to carry out punishment on that
foreign state. The nation is to its territory as the self is to the human body,
and this parallel is so strong that bodies and nations symbolically overlap.
Righteous indignation subsumes human casualties. People learn to iden-
tify the nation as an immense megabody in which their personal agency
can be aligned with the agency of many other people and take on superhu-
man proportions while absolving individuals of accountability for causing
suffering and death to members of that set of humans labeled “the enemy.”
Popular culture can make this idea strikingly literal through nationalist
superheroes like Captain America, a graphic “rescaling icon” internalized by
children to form “a bridge between collective identity and the individual”
(Dittmer 2007:401). This cartoon literalizes the body politic, a notion that
dates from the 15th century and is evident in Hobbes’s writings. In modern
critiques such processes of internalization and externalization are linked
to bio-╉power (Foucault 1978). Not only is agency distributed to a place, but
personal interests are deemed to be subordinate to a “national interest” that
is equated with enforcing certain geographical boundaries (Newman and
Paasi 1998). The needs of this collective body may be imagined/╉imaginary
(Anderson 1991), but they guide militaristic and nationalistic actions that
destroy hundreds, thousands, or even millions of real human bodies.

SUMMARY

Place’s agency is evident through the proliferation and diversity of place-╉

based actions. Place also displays agency by absorbing accountability such
that people can act directly against the interests of others without the
normally attendant sense of guilt for having caused another’s suffering.
Topophilia is the common strand between these complementary aspects of
emplaced action, as “home” distributes agency and accountability at many
scales. Whereas a workshop or studio is reassuring in its homey familiarity
and homely usefulness, a gated community demonstrates the exclusionary
defense of home and a military front seems like the bold advance of an
idealized homeland.

REFERENCES

Adams, P. C. 1995. “A Reconsideration of Personal Boundaries in Space-╉Time.” Annals

of the Association of American Geographers 85(2): 267–╉285.

[â•›218â•›]â•… The Place of Agency

  219

Adams, P. C. 2005. The Boundless Self: Communication in Physical and Virtual Spaces.

Syracuse, NY: Syracuse University Press.
Agnew, J. A., and J. S. Duncan, eds. 1989. The Power of Place: Bringing Together
Geographical and Sociological Imaginations. Oxford/​New York: Routledge.
Retrieved from http://​www.eblib.com.
Anderson, B. 1991. Imagined Communities: Reflections on the Origin and Spread of
Nationalism. 2nd ed. London/​New York: Verso.
Bachelard, G. 1964. The Poetics of Space. Translated by Maria Jolas.
Boston: Beacon Press.
Carlstein, T. 1982. Time Resources, Society and Ecology: On the Capacity for Human
Interaction in Space and Time. London: Allen & Unwin.
Casey, E. S. 1993. Getting Back into Place: Toward a Renewed Understanding of the
Place-​World. Bloomington: Indiana University Press.
Casey, E. S. 1997. The Fate of Place: A Philosophical History. Berkeley/​Los
Angeles: University of California Press.
Cresswell, T. 1996. In Place/​Out of Place: Geography, Ideology, and Transgression.
Minneapolis: University of Minnesota Press.
Davis, M. 1990. City of Quartz: Excavating the Future in Los Angeles. London/​
New York: Verso.
Dittmer J. 2007. “America Is Safe While Its Boys and Girls Believe in Its
Creeds!”: Captain America and American Identity prior to World War 2.”
Environment and Planning D: Society and Space 25(3): 401–​423.
Foucault, M. 1978. The History of Sexuality. Vol. I: An Introduction. Translated by
R. Hurley. New York: Pantheon.
Hägerstrand, T. 1970. “What About People in Regional Science?” Papers of the Regional
Science Association 24: 7–​21.
Husserl, E. 1982. Ideas Pertaining to a Pure Phenomenology and to a Phenomenological
Philosophy: First Book. Translated by F. Kersten. Dordrecht, the
Netherlands: Kluwer.
Irigaray, L. 1993. “Place, Interval: A Reading of Aristotle.” In An Ethics of Sexual
Difference, translated by C. Burke and C. G. Gill. Ithaca, NY: Cornell
University Press.
Manzo, L. C., and P. Devine-​Wright, eds. 2013. Place Attachment: Advances in Theory,
Methods and Applications. London/​New York: Routledge.
Massey, D. 1993. “Power-​Geometry and a Progressive Sense of Place.” In Mapping the
Futures: Local Cultures, Global Change, edited by J. Bird, B. Curtis, T. Putnam,
G. Robertson, and L. Tickner, 59–​69. London: Routledge.
Morley, D. 2000. Home Territories: Media, Mobility and Identity. London/​
New York: Routledge.
Newman, D., and A. Paasi. 1998. “Fences and Neighbours in the Postmodern
World: Boundary Narratives in Political Geography.” Progress in Human
Geography 22: 186–​207.
Sack, R. D. 1997. Homo Geographicus: A Framework for Action, Awareness, and Moral
Concern. Baltimore/​London: Johns Hopkins University Press.
Smith, S. 2012. “Intimate Geopolitics: Religion, Marriage, and Reproductive
Bodies in Leh, Ladakh.” Annals of the Association of American Geographers
102(6): 1511–​1528.
Soja, E. W. 1980. “The Socio-​spatial Dialectic.” Annals of the Association of American
Geographers 70(2): 207–​225.

P l a c e a n d E x t e n ded   A g e n c y   [ 219 ]
220

Tuan, Y.-​F. 1982. Segmented Worlds and Self: Group Life and Individual Consciousness.
Minneapolis: University of Minnesota Press.
Tuan, Y.-​F. 1977. Space and Place: The Perspective of Experience. Minneapolis: University
of Minnesota Press.
Tuan, Y.-​F. 1974. Topophilia: A Study of Environmental Perception, Attitudes, and Values.
Englewood Cliffs, NJ: Prentice-​Hall.
Woolf, V. (1929) 2002. A Room of One’s Own. New York: RosettaBooks.

[ 220 ]  The Place of Agency

╇ 221

CHAPTER 23

How Agency Is Distributed

Through Installations
SAADI L AHLOU

ACTION IS DISTRIBUTED

Artifacts will help, support, or guide users in their activity through their
culturally constructed properties. This idea emerged in various forms in
history (Leroi-╉Gourhan 1965); philosophy (Simondon 1969); psychology
(Gibson 1950, 1982, 1986; Uexküll 1925; Vygotsky 1978), design (Norman
1991), cognitive science (Pea 1994), and so on. Hutchins (1995), observ-
ing jet pilots, demonstrated that what flies the plane is not pilots alone
but the whole cockpit, including instruments, checklists, maps, and so on.
Cognition is “distributed” in its content as well as in its process. Objects
that possess agency are called “actants” (Akrich, Callon, and Latour 2006),
a term including acting humans and material objects alike. When describ-
ing activity emerging from subject-╉ object interaction, some (Gibson,
Norman, Latour) focus more on properties of the object, while others
(Vygotsky, Uexküll) start from the subject. This chapter proposes a larger
systemic view.
Over the last two decades, I  analyzed human activity recorded from
a first-╉
person perspective, with “subcams”—╉ miniature video cameras
attached to glasses of volunteers performing their usual activity, at work or
elsewhere. With participants we analyzed their recordings to make explicit
their cognitive processes and the various elements involved in their action
(Lahlou 2011, 2014).
222

In doing so, it appears obvious that action is distributed and its deter-
minants as well. For example, in the activity of preparing breakfast, we
can observe that, as Mother sets the table, Father goes to buy croissants,
Coffee-╉
Machine brews Coffee, Bottle contains Orange-╉ Juice, Alarm-╉
Clock wakes up Daughter, while Thermostat maintains House warm, and
so on.
To clarify the distribution of agency, let us look in detail at the determi-
nants of action.

INSTALLATION THEORY

When you manage to cycle in heavy traffic, this is the result of simultane-
ously using the affordances of the road; of mobilizing embodied skills; of
protection by the traffic rules that prevent cars from driving you off the
road. Society has constructed the built environment (the road), trained you
to embody skills (cycling, traffic sign reading), and runs control institu-
tions (police, traffic lights, rules of the road). On one hand these three lay-
ers guide and scaffold your individual behavior, enabling you to safely reach
your destination. On the other hand they make you a predictable road user
to others, so we all together co-╉construct a “normal” traffic flow at societal
level. The same mechanism that nudges and empowers you is also a mecha-
nism of control at aggregate level.
The urban street is an “installation”; this installation is not located
within the physical world only or inside your nervous system alone; it is
distributed in the built environment, in educated and disciplined bodies,
in institutions and their enforcing agents. Our everyday environment is
built by humans; centuries of gradual construction have produced “instal-
lations” such as kitchens, conference rooms, or cockpits that make activity
simple, so simple in fact that even novices can perform satisfyingly. Barker
described of such setups as “behavioral settings”: “stable, extra-╉individ-
ual units with great coercive power over the behavior that occurs within
them” (Barker 1968:17). But we must also include the individuals in the
installation.
Installation theory considers how societies guide individual activity by
constructing “installations,” in which individual subjects operate. Subjects
are moved by individual motives and intentions, but an installation has
a momentum of its own. An installation is “a socially constructed system
locally guiding a specific activity, by suggesting, scaffolding and constrain-
ing what society members can/╉should do in this specific situation” (Lahlou

[â•›222â•›]â•… The Place of Agency

  223

2015). Installation theory is a theory for nudging, but also for more intru-
sive control.
As for most important phenomena, a plethora of concepts and theo-
ries have been proposed in the literature, for example, “affordances,”
“disciplining institutions,” “structures,” “assemblages,” “environments,”
“niches,” “infrastructure,” “channels,” “codes,” “norms,” “habitus,” and so
forth. Installation theory tiles three layers of determinants that were so far
separated by disciplinary approach, even though in practice they operate
as a bundle. The various components of an installation are not located in
the same medium; they emerge visibly as a functional unit only when they
locally assemble in action; as a bundle they were gradually constructed in
practice, and only as a bundle can they be understood. The example of the
road above illustrates this: the installation is not limited to the visible set-
ting of the pavement; it includes the representations and driving skills of
the participants, as well as the institutions that regulate the traffic. Each is
meaningless without the others.
Within an installation, behavior is funneled simultaneously by (a) affor-
dances of the physical setting, (b)  subjects’ embodied competences, and
(c)  social control. What characterizes human installations is their inten-
tionality. Their design fits a specific purpose: they support a project of activ-
ity. Humans make installations for all kinds of activities:  roads, houses,
armchairs, hospitals, cradles, graves, and so on.
Individuals do have free will (e.g., they chose to get into the traffic),
but their agency is limited. Clearly, installations have a momentum of
their own. Installation theory considers that humans usually act in such
settings (roads, homes, restaurants, hospitals, etc.) where activity is
channeled by the cultural installation. Depending on how we want to
see it, we can consider that individuals use installations as instruments
to reach their goals; or that installations use individuals to produce an
outcome. For example, the traffic installation (roads, signs, lights, traffic
rules, police, etc.) uses drivers to produce smooth transport at societal
level; the roller coaster or the night club uses Albert as “a client” and as
“a participant” to produce “experience” for users, and profit. Both per-
spectives are valid to a certain extent.
In other words, while Hutchins noted the environment contains medi-
ating structures that process parts of the cognitive operations necessary to
complete tasks, Installation theory considers more radically that our envi-
ronment contains constructed “installations” which coproduce and funnel
activity with the human actors as components. It must be clear here that
installations are not just the context: individuals are part and parcel of

H o w A g e n c y I s Di s t r i b u t ed T h r o u g h I n s ta l l at i o n s   [ 223 ]
224

them. Society does not only install physical artifacts, it also installs inter-
pretive structures (the manual) in individuals so the physical layer is used
“as it should.” For smooth traffic, drivers should know how to drive their
vehicle and there should be rules that are enforced. And this is indeed
the case: society has set up systems to embody competences in individu-
als (e.g., driving lessons, license) and sets of rules of the road, with an
enforcing body (the Police). These are neither independent nor accidental:
the three sets of determinants are obviously part of one single societal
project: smooth traffic. In practice, what the subject can do is the remain-
ing intersection of what is objectively possible (affordances), subjectively
possible (embodied competences), and socially enabled (social control).
Amazingly, while this is common sense, the three layers tend to be studied
independently by different disciplines. In the study of accidents, though,
it becomes clear that problems occur only when the loopholes of the vari-
ous layers of defense of the system combine (Reason 2000).
Installation theory has two aspects. The first is an operational frame-
work to describe the determinants of activity. This framework, above, is
simplistic, with three layers of components; it can be used for analysis and
for change management. It is practical because the layers indicate where
one can act to change the system.
The theory’s second aspect addresses the evolution of installations.
It describes the interactions between the components of the model, and
genetic, functional, and historical mechanisms that produce sustainable
systems. During education individuals will internalize not only various rep-
resentations and practices, but also social rules. Foucault (1975) described
how this embodiment of social rules makes individuals “docile bodies” and
as a result each actor becomes his own controller. Other society members
will also encourage and control others into behaving (from smiling or frown-
ing to giving directions or orders); societies also have specific control forces
(e.g., Police) to ensure enforcement. The rules themselves may be reified
into artefacts which can become actants. Embodied interpretation systems
(e.g., mental representations of objects) and physical objects themselves are
taken in a chicken-​and-​egg reproductive cycle, with dual selection operat-
ing in each of the two forms (embodied and material). Institutions are in
charge of designing, operating, maintaining, and evolving the installations;
they monitor the dual selection process that make the installation a single,
coherent, functional unit. In this short chapter we will use only the first part
to provide a framework distinguishing three types of agency.
Installation theory considers three types of components that determine
action: affordances in the physical context, embodied competences of the sub-
ject, and social influence—​direct or through institutions. Affordances afford;

[ 224 ]  The Place of Agency

  225

competences enable, and rules empower and control. The combination

of these three layers generates a limited tree of possibilities for action at
every step; so actors are funneled into the very activity the installation is
designed for.
At a physical level, objects have affordances that both limit some behav-
iors and call for some others. Gibson considers that objects immediately
display their affordances to the senses. For example, a door handle signals
how it can be handled and turned (Norman 1988). Uexküll placed more
emphasis on the interpretation by the subject, considering the object pro-
vides “connotations for activity” (Uexküll 1925). Unless one is in the wil-
derness, we are surrounded by man-​made objects, which, as stated above,
carry some intentionality and affordances by human design. The environ-
ment has been planted with physical installations.
The physical level can be used by Humans because they have embodied
competences to interpret these affordances and perform action. In practice
we learn and embody new affordances. Someone who never used a fancy
door handle (e.g., electric button) could stay trapped in the room even if
the door is not locked. Representations of common objects, material or not
(doors, hospitals, democracy, etc.), are shared by the population: they are
“social representations” (Moscovici 1961; Abric 1994). Social representa-
tions are the World’s user’s manual. As we see with the example of the door,
“interpreting” is not only a matter of chaining ideas to perception, but also
chaining action to ideas. Interpreting a door includes the motor action to
open it—​or close it. Just as the physical environment is planted with physi-
cal installations, the minds and bodies of Humans are an embodied layer
that has been installed with representations and motor skills; as a com-
bined result of biology, education, and experience.
Finally, we take into account what others do and want us to do (or not
do). Others can influence us by force or menace; most often on our own
behalf we are influenced by imitating others or anticipating consequences
of our actions. Social psychology experiments show interpersonal or group
influence in laboratory (for extreme cases see Haney, Banks, and Zimbardo
1973, or Milgram 1963); in real life influence is often mediated by institu-
tions, and the expected behaviors come in the form of rules. There lay the
third, social layer of installations: its components are distributed over “other
people.” This level can hamper and control as well as empower and support.
It is the most diverse in form since influence can come directly from indi-
viduals or from groups, and can be mediated by institutions and rules.
Let’s take again breakfast in the family: Father brings four Croissants;
Croissants have affordance as “food.” Children will interpret these as edi-
ble; they know how to eat them. The institutional rules of the family are

H o w A g e n c y I s Di s t r i b u t ed T h r o u g h I n s ta l l at i o n s   [ 225 ]
226

that Croissants should be shared equally, one per person, so if Son wants
to take an extra one, other participants will prevent him from doing so.
Another example in a hospital installation: a cancer patient may afford sev-
eral treatments among which the competent doctor will make a choice; but
hospital rules will force the doctor to ask the patient’s consent.
In short, Installation theory states action is supported and guided at three levels:
affordances planted in the environment, competences embodied in the subject, other
people and institutions. It suggests that sustainable behaviors are the ones that are
simultaneously supported by and compatible with these three layers altogether.
Agency, as the potential for action, is therefore distributed in these three layers.

HOW AGENCY IS DISTRIBUTED

Activity is a process that gradually transforms one state of things into

another. We can consider that agency, for a subject who has a goal (“a rep-
resentation of a specific desired final state”), is the capacity to reach the
goal in the conditions given.
The final state is the result of a series of actions performed in a distrib-
uted way, by the subject and a series of actants: other people and objects.
From the perspective of the subject, these actants can be resources or
constraints, depending whether they support or hamper the process of
reaching the goal. Agency of a subject, as the capacity to reach a goal, is a
function of access to resources and liability to constraints.
We highlighted three layers of determinants: physical affordances,
embodied competences, and social control. In that framework, what does
agency mean? Capacity to reach the goal in the conditions given translates
into capacity to use the necessary components in the three layers above.
Moreover, as we saw that many contexts were constructed as “installa-
tions,” having access to an installation that is designed to reach the goal
provides agency for that goal. This is a first finding: capacity to use (i.e., enti-
tlement and competence) an installation provides agency for the goals supported
by this installation. But not everything is an installation; many goals will
not match exactly with the available installations; some components will
be general enough to be used across the board; and so on.
Let us then look more generally at agency as the capacity to mobilize the
necessary components of activity.
Material agency is access to the resources in the physical layer. It includes
direct and indirect access and control over affordances. For example, I can
use my kitchen, my car, and the roads in my country.
Access to resources is not enough. The subject must be able to use them
in a relevant way. The subject’s embodied agency is the set of embodied

[â•›226â•›]â•… The Place of Agency

  227

education, experience, and more generally psycho-​physical capacity to

interpret. It includes mental representations, motor skills, memories of
past experience that can serve to interpret situations, and so on. Note that
these competences can be used for the subject’s goals, but also to contrib-
ute to other people’s goals, as shown above.
Activity is constrained by other people. They define and enforce, usu-
ally collectively, rules in their local domains of control. When operating in
these domains, subjects must obey local rules, or face the consequences.
Social agency is the capacity to deal with social forces. This capacity is often
the result of the subject’s history in that specific community, especially
his or her status and roles1. From a subject’s perspective, social identity
(belonging to various communities) provides access rights to some com-
mons (protection, support, common resources). It also makes one subject
to those rules and accountable to the communities, for good or bad. Being
an inhabitant of a village may grant access to common meadows for one’s
cattle, local clubs, hospital, and so on. Being a national of a country pro-
vides access to public services, infrastructure, and so on. Family, town,
country, organization, religious community, party, professional body,
and so on enable their members to access resources and provide support
through rules of use. They also forbid certain behaviors.
Finally, in some cases restrictive laws do not apply to trusted members
of the community:  top managers have no fixed and controlled working
hours, some government heads are immune to laws during their mandate,
and so on.
We can call institutional agency the agency granted to the individual by a
community that provides support for action. Institutional agency is about
what behavior is authorized or forbidden, and this agency is mediated by an
institution at the moment of acting (an institution is a stable common set
of rules in a community). The various layers of agency are not independent.
While this triple layered analysis may appear commonsensical or too
encompassing in theory, it is handy in practice. Key here is that an installa-
tion work as a bundle: sustainable performance requires agency in the three
realms simultaneously. For example, having a car and a road is not enough
to drive from point A to point B: one must also know how to drive, and
drive on the right side of the road; otherwise one may face serious prob-
lems. By separating the various levels of agency, the installation framework
facilitates the analysis of cases where agency is contested, hazardous, and
so on, and also provides handles to act—​to empower or to restrict.
One can be surprised here not to find “power,” as the capacity to force
other people to perform behaviors, in our analysis of agency. In fact power
can appear as brute force in embodied agency, as economic power in mate-
rial agency, as institutional power in entitlement. Cognitive authority is

H o w A g e n c y I s Di s t r i b u t ed T h r o u g h I n s ta l l at i o n s   [ 227 ]
228

one variant of power active in social influence. The classic approach of

financial capital, cultural capital, and social capital (Bourdieu and Passeron
1990) seems to map the three levels of agency outlined by Installation the-
ory: access to material goods, embodied skills, and social support.
Finally, we should note that agency is a situated notion: it is agency to
do something specific; it is goal-╉dependent and context-╉dependent. Agency
to make a pair of shoes is different from the agency to vote a law or to feed
a village.
In sum, we saw three main forms of agency: material agency (access to
affordances—╉direct or through the proxy of social entitlement); embod-
ied agency (interpretative competences, where interpretation involves
motor skills as well as cognitive); and social agency (often mediated by
an institution, including capacity to evade local rules). Robust agency
must combine the three layers. In practice we can consider the palette
of these three layers to improve a given situation. By distributing agency
over the layers, we can design interaction and foster or hamper different
types of behavior. In doing so we can empower (or control) groups or
individuals and support political or organisational change. Because such
approach can be very efficient, and will partly rely on nonhuman actants
that have no moral concerns, we must be especially cautious and ethical
in the design process.

ACKNOWLEDGMENTS

This chapter benefited from a residency at Paris Institute for Advanced Study
(France) as a EURIAS senior fellow, with support of the European Union
7th Framework Programme for research (grant agreement no.  246561),
and from constructive reviews from Paul Kockelman and Nick Enfield.

NOTE

1. Stoetzel (1963) defines role as the set of behaviors that can be legitimately
expected from the subject, and status as the set of behaviors the subject can
legitimately expect from others.

REFERENCES

Abric, J.-C. (1994). Pratiques sociales et représentations. Paris: Presses Universitaires de

France.

[â•›228â•›]â•… The Place of Agency

  229

Akrich, M., M. Callon, and B. Latour. 2006. Sociologie de la traduction: textes fondateurs.

« Sciences sociales ». Paris: Mines Paris, les Presses.
Barker, R. G. (1968). Ecological Psychology: Concepts and Methods for Studying the
Environment of Human Behavior. Stanford University Press.
Boltanski, L., and L. Thévenot. 2006. On Justification: Economies of Worth. Princeton,
NJ: Princeton University Press.
Bourdieu, P., and J.-​C. Passeron. 1990. Reproduction in Education, Society and Culture.
Volume 4: Reproduction. London: Sage.
Foucault, M. 1975. Surveiller et punir : naissance de la prison. Bibliothèque des histoires.
Vol. 2004. Paris: Gallimard.
Gibson, J. J. 1950. The Perception of the Visual World. Boston: Houghton Mifflin.
Gibson, J. J. 1982. “Notes on Affordances.” In Reasons for Realism: Selected Essays of
James J. Gibson, edited by E. Reed and R. Jones, 401–​418. London: Lawrence
Erlbaum Associates.
Gibson, J. J. 1986. The Ecological Approach to Visual Perception. London: Lawrence
Erlbaum Associates.
Haney, C., Banks, C., and Zimbardo, P. (1973). Interpersonal Dynamics in a Simulated
Prison. International Journal of Criminology and Penology, 1, 69–97.
Hutchins, E. L. 1995. “How a Cockpit Remembers Its Speed.” Cognitive Science
19: 265–​288.
Lahlou, S. 2011. “How Can We Capture the Subject’s Perspective? An Evidence-​Based
Approach for the Social Scientist.” Social Science Information 50(3–​4): 607–​655.
Lahlou, S. 2014. “Social Representations and Social Construction: The Evolutionary
Perspective of Installation Theory.” In Handbook of Social Representa,
edited by G. Sammut, E. Andreouli, G. Gaskell, and J. Valsiner, 193–​209.
Cambridge: Cambridge University Press.
Leroi-​Gourhan, E. 1965. Le geste et la parole. Paris: Albin Michel.
Milgram, S. (1963). Behavioral Study of obedience. Journal of Abnormal Psychology,
67(4, Oct 1963), 371–378.
Moscovici, S. (1961). La psychanalyse, son image et son public. Etude sur la représentation
sociale de la psychanalyse. Presses Universitaires de France.
Norman, D. A. (1988). The Design of Everyday Things. New York: Basic Books.
Norman, D. A. 1991. “Cognitive Artifacts.” In Designing Interaction: Psychology at the
Human Computer Interface, edited by J. Carroll, 17–​38. New York: Cambridge
University Press.
Pea, R. D. 1994. “Seeing What We Build Together: Distributed Multimedia Learning
Environments for Transformative Communications.” Journal of the Learning
Sciences 3(3): 285–​299.
Reason, J. T. (2000). Human error: models and management. British Medical Journal
(Clinical Research Ed.), 320(7237), 768–770.
Simondon, G. 1969. Du mode d’existence des objets techniques. Paris: Aubier.
Stoetzel, J. (1963). La psychologie sociale. Paris: Flammarion.
Uexküll, J. von. (1925) 1965. “Mondes animaux et monde humain.” In Mondes
animaux et monde humain, suivi de Théorie de la signification, edited by J. von
Uexküll, pp. 1–​90. Paris: Denoël, 1965.
Vygotsky, L. S. 1978. Mind in Society: The Development of Higher Psychological Processes.
Edited by M. Cole, V. John-​Steiner, S. Scribner, and E. Souberman. Cambridge,
MA: Harvard University Press.

H o w A g e n c y I s Di s t r i b u t ed T h r o u g h I n s ta l l at i o n s   [ 229 ]
  231

PART TEN

From Cooperation to Deception

and Disruption
╇ 233

CHAPTER 24

Cooperation and Social Obligations

DAVID P. SCHWEIKARD

P henomena of distributed agency are deeply entrenched in our lives as

agents. We go for walks and visit concerts together, we found clubs and
organizations, build research teams and political parties. Curiously, schol-
ars of the philosophy of mind and action have only relatively recently, that
is, in the mid-╉1980s, started devoting their attention to these phenomena.
As more elaborate accounts are being put forth and discussed, it becomes
clear that core issues revolve around two main foci: the intentionality and
the normativity of distributed agency. Under these headings, scholars seek
to understand, on the one hand, the structure of the intentional attitudes
that guide joint action and by which joint action is explained. And on the
other hand, they inquire into the questions as to whether agents incur, by
cooperating with others, specific mutual obligations and, if they do, how
these obligations come about, what kind they are, and how they are justi-
fied. In what follows, I shall provide a brief account of these two sets of
challenges and expound ways of meeting them. My focus will be on small-╉
scale cooperative joint action, which I describe more thoroughly in the next
section, and the question as to whether and, if so, how specifically social
obligations between the agents involved come about.

SORTING TYPES OF AGENCY

One way of sorting phenomena of distributed agency proceeds by ask-

ing how many agents can or need to be involved in the performance of an
234

action. I shall lay out such an approach by appealing only to actions human
agents normally can and do perform (together). This will serve to narrow
the scope of the following reflections without denying the significance of
classifying and understanding the cooperative behaviors of animals or the
actions of and interactions with technical systems.
That said, it seems helpful to distinguish between actions human agents
can perform on their own and actions they can only perform together. The
action of going for a walk, for instance, can be performed by a single agent
and by two or more agents; by contrast, actions such as playing tennis or
lifting a heavy object can (under normal conditions) be performed only by
two or more agents. If we call actions performed by two or more agents
joint actions, we can capture this difference between types of action by say-
ing that contingently joint actions are those that could also be performed by a
single agent, and that necessarily joint actions are those that require the par-
ticipation or contribution of two or more agents. Joint actions come in a
vast variety, ranging from spontaneous and one-╉time cooperation to regular
and highly structured collective agency. I shall return to core characteristics
of joint actions in due course, although not without mentioning that it has
proved helpful to treat questions of group agency separately from analyses of
joint action (cf. Pettit and Schweikard 2006). Even though it is quite common
to refer to actions performed by two or more agents as actions of a group, the
design and structure of some groups raises the question whether they qualify
as agents in their own right. Corporations, for instance, are complex social
entities whose internal makeup and structures of decision making justify
treating their ways of operating as decisively different from that of random
collectives or pairs of cooperating agents. Interest in the structure of groups
that may count as agents in their own right may be prompted primarily by
issues of collective responsibility, but there are certainly also fundamental
questions in action theory and social ontology that deserve attention.
In what follows, I shall set aside questions raised by complex social for-
mations and concentrate instead on small-╉scale cooperative joint action.
Prototypical cases such as two agents going for a walk together or lifting a
heavy object together can be used to illustrate two central challenges in the
philosophy of joint action, and in the theory of distributed agency more
generally.

THE INTENTIONALITY OF JOINT ACTION

A simplified and preliminary way of formulating these two challenges

posed by small-╉scale joint action on which I concentrate here is given by

[â•›234â•›]â•… From Cooperation to Deception and Disruption

  235

asking what is special about joint action and what joint action implies nor-
matively. In a nutshell, the question “What is special?” refers to those ele-
ments of joint actions that set them off against parallel individual actions,
and the question concerning normative implications is aimed at the mutual
obligations and entitlements that may (or may not) hold between cooper-
ating agents. The issues thus brought to the fore are the intentionality and
the normativity of joint action.
Starting with intentionality, let us consider the following case. Imagine
three agents (A, B, and C) walking next to each other in the same direction
and that from what is observable it is unclear whether they or a subset of
them are walking together or not; imagine further that we know that two of
them (A and B) are going for a walk together, while the third (C) is just acci-
dentally walking in parallel to them. With respect to this imagined case, we
may ask what makes for the difference between A’s and B’s walking together
(or jointly) and B’s and C’s walking in parallel, where A and B perform a joint
action and B and C (in abstraction from A’s presence) perform parallel, but
otherwise unrelated, individual actions. A  minimal consensus about this
issue among philosophers of action is best captured in John Searle’s sloga-
nesque formulation “If there is anything special about collective behaviour,
it must lie in some special feature of the mental component, in the form
of the intentionality” (Searle 1990:402). But how should this specific form
of intentionality be described? Which specific intentional attitudes guide
and serve to explain joint actions in contrast to coincidentally parallel indi-
vidual actions?
Once it is accepted that the intentional attitudes that are in play in joint
action cannot be simply the same as those that guide solitary individual
action, a familiar move in the debate is to guard against postulating a form
of group mind that hovers above, as it were, and controls the individuals
involved. It seems just obvious that explaining joint action does not call for
such a postulate of a supra-​individual entity. What is required, instead, is
an account of how exactly the intentionality characteristic of joint action
differs from ordinary individual intentionality; put differently, what is
sought is an answer to the question “What is collective about collective
intentionality?” (cf. Schweikard and Schmid 2013). In short, the specific
jointness or collectivity of the intentionality in play can rely in its content,
its mode, or its subject. I take this last “or” to be inclusive and shall go on
to sketch a relational account of joint intentions that takes all three aspects
to take a specific collective form.
In order to intentionally guide their joint walk, A and B in our previous
example both have to intend their joint walk, in other words, their joint
walk has to be represented in the content of their intentions. On a first

C o op e r at i o n a n d S o c i a l O b l i g at i o n s   [ 235 ]
236

approximation, they each need to form intentions of the form “I intend

that we go for a walk” (cf. Bratman 1999: ch. 5; Bratman 2014), since if each
intended only to go for walk, this could already be fulfilled if each walked
completely independently of the other. Capturing what makes their walk-
ing a case of intentionally walking together in this form, however, attracts
a number of problems. Thus, one may hold that intentions of the form “I
intend that we … ” contradict the otherwise plausible assumptions that
one can intend only one’s own actions, those one can control or whose per-
formance is up to oneself (cf. the discussion in Bratman 1999: ch. 8). These
worries can be deflected by arguing that the intentions in question are dif-
ferent from ordinary intentions to do something, in that their formation
and enactment are dependent on all individuals involved forming structur-
ally analogous intentions. A may not be in a position to fully intend, con-
trol, or settle deliberation concerning a joint walk with B, but if the walk is
to be a joint walk, then it seems plausible to require that each refer to it in
the content of their intentions. Both these intentions are, in other words,
adequately formed only in relation to, and in combination with, the respec-
tive other. But accepting this “collectivity condition,” as Raimo Tuomela
(2005) has termed it, calls into question whether the subject place of the
verbalization of the intentions can plausibly be rendered in the first person
singular. Intentions of the form “I intend that we … ” then appear to be
circular in that their formation presupposes the “we” they are meant to
constitute (cf. Bratman 2014: ch. 2).
One way of avoiding this problem lies in postulating that the relevant
intentions not only have a collective content, but that they also stand in
a specifically collective mode, the we-​mode. Tuomela (2005), for instance,
takes intending in the we-​mode to capture individuals’ “intending as a
group-​member” as opposed to their “intending as a private person.” On
his account, individual agents who shift to the we-​mode thereby take the
respective group’s perspective and form corresponding contributory inten-
tions. A theoretically less demanding approach restricts the adjustment
called for earlier to the modification that the individual agents involved
intend in the form “We intend that we… . ” This does not appeal to a dif-
ferent form of intending—​we-​intending (Tuomela) or primitive we-​inten-
tions (Searle 1990) but only to a pluralized verbalization of the intentions
in question. By thus intending in the plural, cooperators such as A and B
refer to themselves and each other using the plural indexical “we,” which in
the context at hand is correctly used only when it picks out a plurality each
of them represents. Their intentionally guiding the joint walk thus requires
that A and B each form and enact intentions of the form “We intend that
we go for a walk.”

[ 236 ]  From Cooperation to Deception and Disruption

  237

But whose intention is the intention that guides the joint walk? Who is
the subject of this intention? Having set aside the postulate of a mysterious
supra-​mind that could be the subject of this intention, we should be drawn
to the conclusion that the intending individuals together form the subject
of this intention. With respect to both the content and the mode of the
intentions of the individual agents involved, we noted that (1) collectivity
has to feature in both of these aspects, and (2) it has to feature by high-
lighting the relations between the agents involved. In a two-​person case,
only both of them can give sense to their each intending the joint activity
and to their using the first person plural indexical “we” in verbalizing these
intentions. And if it takes two to set up these interrelated intentions, then
the same two, together, should be viewed as the subject of the intention
that guides the joint walk.
This result calls for (at least) one immediate clarification regarding the
number of intentions involved. I have referred to the individual agents’
intentions and to the intention that guides the joint walk, possibly as
if those were simply on a par. If we view as intentions only such mental
phenomena that are had and embodied by individual agents, then only
the participatory intentions of the individual agents qualify as inten-
tions. This individualistic picture runs the risk of losing sight of the ana-
lytic and explanatory task set out at the beginning of this section. On the
other hand, if we allow all structures of attitudes that guide and explain
actions to be intentions, then specifically related intentions, such as
those of cooperating agents, can play that role together. In line with the
latter understanding, we can call the structure of interrelated attitudes
that guides A’s and B’s joint walk the joint intention whose subject A and
B are together.
Strictly speaking, the analysis of joint intentions should not be lim-
ited to the actual intentions regarding the joint activity, as these will typi-
cally be accompanied by further intentions and beliefs on the part of the
agents involved. In these respects it seems plausible to demand that the
agents’ further intentions, for example, other goals A may have in mind
in relation to the walk with B, do not conflict with their joint endeavor;
that they each exhibit a sufficient degree of readiness to follow through
with their endeavor; and that they each are aware that what and how they
intend together is out in the open between them. Requiring an elaborate
form of common knowledge among the participants may seem inadequate
or overly intellectualized in some contexts, but a weaker “no conflict”–​
condition regarding their awareness and understanding of the situation
may nevertheless be necessary for the attitudinal nexus to adequately pro-
vide guidance.

C o op e r at i o n a n d S o c i a l O b l i g at i o n s   [ 237 ]
238

Returning to the laxly formulated question as to “what is special?” about

joint action, the answer suggested by the relational account just sketched
is that joint action is defined as being guided by a specific complex of inter-
related attitudes of the individual agents involved. This complex of atti-
tudes, the action-╉guiding joint intention, is in the main constituted by the
participants’ intentions of the form “We intend that we … ,” and its sub-
ject are the participants together. Understanding the one guiding attitude
(the joint intention) and its subject as being thus constituted by specifi-
cally interrelated individual attitudes and by interrelated individual agents,
yields a particular perspective on what it means to say that agency is dis-
tributed. The claim that interrelations between individual attitudes are
characteristic of cooperation and distributed agency leads to the question
as to whether these phenomena also characteristically involve normative
relations between the agents involved. This is the question we turn to in
the next section.

THE NORMATIVITY OF JOINT ACTION

When it comes to the normativity of joint action, the interesting question

is not whether any mutual obligations exist between the agents involved,
but whether it is characteristic of joint action that the agents have or incur
obligations to each other precisely in virtue of acting together. According
to Margaret Gilbert’s account, such specific mutual obligations are funda-
mental to phenomena of joint action (cf. Gilbert 2014). Gilbert claims that
such obligations are implied by the fact that in getting together to perform
a joint action, agents enter (what she calls) a “joint commitment,” which
she characterizes as an irreducible, shared practical attitude the agents cre-
ate by agreeing on the pursuit of a joint endeavor, which is described as an
attitude of the group that is possibly discontinuous with the intentions of
the individual agents, and which can be annulled only unanimously. If two
agents are thus jointly committed to act together, each is obligated to the
other to do his or her bit and each is entitled to the other’s contribution
or to rebuking the other in case of noncompliance. Gilbert’s view is that
joint commitments form the basis of jointly intentional action and that
the inherent obligations and entitlements, which are not understood as
moral but as social normative relations, are fundamental to all kinds of
joint action.
The problem with this account is certainly not that it is never true that
agents cooperate on the basis of binding mutual agreements, or that it is
inconceivable that cooperators sometimes treat each other as if a contract

[â•›238â•›]â•… From Cooperation to Deception and Disruption

  239

existed between them even though their agreement was nonverbal. But
some authors have remained unconvinced by some of the aforementioned
details of the account and by the view that normative structure that comes
with joint commitments is indeed fundamental to all kinds of joint action.
For instance, below the level of highly organized group agency, and unless
a somehow emergent subject of the relevant intentional states is appealed
to, it does not seem plausible to claim that the intention that explains a
joint action can diverge from the participatory intentions of the individual
agents involved. And the feature of joint intentions Gilbert terms the “con-
currence criterion” (2014:106–​108), which is meant to capture that joint
intentions can only be modified or annulled unanimously, also seems too
demanding to be a general feature of joint action. For even if a degree of
unanimity, a shared belief as to what they are doing together, is required to
keep the cooperators on track, as it were, it would seem exaggerated to hold
that in all cases of joint action one participant is entitled to hold all others
to the original joint commitment. Gilbert’s view thus represents phenom-
ena of joint action as in principle akin to scenarios in which agents have
entered a contract or exchanged promises, with all the normative impli-
cations involved by that. Without disputing that some cases exhibit that
sort of normative structure, it seems doubtful that all cases are like that
and thus that joint commitments are essential to joint action (cf. Bratman
2014: ch. 5).
Do we have to conclude that joint action generally does not involve
mutual obligations? Or does joint action imply a weaker sort of norma-
tive relationship? Against proponents of the view that only certain kinds
of joint action involve mutual obligations, such as Michael Bratman (2014),
others have suggested to regard practical attitudes such as reliance (Alonso
2009) or trust (Schmid 2013) as grounding normative interpersonal rela-
tionships and as being fundamental to joint action. The normative bonds
grounded in these attitudes can be described as weaker than the relation-
ship Gilbert conceives of in terms of joint commitments. The unifying idea
behind accounts that refer to reliance or trust is that if one agent forms (and
exhibits or expresses) this kind of attitude regarding one’s partner’s contri-
bution to a joint action, the other is led to form and endorse the same kind
of attitude. To illustrate this, assume that A and B are in a position to move
the piano across the room, where this is something neither can do alone,
but perfectly doable in unison. Assume further that each of them not only
intends to move the piano, as in the analysis outlined in the previous sec-
tion, but that in addition each of them relies on the other doing his or her
bit, in other words, push at his or her end of the piano (cf. Alonso 2009). Or
assume that each of them, besides intending to move the piano, trusts that

C o op e r at i o n a n d S o c i a l O b l i g at i o n s   [ 239 ]
240

the other will in fact push at his or her end (cf. Schmid 2013). The relevant
contention put forth with these accounts is that, characteristically, cooper-
ating agents create mutual obligations—​or, in other words, give each other
normative reasons to contribute—​by forming attitudes of reliance or trust
regarding the other’s contribution. Thus, in our example, A’s relying on (or
trusting) B to push his or her end of the piano creates a reason for B to do
so, and vice versa; let us assume that this reason is comparatively strong,
so that it corresponds to an obligation for B to contribute. Obviously, not
all instances of reliance or trust can generate reasons and obligations in
this way, but under normal conditions, and especially in cases of mutual
reliance and mutual trust, in which the attitudes can be understood as
mutually reinforcing, the generative effect seems to occur. And it should be
emphasized that the obligations in question are not moral, but only social
obligations, which means that their force is explained by reference to the
particular social situation, including the attitudes and relations that fea-
ture in it, and not by reference to more general moral principles or norms.
However, the question remains whether it is necessary to posit the pres-
ence of a different kind of attitude as fundamental in order to explicate a
basic normative relationship implied by joint action. Defenders of accounts
that refer to reliance or trust seem to answer in the affirmative, but an
account that builds on the notion of an individual social commitment can
confine the conceptual resources to fulfill this task to attitudes more closely
related to the intentions and beliefs that accompany joint action. Abe Roth
(2004) has developed a variant of such an account, but I shall conclude by
outlining the main elements of a different variant. These regard the con-
ceptual resources made use of in this account, the way it explains the nor-
mative relationship involved in joint action, and the connection it draws
between commitments and obligations.
First off, talking of commitment in relation to intentions does not intro-
duce an additional type of attitude; it rather highlights another element
of the structure of intentional activity. In the case of joint intention, this
amounts to making explicit that the socially oriented intention of the form
“We intend that we … ” implies a specific type of commitment, a social
commitment. Whenever an agent intends to do something, they commit,
inter alia, to choosing means appropriate to their end and to avoiding any-
thing that could render their attaining that goal impossible. Building on
this, we can say that when an agent intends a joint action, as in the cases
considered above, they assume a social commitment that regards both
(1)  their own contribution to the joint action and (2)  the other agent(s)
involved. The first aspect of this commitment captures the aformentioned
implication of intending; the second reveals the social orientation of these

[ 240 ]  From Cooperation to Deception and Disruption

  241

practical attitudes and corresponds to the relational structure of the occur-

rences of “we” in “We intend that we. … ”
Now, if we require that joint action be guided by an intentional struc-
ture constituted by intentions of the form “We intend that we … ” on
the part of each agent involved, and if we understand this as implying
that each undertakes a social commitment to the joint activity and to the
other(s) involved, then we are led to accept that joint action involves a
kind of normative relationship. The normativity in play here encompasses
both the norms that govern the performance of intentional activities and
the social norms that govern attitudes toward other agents. Being com-
mitted to another agent, as implied in joint action on this account, does
not presuppose an agreement, but it can arguably be construed as being of
consequence for the respective other. For this kind of social commitment,
in conjunction with the commitment to contribute to a particular joint
action and especially when made explicit, gives the addressee a reason to
contribute. In the familiar case, if A signifies that they are committed to
B to move the piano with B, then this gives B a reason to move the piano
with A, very much like attitudes of reliance and trust can be reinforcing by
inviting the respective other to rely and trust. The temporal asymmetry
of this example is no precondition for this generation of reasons to be
in place.
The last element of this brief outline of an account that certainly needs
further elaboration and defense regards the connection between social
commitments and obligations. Assuming a scenario in which two agents
each undertake a social commitment as specified in the previous para-
graphs, what becomes decisive is that each gives the other a reason to coop-
erate with them. Unless they have a specific practical authority vis-​à-​vis the
other, which we can exclude from the typical case of joint action, an agent
cannot create an obligation for another. Each can, however, and especially
in the symmetric case in which both are socially committed or at least dis-
posed or ready to cooperate, change the other’s normative situation to the
effect that each agent is obligated to at least respond to the other’s com-
mitment, or event to contribute to the joint action. As full-​blown agents,
they are required to be susceptible to the reasons that apply to them, and
in the social case, this requirement is extended to comprise responsive-
ness to others’ commitments. Once the translation of this talk of reasons
into talk of obligations is granted, we have at our disposal a conceptually
conservative account of the social obligations involved in joint action. These
obligations pertain to responses to others’ commitments and to engage-
ment in joint action, where the force of the latter part of such an obligation
is supported by one’s own social commitments.

C o op e r at i o n a n d S o c i a l O b l i g at i o n s   [ 241 ]
242

This concludes the brief portrayal of some characteristic intentional and

normative structures of joint action from the perspective of the philoso-
phy of action. How exactly these outlines of a relational account of joint
intention and an account of the normativity of joint action in terms of
social commitment and social obligation extend to other types of distrib-
uted agency will have to be left to other studies. And the plausibility of
the reflections presented here may well depend on, or be supported by, a
synopsis with other perspectives on similar phenomena represented in this
volume.

REFERENCES

Alonso, F. M. 2009. “Shared Intention, Reliance, and Interpersonal Obligations.”

Ethics 119(3): 444–╉475.
Bratman, M. E. 1999. Faces of Intention: Selected Essays on Intention and Agency.
Cambridge: Cambridge University Press.
Bratman, M. E. 2014. Shared Agency: A Planning Theory of Acting Together.
New York: Oxford University Press.
Gilbert, M. 2014. Joint Commitment: How We Make the Social World. Oxford: Oxford
University Press.
Pettit, P., and D. Schweikard 2006. “Joint Actions and Group Agents.” Philosophy of the
Social Sciences 36(1): 1–╉22.
Roth, A. S. 2004. “Shared Agency and Contralateral Commitments.” The Philosophical
Review 113(3): 359–╉410.
Schmid, H. B. 2013. “Trying to Act Together: The Structure and Role of Trust in Joint
Action.” In The Background of Institutional Reality, edited by B. Kobow, H. B.
Schmid, and M. Schmitz, 37–╉55. Dordrecht, the Netherlands: Springer.
Schweikard, D. P., and H. B. Schmid. 2013. “Collective Intentionality.” The Stanford
Encyclopedia of Philosophy (Summer 2013 Edition), edited by Edward N. Zalta.
http://╉plato.stanford.edu/╉archives/╉sum2013/╉entries/╉collective-╉intentionality/╉.
Searle, J. R. 1990. “Collective Intentions and Actions.” In Intentions in Communication,
edited by P. Cohen, J. Morgan, and M. E. Pollack, 401–╉415. Cambridge,
MA: Bradford Books, MIT Press.
Tuomela, R. 2005. “We-╉Intentions Revisited.” Philosophical Studies 125:327–╉369.

[â•›242â•›]â•… From Cooperation to Deception and Disruption

  243

CHAPTER 25

Deception as Exploitative Social Agency

R ADU UMBRES

W hen we deceive others, are we using them as tools? After all, when
we deceive, we use others as means to our ends, irrespective of their
intentions and welfare. But I  argue that such a view is only superficially
valid once we consider the mechanisms of deception as a form of social
agency. A  more insightful understanding of deception must account for
the interlocking of agency between deceiver and deceived at several levels,
based on uniquely human meta-​representational capacities.
The idea that deception means using humans as tools has widespread
intuitive appeal. Often, people deceived by others feel “used” or “played
with” or “powerless,” attributes better fitting an inanimate object than an
active, intentional agent. One of its guises, manipulation, strongly evokes
a puppeteer handling the mechanics of a tool-​like puppet. In slang, a “tool”
is an individual deemed too slow-​witted to realize that he is being taken
for a fool. The link between deception and tool use has even more presti-
gious advocates. Immanuel Kant opposed any form of lying (even when
the lie would save an innocent life) because it would contradict the first
formulation of the categorical imperative (to act only in accordance with
the maxim through which you can at the same time will that it become a
universal law) but also, and more relevant to our point here, the second
formulation—​that we are to treat others as ends in themselves, and not
only as means:  “The one who has it in mind to make a lying promise to
another will see right away that he wills to make use of another human
being merely as means, without the end also being contained in this other.
244

For the one I want to use for my aims through such a promise cannot possi-
bly be in harmony with my way of conducting myself toward him and thus
contain in himself the end of this action” (Kant 1993:429–╉430). While Kant
is making a moral point about deception, the liar’s conduct (and hence his
moral failure) toward the other evokes exactly the use of a device for one’s
own ends, without acknowledging the humanity of the victim.
Intuition and moral metaphysics aside, this chapter explores aspects of
deception that clearly distinguish it from mere tool use and bring it con-
ceptually closer to other forms of social agency. It is argued that decep-
tion builds upon co-╉opting several aspects of the victim’s agency, whereas
no agentive qualities are needed from a tool aside from its affordances.
Moreover, although deception feeds on an asymmetry of agency, it is not
merely a unilateral process. The target of deception may fend for himself in
ways that are unthinkable for an object. Together, these facets reveal the
entangled, reciprocally responsive agency of actors involved in deception
and counterdeception.

THE COMPLEXITY OF DECEPTION

The first thing about deception that sets it apart from treating others as
mere tools is its constitutive social nature. Few scholars if any would dis-
agree with the fact that deception is a form of social agency, yet it is con-
spicuously absent or undervalued in major works in the philosophy and
psychology of social agency. As illuminating as they are, developments
in the burgeoning field of studies and theoretical proposals dealing with
shared agency might suggest an unwarranted identity between “shared”
and “social” agency by overlooking a vast realm of sociality. Since the point
made here is a fairly general one, the umbrella-╉term “shared” covers many
forms of social agency. They may share only a “family resemblance,” includ-
ing concepts such as “shared,” “joint,” “collective intentionality,” “joint
action,” “common ground,” “joint commitment,” “group,” or “collective
agency” (Gilbert 1990, Searle 1990, and Tomasello et  al. 2005; Bratman
2014 is an exception, but even he tries to demonstrate how types of decep-
tion may still count as shared agency).
Deception and coercion are exemplary cases of those other forms of
agency that are irreducibly social but have features that set them apart from
the “shared agency” family. These different forms are ontologically based on
asymmetry rather than symmetry, on exploitation rather than mutuality,
on opposition rather than convergence, on competition rather than coop-
eration, on the circulation of incomplete or misleading information and

[â•›244â•›]â•… From Cooperation to Deception and Disruption

  245

imperfectly aligned or even conflicting interests. One could hardly consider

non-​cooperative social agency as a rare phenomenon, and, if you doubt its
relevance for human societies, consider the role of coercion in the slave-​
master relationship behind the building of the pyramids or the expanse of
totalitarian political regimes in the past century. As for deception, we can
only speculate how the crucial invasion of Sicily would have enfolded were
it not for Operation Mincemeat, in which the Allied forces tricked Nazi
commanders (including Hitler) that the landing target was Greece, staging
a body washed up on the shores of Spain who appeared to be a dead spy-
master stacked with orders carefully faked by the British.
One of the possible reasons why deception stays under the radar of
scholars of social agency is the weight given to human cooperative incli-
nations as the key to our species’ capacity for culture and for building
and abiding by social institutions. In contrast, deception figured more in
the scientific mapping of cognitive capacities of nonhuman primates and
other animals, especially tactical deception, which brings certain primates
closer to human capacities for attending to cues such as field of vision or
attention, or even reading minds. While some scholars discussed to what
extent chimpanzees deploy a Machiavellian intelligence with deception as
a crucial aspect of competitive interactions (Whiten and Byrne 1997), the
uniqueness of human thinking was attributed to our cooperative inclina-
tions unparalleled among other primates (Tomasello 2014).
The role of cooperation for human social agency could hardly be denied,
but this only makes human deception an even more complex phenomenon.
Compared to humans, the most Machiavellian of chimps deploys rudimen-
tary ruses, just as limited as its cooperative scope. Human deception may
be the opposite of human cooperation, but it is causally built upon the same
cognitive capacities, exploits the same propensities for acting and thinking
together, and even mimics or parasitizes cooperation to exploitative pur-
poses. In fact, the cooperative aspect of human agency (in forms of trust,
shared symbols, expectations, and obligations of mutuality, etc.) opens up
vast possibilities for exploitation that are unavailable for less cooperative
species. Because the engine of cooperation depends on the mutual good
faith of parties, a deceiver may mimic a bona fide cooperator and allow the
other party to (mistakenly) think that the goal of interaction is common
and mutualistic, but without having the positive, shared, “we” intentions
of true cooperation (Tomasello 2014). In a foundational work, Nicholas
Humphrey (1976) has argued that human intelligence evolved as a func-
tion of social complexity. The presence of cooperation does not just make
human sociality more complex in cognitive terms (in contrast with other
primates). The dramatic effect is that cooperation extends the capacity of

De c e p t i o n a s E x p l o i tat i v e S o c i a l   Ag e n c y   [ 245 ]
246

social agency to involve a huge array of flexible and interweaving processes

ranging from full-╉on cooperation to pure competition with myriads of pos-
sibilities in between.

DECEPTION AS SOCIAL AGENCY

If we take agency to be “the relation between a person and a course of action

and its effects” (Enfield 2013), the provisional definition of deception
that I propose and evoke in empirical examples is the act of an agent (the
deceiver) to manipulate information in order to use another agent’s agency
(the deceived) without the victim’s awareness or consent. Deception is thus
a form of social agency in which two (or more) actors have their individual
agencies interlocked in a chain of information manipulation. To further
understand the relationship, we may further unpack the concept of agency
into flexibility and accountability (Kockelman 2007, Enfield 2013) to dis-
cover the kinds of asymmetries exploited by deception as social agency.
As a starting point for understanding deception, the subjective inter-
ests of the agents involved are not aligned (or not aligned enough for the
purposes of the deceiver). This is important to rule out cooperation with
its relatively balanced distribution of interests as a flexible alternative to
deception. Unless we think of all deceivers as pathological (which they are
obviously not), deception starts from an imbalance between the initiator’s
and the target’s desires that cannot be solved efficiently through coopera-
tion or other means. Consequently, the deceiving agent pursues a course of
events that builds upon and exploits to his own advantage an asymmetry
with the victim at the level of knowledge and beliefs. The epistemic imbal-
ance is clear: the deceiver knows something more than the victim, in other
words, that an act of deception is under way, that certain signs or symbols
are false or misplaced, that the victim will act against its own interests, and
so on. The deceiver either does nothing to correct the error of the victim or,
most often, actively creates that error in order to profit from the imbalance.
Despite the imbalance, deception relies on the interlocking of individ-
ual agencies, leading to causal effects beyond the scope of any single one
of them. A simpler way of putting this is that neither party brings about
by itself the deception, just as it is true for any communicative or social
action. On the deceiver’s side, he or she must manipulate something in
the cognitive environment of the deceived. On the other side, the deceived
would not pursue the course of action by him-╉or herself were it not for
the manipulation. Standing by silently while municipal workers water your
plants mistakenly taking them for public property may engage someone’s

[â•›246â•›]â•… From Cooperation to Deception and Disruption

╇ 247

agency in your favor by means of misinformation, but you did not deceive
anyone—╉the workers are simply mistaken. Contrast this with the case
of, for example, a Nigerian inheritance scam, where the victim acts and
(falsely) interprets the situation as a cooperative event, apparently acting
jointly with the deceiver, but in reality playing into his or her hands.
In contrast, tool use may be considered a part of an individual’s agency.
Through its affordances, its embodied history and origin, an object may be
part of the distributed agency of an individual or a collective of individu-
als (Hutchins 1995). The elements of the cockpit form a constitutive part
of the pilot’s agency in flying a plane as a ship may be part, and mediate,
the social agencies of a crew. The tool becomes part of individual or social
agency as long as it is engaged into cognitive and physical work by at least
one human being. Deception, on the contrary, is essentially social. It takes
(at least) two to create a deception (leaving “self-╉deception” aside), just as
it takes two to have a conversation or to jointly carry an object. This cre-
ates a discrepancy at the level of flexibility, as the tool, unlike the victim of
deception, may only be co-╉opted passively, while the deceived is induced
to make choices, comprehend, and communicate. If I  push you in front
of a trolley (perhaps to save five people as a good utilitarian), I  am not
engaged in any social agency, but merely use your body as a tool for derail-
ing the trolley. A different event happens if I persuade you that it is safe
to lay on the tracks by your own volition. Between victim and perpetrator
there is an imbalance of epistemic power, but they are still ultimately cog-
nizing actors, while a tool is not. From the second perspective on agency,
the proper accountability of tools is usually derived from their users. An
Azande granary is the tool used by the sorcerer to kill its victim, but the
legal and causal actor is the (invisible) sorcerer.

DEFENSE AND ACCOUNTABILITY IN DECEPTION

The previous section argued that deception is a form of social agency whose
flexibility involves the imbalanced relationship at the level of knowledge
between the agencies of two actors. The epistemic asymmetry is key to the
success of deception, as the deceiver controls the event and adjusts his or
her inputs according to how he or she anticipates the behavior of the vic-
tim. Yet the victim is not a simple tool, as he or she contributes with his
own interpretation of facts, his or her own sources of information, and
his or her own capacity for choice and control. Moreover, deception is not
the automatic outcome of a single agent—╉the deceiver—╉since the target is
never just a sitting duck.

De c e p t i o n a s E x p l o i tat i v e S o c i a l   Ag e n c y â•… [â•›247â•›]

248

Sperber and colleagues (2010) have proposed a model of epistemic

vigilance that exposes the flexibility in the (potential) victim’s agency. The
deceiver may try to manipulate information, but the victim is not a pas-
sive recipient (a tool, one might say). Sperber et al. argue that the human
mind is endowed with a set of mechanisms geared for filtering true and
relevant information. Such mechanisms have evolved to avoid both inten-
tional deception as well as unintentional misinformation from malevolent
or incompetent sources. Epistemic vigilance consists of assessing a poten-
tial deceiver’s reputation and subjective interests (do not trust when a man
you know to be a robber tells you to go left into the woods), as well as for
consistency between content and prior knowledge (do not believe that this
hen for sale lays three eggs a day when no hen ever did that).
Such defense mechanisms enormously complicate the interlocking
of the agencies of deceiver and deceived. Actors deploy layers of meta-​
representations in deception and counterdeception (I know that she knows
that I know that she … ), flexibly adjusting their choices and symbols to
the other actor’s inputs. Such to-​and-​fro is unthinkable for the relationship
between a tool user and a tool, where the flexibility is only attached to the
human element.
If the victim is not merely a passive element in the interlocking social
agency that characterizes deception, is he or she to blame? Who is account-
able for the deceptive act? What are the rights and the duties attached to
deception? Here, cultural variation can enlighten our understanding of the
agency behind deception. A Lebanese village society (Gilsenan 1976) abides
by a rule of generalized potential deception. “Everything is kizb,” an infor-
mant declares as a justification for a world in which people patiently plan
the humiliation of others using lies and embellishment of facts, where an
epistemically vigilant commoner uses a trick to reveal the dishonesty of
a visiting sheikh, where everyone needs, and is expected, to lie at least a
little bit to keep up pretenses of honor. Likewise, in the Romanian village
I have studied for the past ten years, deception coming from people outside
a circle of trust and reciprocity is an ever-​present danger and makes a way
of life from a heightened sense of mistrust. What makes such communities
special is that the blame for being deceived often falls upon the victim. An
ethic of “devil take the hindmost” applies to everyday encounters includ-
ing communicative acts in which one ought to interpret the utterances of
others with utmost vigilance and adequate information, lest he or she be
deceived by a village competitor.
Apportioning (at least part of the) blame to the victim, villagers seem to
ascertain the role played by him in the social agency underpinning decep-
tion. Had the victim been more competent, more astute, better prepared

[ 248 ]  From Cooperation to Deception and Disruption

  249

by his peers, the deceiver would have failed. In a way, the deceiver did what
is expected of most people—​to further his or her interests even at the
expense of others, using such means as possible and useful. To try a com-
parison, deception flies in the air just as your inbox swells with Nigerian
scam letters. One could hardly stop feeling that the unfortunate client of
such scams is at least an active party to the deception with inadequate
epistemic vigilance in this day and age. The same apportioning of blame is
used in ritual pranks of initiation, in which the incompetence of the victim
justifies and legitimates the social and epistemic power of the tricksters
(Umbres 2013).
But, to return to our initial problem, why do people feel that deception
involves them as mere tools of the deceivers? I  conclude with a specula-
tion about the accountability of deception. In modern, centralized societies
driven by the universalistic morality of religion, or citizenship, or Kantian
ethics, people do not expect fellow humans to deceive them, at least not
as a rule. The norms of social interaction prescribe (at least minimally)
cooperative agents, engaged in true and relevant communication. They feel
obligated to speak the truth (or at least not to tell a lie knowingly) and feel
entitled to the same reciprocal treatment. The social contract of everyday
interaction punishes deceivers, either formally or informally, from jail sen-
tences to social ostracism. Not all social settings, of course, require such
a social norm. Strudler (2009) argues that deception is socially accepted
when used as legitimate self-​defense. The norm of trust1 does not apply
automatically and exhaustively in legitimate negotiation maneuvers such
as communicating a false reservation price when buying a house.
In high-​trust societies, the sentiment of reciprocity in truth-​telling
makes everyone lower their epistemic guards, leading to huge savings in
the transaction costs of communication and social interaction, and fur-
thermore to successful and efficient cooperation (compare with Keenan
1976 for a culture in which truth and relevance are not conversational
postulates). But the reverse of this generalized amnesty of distrust is ced-
ing part of our agency to others. When we trust someone, we give him
or her power over our behavior, over our representations, over the flex-
ibility of our agency. We do this in the expectation that her or she (1) will
be benevolent and cooperative and (2) would do the same for us. This is
why deception appears, for such enlightened altruists, as a way of using
people as tools, as means to private, egoistic ends. The comparison helps
us understand the way deception is accounted for in some societies, but it
is misleading as a metaphor or theoretical inspiration. In deception, as in
any form of social agency, it takes more than one individual to act and to be
responsible for their actions.

De c e p t i o n a s E x p l o i tat i v e S o c i a l   Ag e n c y   [ 249 ]
250

ACKNOWLEDGMENTS

The writing of this chapter began while I was Fyssen Foundation postdoc-
toral fellow at Institut Jean Nicod, Paris. I am grateful to Nick Enfield and
Dan Sperber for their comments on draft versions, and I thank the partici-
pants of the “Foundations of Social Agency” scientific retreat organized by
the Max Planck Institute for Psycholinguistics for their suggestions.

NOTE

1. Strudler also links trust with a yielding of autonomy, and deception as breach of
trust with a compromise of autonomy, lending further support to the idea that
deception is a form of intertwining social agency between parties.

REFERENCES

Bratman, M. E. 2014. Shared Agency: A Planning Theory of Acting Together.

New York: Oxford University Press.
Enfield, N. J. 2013. Relationship Thinking: Agency, Enchrony, and Human Sociality.
Oxford: Oxford University Press.
Gilbert, M. 1990. “Walking Together: A Paradigmatic Social Phenomenon.” Midwest
Studies in Philosophy 15(1): 1–╉14.
Gilsenan, M. 1976. “Lying, Honor, and Contradiction.” In Transaction and
Meaning: Directions in the Anthropology of Exchange and Symbolic Behavior,
edited by Bruce Kapferer, 191–╉219. Philadelphia: Institute for the Study of
Human Values.
Humphrey, N. 1976. “The Social Function of Intellect.” In Growing Points in Ethology,
edited by P. P. G. Bateson and R. A. Hinde, 303–╉317. Cambridge: Cambridge
University Press.
Hutchins, E. 1995. Cognition in the Wild. Cambridge, MA: MIT Press.
Kant, I. 1993. Grounding for the Metaphysics of Morals. Indianapolis: Hackett.
Keenan, E. O. 1976. “The Universality of Conversational Postulates.” Language in
Society 5(1): 67–╉80.
Kockelman, P. 2007. “Agency.” Current Anthropology 48(3): 375–╉401.
Searle, J. R. 1990. “Collective Intentions and Actions.” In Intentions in Communication,
edited by P. R. Cohen, J. L. Morgan, and M. E. Pollack, 401–╉415. Cambridge,
MA: MIT Press.
Sperber, D., F. Clément, C. Heintz, O. Mascaro, H. Mercier, G. Origgi, and D. Wilson,
2010. “Epistemic Vigilance.” Mind and Language 25(4): 359–╉393.
Strudler, A. 2009. “Deception and Trust.” In The Philosophy of Deception, edited by
C. W. Martin, 139–╉152. New York: Oxford University Press.
Tomasello, M. 2014. A Natural History of Human Thinking. Cambridge, MA: Harvard
University Press.

[â•›250â•›]â•… From Cooperation to Deception and Disruption

  251

Tomasello, M., M. Carpenter, J. Call, T. Behne, and H. Moll. 2005. “Understanding

and Sharing Intentions: The Origins of Cultural Cognition.” Behavioral and
Brain Sciences 28(5): 675–​691.
Umbres, R. 2013. “Chasse au dahu et vigilance épistémique.” Terrain 61: 84–​101.
Whiten, A., and R. W. Byrne. 1997. Machiavellian Intelligence II: Extensions and
Evaluations. Cambridge: Cambridge University Press.

De c e p t i o n a s E x p l o i tat i v e S o c i a l   Ag e n c y   [ 251 ]
  253

CHAPTER 26

Disrupting Agents, Distributing Agency

CHARLES H. P. ZUCKERMAN

INTRODUCTION

One afternoon in Luang Prabang, Laos, I  was gambling with three men
over a game of pétanque. For the unfamiliar, pétanque is a game played
like bocce or lawn bowling and it is a common way for people—​especially
men—​to gamble in Luang Prabang.1 On this day, I  filmed as we played.
Our game had drawn an audience of a dozen or so. The spectators lined
the benches—​some of them gambling, some of them just passing the time
and enjoying the show. About an hour into the game, my teammate Bii
began to take a shot when one spectator named Can2 wandered from his
seat on the bench and onto the court. Can then suddenly lifted his arms
and screamed. Startled, Bii missed his shot. He glared at Can and barked
accusatively, “You!”3 The audience burst into laughter and Can scampered
back to his seat as the two exchanged threats.
Clearly, Can was trying to distract Bii and make him miss the shot. His
scream was not merely a scream but an especially aggressive heckle in a
match already brimming with trash talk. A few seconds after Can’s heckle,
two audience members imitated Bii flinching in reaction: one man jerked
his head, mimicking Bii’s shaking body, while the other commented, “He
was surprised, all right, he went like this,” twitching his own arm to reenact
Bii’s flinching arm.4
To everyone, there was no doubt that Can’s heckle had caused Bii to miss
the shot, that it had affected Bii’s body and surprised him in ways beyond
254

his control. But when I watched the video recording of the event, matters
became less clear. Despite my repeated viewings, the recording showed no
signs of Bii flinching. Instead, the evidence that Bii was startled resided
entirely within Bii’s and the spectators’ own reactions. Bii had responded to
the heckle in a manner implying it was effective and audience members had
said that the heckle was effective, that it had made Bii flinch. While Can’s
scream was so loud, well-​timed, and explicitly addressed toward Bii that it
seems unlikely the scream did not affect Bii’s shot, we cannot know with
any certainty whether it did have an effect. Put another way, it is impos-
sible to know whether Bii would have made the shot—​which he only barely
missed—​had Can not screamed.
There is an inevitable and irresolvable uncertainty here; one that is, in
fact, helpful to better understanding agency. Acknowledging that we can-
not always know what causes what guides us to move away from concep-
tualizing agency in terms of moments of causality isolated from human
interpretation. The question worth asking becomes not did Bii cause Can
to miss the shot, but how did Bii and the spectators make this causal
relationship—​real or fictive—​visible? In other words, how are causal rela-
tions understood and made apparent in interaction?
In this chapter, I explore the prevalence and subtlety of attributions of
agency on the pétanque court. By unpacking a video recording of a differ-
ent pétanque game, I show that a player can respond to a heckle such that,
through his response, he helps to frame it as effective or ineffective. In
other words, I show how people can retroactively attribute and (re)distrib-
ute agency through their responses to events. I argue that we should study
agency not as a static, perspective-​free property of the world (e.g., “In this
moment of heckling, agency is distributed across three actors”), but as part
of an ongoing semiotic process through which actors ascribe agency. To put
it simply, this chapter argues for an interactional approach to distributed
agency that treats the “distributed” in “distributed agency” as more verb
than adjective.

“TELLING STORIES”, DISTRIBUTING AGENCY

In his classic essay “Response Cries,” sociologist Erving Goffman has us

imagine a man walking along a busy street. While walking, the man trips on
a piece of broken sidewalk and catches himself. “Up to this point,” Goffman
writes, “[the man’s] competence at walking had been taken for granted by
those who witnessed him … [but] his tripping casts these imputations

[ 254 ]  From Cooperation to Deception and Disruption

  255

suddenly into doubt. Therefore, before he continues he may well engage in

some actions that have nothing to do with the laws of mechanics” (Goffman
1978:88–​89). He might smirk to himself, signaling to any onlookers the
uncharacteristic nature of the event, or “ ‘overplay’ his lurch,” suggesting
he was playing the clown. Or perhaps he will “examine the walk, as if intel-
lectually concerned … to discover what in the world could possibly have
caused him to falter.”
These potential responses, as Goffman puts it, “tell stories to” the man’s
stumble. In different ways, they guide how others view the stumble’s cause
just as they guide how others conceive of the man, whether as “professional”
or “drunkard.” The man’s “stories” point toward what caused the stumble
and away from what did not (e.g., momentary clumsiness or inebriation).
They serve to distribute and displace responsibility for the stumble.
This kind of “story”-​telling abounds in interaction. People often sig-
nal what has occurred, what is occurring, and what will occur: the causes,
effects, and agents that should be held responsible. While it might be
tempting to dismiss these “stories” as folk understandings of agency or
a cloaking of “real” causal relations, doing so would ignore the prevalence
and importance of “stories” in social life. These “stories” are the means
through which agency is established, communicated, and negotiated in
interaction. They are often not just depictions of causality but also the
evidence of it. Take Goffman’s example of the tripping man “overplaying”
his stumble. Where does the stumble end and the exaggeration begin?
Drawing such a line—​while sometimes satisfying—​only lures us into the
false premise that causal processes (i.e., the stumble) and attributions of
causality (i.e., the exaggeration) are necessarily distinct in time and space.
That is, it is sometimes impossible or at least unproductive to distin-
guish the “stories” that attribute responsibility for causal processes from
the processes themselves. Many of the “stories” people tell are not nar-
ratives but wordless acts—​smiles, looks, exaggerated falls—​that blur
doing and display—​actions doubling as interpretations.5 Of course, “sto-
ries” can be separate in time and space from the acts they depict. For
example, a stumbling man can say “That was funny” after he has tripped
instead of overplaying his lurch while in the midst of tripping. “Stories”
can also vary in how they attribute agency: referentially explicitly, tacitly,
et cetera. As I show below, even an act as small as a pétanque player ignor-
ing a spectating heckler can tacitly deny that the heckler is responsible
for a missed shot. Ultimately, however, how one tells a “story” matters, as
different methods of attributing responsibility can have different interac-
tional effects and different consequences for how agency is distributed.

Di s r u p t i n g A g e n t s , Di s t r i b u t i n g   Ag e n c y   [ 255 ]
256

Saj jaa

What I  have been calling “heckling” is what Lao pétanque players most
often call saj jaa.6 Saj jaa literally means “to apply medicine” or “to dose”
and it is a local label for actions on the pétanque court that disrupt and
destroy the ability of others to focus, play properly, or control their bod-
ies and minds. The trope is that saj jaa can affect its victims with the same
efficacy that a pill cures a headache or a dose of methamphetamines keeps
one awake. Although there is a range of techniques for saj jaa—​being loud,
standing too close, bickering about the score, telling an opponent he will
miss his shot, even betting—​the techniques all presumably share a goal:7
to disrupt the attention, focus, and calm of the targeted player; to affect his
“heart;” and to make him lose, both the game and his cash.
All players are vulnerable to saj jaa, but most agree that some are more
susceptible than others. The most vulnerable are said “not to have the heart
for it” (caj3 bòø-​daj4) or to be “weak-​hearted” (caj3 qòòn1). They “tremble”
(san1 kathùan2) in response to trash talk like “spring chickens” (kaj1 qòòn1)
with “soft, porous skin” (nang3 pùaj1). Hecklers make them feel “angry”
(caj3 haaj4) and “rushed” (caj3 hòòn4, literally, “hot-​hearted”), and they play
poorly as a result. Those better at resisting saj jaa are said to have stron-
ger, more capable hearts (caj3 khaw2 daj4). They are tough and inured, like
“experienced, older chickens” (kaj1 kêê1) with “tight, rubbery skin” (nang3
niaw3) and stay “calm and cool” (caj3 jèn3, literally, “cool-​hearted”) under
pressure, allowing them to play their best. A strong heart, as one longtime
pétanque coach told me, is a key trait in good athletes. Being invulnerable
to saj jaa is, thus, both valued and valuable.
While players are sometimes said to be habitually “hot” or “cool-​
hearted,” the terms “hot” and “cool-​hearted” are also used to describe tran-
sitory emotional states.8 People often comment upon how others appear to
be feeling. Does he look “cool” and relaxed, or does he seem “hot” and dis-
tracted? Observing players’ emotional states like this can provide valuable
information. For example, take a moment from a money-​gambling game: a
player rushed his way through multiple shots and his opponents howled
and heckled after each poor shot. Seeing this, one spectator remarked to his
friend, “Don’t bet on him, he’s hot!”9 Monitoring a player’s emotions can
also inform a heckler’s use of saj jaa. As one man put it, a good pétanque
player should observe his opponent’s “heart” to map out his weaknesses
and discover what kind of “dose” might best affect him.
One way a player can display his “heart” is in his responses to saj jaa.
These responses also can display his autonomy as an agent. If a player
maintains a “cool heart” in the face of medicine, he is likely to be seen as

[ 256 ]  From Cooperation to Deception and Disruption

  257

unaffected by the medicine, and thus, as a relatively autonomous agent.10

By contrast, if he gets “hot” when given a dose of medicine, he is likely to
be seen as being affected by the medicine.11
On the whole, saj jaa invites those witnessing and experiencing it to
discuss and, perhaps interpret, what has happened in causal and agentive
terms. Once it is clear a heckler has used medicine and that the player has
perceived it, the player’s next action is liable to be taken as a sign of the
extent of the heckle’s efficacy. His next action then becomes not merely
another action but a reaction to (or, in semiotic parlance, an interpretant of)
the heckle. This phenomenon—​where a sign frames a subsequent action
as a reaction—​is common in discursive interaction. Let’s say, for example,
that someone asks you, “What’s up?” and it is clear you have heard and
understood her; if you stay silent, she is likely to take your silence not as
simple inaction but as a cold “response,” as a noticeably absent, unfilled
pause and perhaps as a hint not to bother you.12 Similarly, once it is clear
that medicine has been applied and that a player has perceived it and
missed his shot, questions tend to present themselves: Why was the shot
missed? The general ebb and flow of the match or distraction?13
How a player responds to a heckle often answers these questions
(whether tacitly or explicitly). Ignoring a heckler entirely is paradigmati-
cally “cool.” It can be a sign that a player is not distracted, but it can also
serve as a tool for maintaining his focus. Much like taking notes during a
lecture can help the note taker pay attention while also displaying focus
to others, not looking at, or not responding to, a distraction can help one
concentrate while simultaneously displaying concentration to others. Not
ignoring the heckler, in contrast, is paradigmatically “hot.” Talking to or
looking at a heckler, for example, is often taken as evidence that his medi-
cine is working.14 Because of this, players often tout ignoring the heckler
as the ideal. As one player put it, “No, [I never respond to hecklers]. [When
I play], I’m not interested [in medicine]. That’s me. I pay attention to [the
shots I take] and that’s it. I’m not interested in anything [else].”15

“WHO DO YOU THINK YOU ARE PLAYING WITH?”

I now turn to an example in which a player ignores his heckler and, in

doing so, tacitly distributes agency. Taa, a man in his late thirties, is play-
ing pétanque for money against Phuumii, a man in his forties. During the
game, Phuumii repeatedly uses medicine on Taa, while Taa mostly ignores
him. Below I present two transcripts of the event: one a more traditional
interlinear transcript familiar to linguistic anthropologists, the other a

Di s r u p t i n g A g e n t s , Di s t r i b u t i n g   Ag e n c y   [ 257 ]
258

more cartoonish transcript, created by loosely outlining screenshots of my

video recording of the event.16 I then walk the reader through these tran-
scripts and show how Taa’s seemingly minute actions serve to distribute
agency on the court.
But first, a basic outline of the event. It is late morning and I am film-
ing on one of Luang Prabang’s money-​gambling pétanque courts. Taa and
Phuumii are playing in front of an audience of about a dozen.17 The audi-
ence members are scattered at the edges of the compacted-​dirt court: some
lean against nearby trees, others sit on wooden benches, and a few lounge
on a newly acquired couch. They watch Phuumii and Taa’s game with vary-
ing degrees of focus and concern. Some are betting on the outcome and
watch the game closely, while others chat casually.
Where my transcript begins, Phuumii is on the verge of defeat. He then
makes a good shot and offers to bet Taa additional money. This offer to
bet (Figure 26.1: 1) is, among other things, a dose of medicine18 and Taa
treats it as such, rejecting it out of hand: “Hah … who do you think you are
playing with?” (Figure 26.1: 2), he says. Taa then takes his shot and misses
(Figure 26.2: 3). His miss seems to encourage Phuumii, who now even more
enthusiastically offers to bet him. Taa ignores Phuumii’s offers and makes
his second shot.
To summarize, during this short segment of interaction, Taa misses his
first shot and makes his second, Phuumii heckling both times.
Taa’s disengagement from Phuumii “tells a story to” his first missed
shot and to Phuumii’s heckles. The “story” is that Phuumii did not make
Taa miss, that he has not gotten into Taa’s heart. As is often the case, Taa
tells this “story” not through explicit comments but with a series of move-
ments, cries, and partially addressed utterances.
As mentioned above, throughout this entire segment of interaction, Taa
directly addresses Phuumii only once (Figure 26.1: 2), with a softly spoken
response to Phuumii’s offer to bet: “Hah … who do you think you are play-
ing with?” With this retort, Taa treats Phuumii’s offer as a frivolous kind of
saj jaa, meant to distract him and make him shoot poorly rather than put
more money on the line; in reminding Phuumii who he is playing with, he
dismisses and rejects Phuumii’s offer out of hand. And while this response
is addressed to Phuumii, there is some evidence that Taa is still not fully
engaged with him. He does not look at Phuumii as he speaks but stares
straight ahead at the pétanque balls.19 Taa also speaks much more softly
than Phuumii, an incongruence that suggests an interactional distance
between them. This partially oriented rejection of Phuumii’s offer fore-
shadows what is to come: Taa’s complete lack of engagement with Phuumii.

[ 258 ]  From Cooperation to Deception and Disruption

   259
TRANSCRIPTION 26.1 
260

Figure 26.1 Scene 1.

From lines 3 to 11b, Phuumii addresses Taa multiple times but Taa
never engages him. As Phuumii becomes more animated, Taa stares only
at the pétanque balls. When Taa does talk, he seems to address the court
generally, rather than Phuumii. Take, for example, his frustrated cry, “Oh!”
in line 6.  This cry happens immediately after the miss and is apparently
a reaction to it. It suggests Taa’s surprise and seems to be a kind of “self-​
remarking,” addressed to no one in particular (Goffman 1981b:97).20 That
this frustration is not addressed to Phuumii distributes responsibility for

[ 260 ]  From Cooperation to Deception and Disruption

  261

Figure 26.2 Scene 2.

the miss through a kind of negative evidence: Taa is frustrated at some-

thing for making him miss, perhaps himself, but not at Phuumii.
At every turn, Phuumii meets Taa’s aloofness with exaggerated engage-
ment. In lines 7–​8 (Figure 26.3), Phuumii encroaches on the space that Taa
is going to throw toward (and where Taa is looking) and points to the target
ball. Phuumii’s pointing gestures hover over where the pétanque balls sit,
invading Taa’s line of vision. Phuumii shouts at Taa and lies to him: he says

Di s r u p t i n g A g e n t s , Di s t r i b u t i n g   Ag e n c y   [ 261 ]
262

Figure 26.3 Scene 3.

that if Taa can knock the target ball off the court, he will score five points,
which is not true. Phuumii floods the court with stimuli addressed to Taa,
and retreats from where the pétanque balls sit only just before Taa takes
his shot.
During Taa’s second turn in the game (Figure 26.4), when he successfully
hits the target ball, he continues to ignore Phuumii and again does several
things that tacitly attribute responsibility for his initial missed shot to him-
self and not Phuumii. For example, as he throws his ball and it flies through
the air, he and Phuumii yell simultaneously: Taa yells, “[Knock] the little
ball [out]!” (line 11a), and Phuumii yells, “Hah!” (line 11b). Phuumii’s “Hah!”
again floods the court with stimuli directed toward Taa. Taa’s shout, in con-
trast, works more subtly: in this situation, Taa would not want to knock the
little ball, also called the jack, out at all (as this would score him fewer points
than if he knocked his target ball out); he is thus shouting a kind of anti-​
wish, expressing what he does not wish to happen. Anti-​wishes like this are
not uncommon on the pétanque court and function to anthropomorphize
and challenge Taa’s pétanque ball as it flies through the air; perhaps akin to
telling a problematic copy machine, “Jam again, I dare you.”21

[ 262 ]  From Cooperation to Deception and Disruption

  263

Figure 26.4 Scene 4.

When Taa does make his shot (Figure 26.5), he walks back to the other
end of the court and says (line 8), perhaps a bit tongue-​in-​cheek, “Ohhh, [I
was] scared of hitting the little ball.” As he says this, and for the first time in
this segment of the interaction, he briefly glances at Phuumii. Like much of
what Taa has done, this utterance implies the ineffectiveness of Phuumii’s

Di s r u p t i n g A g e n t s , Di s t r i b u t i n g   Ag e n c y   [ 263 ]
264

Figure 26.5 Scene 5.

doses. Even though Taa might not actually have been “scared of hitting the
little ball,” his utterance nevertheless functions to foreclose the interpreta-
tion that he was scared of, or even attending to, Phuumii’s talk—​he was
scared only of hitting the little ball. This utterance also seems to explain
away Taa’s first, missed shot. Perhaps, it implies, Taa had missed that shot
because he had overcompensated in fear of hitting the “little ball.”
That Taa successfully made his second shot helps him tell a “story” about
his first, missed shot. Because Phuumii heckled Taa during both shots,
Taa’s ability to make his second shot while being heckled implies that the
cause of his first, missed shot, was something other than heckling.
In the interaction as a whole, Taa and Phuumii are talked about, and
treated as, different kinds of agents. While Taa comes off as cool and in
control, Phuumii exhibits a kind of frantic “hotness” aimed at sensorial
overload. During my research stint more generally, I noticed that Phuumii
tended toward such states; others said so as well. During this game, in fact,
a number of spectators commented on the ragged state of Phuumii’s heart.
One man shouted out, for instance, just before Phuumii was offering to
bet Taa and the above interaction began, that Phuumii’s heart was “con-
fused” and “disordered,” that he “couldn’t even sit still.”22 Later during the
game, Phuumii was again accused of losing his cool. He had just shouted

[ 264 ]  From Cooperation to Deception and Disruption

  265

and missed a shot when a spectator called out that in using medicine on
Taa, Phuumii had dosed himself instead. Like a chemist organizing danger-
ous chemicals and accidently spilling them on his skin, Phuumii was falling
victim to his own medicine.23 According to the spectator, Phuumii was an
agent of self-​destruction, “hot” and uncontrollable. Of course, ascriptions
of agency and responsibility like these can in and of themselves be effec-
tual. In saying Phuumii was “hot,” the spectators were perhaps themselves
“dosing” Phuumii; compare the frustration one feels when repeatedly told
to “calm down” during an argument.

CONCLUSION

“Stories” are told not only on pétanque courts. Much like a man stumbling
on a sidewalk can “tell a story” about why or how he tripped, people—​with
varying degrees of explicitness—​often frame what has happened and who
is responsible. This sometimes has serious consequences. As it does, for
example, in a court of law, often the most ritualized and important venue
for attributing agency. Take the state trial of the officers who beat and bru-
talized the late Rodney King, which resulted in the officers’ acquittal. In the
trial, the defense used a number of visual and linguistic devices to guide how
people viewed the video of King’s body being beaten. As Charles Goodwin
(1994:621) writes, the video of the beating was framed such that “a rise in
King’s body [was] interpreted as aggression, which in turn justifie[d]‌an
escalation of force [by the officers]”. In this case, the defense “told a story”
about the beating, and attributed responsibility for it to King, not the offi-
cers. They presented King’s writhing body not as the victim of the officers’
force, but as the entity with agency, in control of the situation.
If we take attributions of agency seriously, some dimensions of agency
rise to the fore. First, attributions of agency are often contested. People can
and often do disagree about who or what is responsible for an act, and, thus,
about how agency is distributed (see Hill and Irvine 1993). Furthermore,
some people may be better at telling their “side of the story,” because actors
have differential access to authority and the resources to represent agency
(e.g., legal training, ownership of a projector, etc.—​see Kockelman 2007
and Kockelman this volume for a discussion of the agency of representa-
tion itself). Would Taa’s “story,” for instance, be as convincing if he had not
made his second shot? Second, “stories,” or whatever we might want to
call attributions of agency, often distribute agency differently across time
and space. That is, agency and the distribution of agency are interactionally

Di s r u p t i n g A g e n t s , Di s t r i b u t i n g   Ag e n c y   [ 265 ]
266

emergent, not fixed properties of things or people. Third, agency is often

attributed gradiently. That is, the attribution of agency is not an all-​or noth-
ing-​act; people can be held responsible to different degrees. And, fourth
and finally, causality—​and what we might call agency24—​can be distrib-
uted to groups or dimensions of humans other than biological individuals
and to things we might not typically consider “agents.” In pétanque, for
instance, players and audience members often attribute responsibility for
a missed shot to a variety of things: slippery pétanque balls, stray bottle
caps, poorly placed stones, bad luck, inebriation, et cetera.
Following this approach, the relevant question becomes not so much
who is responsible for what, but, how does responsibility emerge interac-
tionally, how is it distributed, and how might these distributions be negoti-
ated and contested?

ACKNOWLEDGMENTS

The research behind this chapter was supported by a Fulbright Hays

Doctoral Dissertation Research Abroad fellowship and a Wenner Gren
Dissertation Fieldwork Grant. While preparing the chapter, I have received
guidance and assistance in a number of ways. Versions of the paper,
and the data behind it, were presented—​and vigorously and helpfully
discussed—​at the University of Michigan Linguistic Anthropology Lab, at
a Georgetown University Data Session, and at a vibrant National Capital
Area Linguistic Anthropology “mini-​conference” held at the University of
Virginia. I would particularly like to thank Barbra Meek and Mark Sicoli for
inviting me to participate in these events. Michael Lempert, Webb Keane,
Alaina Lemon, Perry Sherouse, John Mathias, Meghanne Barker, Jeffrey
Albanese, Kimberly Ang, Adrienne Lagman, Elizabeth Batiuk, and Dana
Nichols all read and commented on drafts of this chapter in one form or
another. Finally, I would like especially to thank the other conference par-
ticipants and the editors of this volume, Paul Kockelman and Nick Enfield,
for their support in regard to this project and beyond. To repeat the com-
mon “story”: all mistakes are my own.

NOTES

1. Because of limitations in space, I am here not addressing the way in which the
game fits into broader life in Laos—​its sociological, economic, and historical
context, its gendered dimensions, the discourses about its unseemliness, and so

[ 266 ]  From Cooperation to Deception and Disruption

  267

on. These topics are, obviously, hugely important to fully understanding what
is happening on the court. Note, too, that all of my examples come from games
in which men are playing for money (as opposed to beer, which is also often
gambled).
2. This name, along with all the other names of players in this chapter, is a
pseudonym.
3. In Lao, “caw4 nii4 naq1.” For transcription of Lao in this paper, I am following
the system outlined in Enfield (2007).
4. In Lao, “tùùn1 qoo1 con3 vaa1 hèt2 cang3 sii4 leej2 mùù4 kii4 naq1.”
5. See Duranti (2004:454) and Hill and Irvine (1993) for discussion of the
discursive construction of responsibility and the relation between responsibility
and evidence. Generally, the argument in this chapter builds on some prevalent
arguments in linguistic anthropology concerning “metapragmatics” and
“reflexivity” (e.g., Silverstein 1993). Others, especially Ahearn (2010), have built
on these arguments somewhat differently. Note that, in contrast to some classic
statements on agency from Ahearn (2001) and Duranti (2004), and in line with
the work of the editors of this volume, my argument moves beyond a primary
focus on grammatical categories (e.g., ergatives) and explicit framings of agency
to include the alleged suburbs of language: gesture, gaze, and so on.
6. Saj1 jaa3 is a verb, and the nominal form is kaan3 saj1 jaa3. For ease of reading,
I will refer to it as simply saj jaa, without tone markers.
7. This is a slight simplification, because sometimes actions that could be
considered medicine are claimed to be “only a joke.” In the final analysis, what
counts as saj jaa is, like agency, negotiated in interaction.
8. For further discussion of “hot” and “cool” heartedness in a similar context, see
Cassaniti (2009).
9. Emotional moments like this shape players’ reputations, but, of course,
reputations are not merely an aggregate of these moments. Some moments are
inevitably more lasting, salient, or important than others.
10. Or at least autonomous vis-​à-​vis the heckler.
11. Note that this is a simplification for multiple reasons. First, people can interpret
these signs in very different ways. Second, the distribution of agency is often
gradient. Players or audience members can, for example, downplay a heckler’s
agency by saying the heckler “joked just a tiny bit,” implying the missed shot
was due to the player’s heart, not the strength of the medicine used. In contrast,
people can characterize medicine as being so powerful that anyone would be
affected. In fact, this is what happened later in the interaction between Bii and
Can when Can screamed in Bii’s face. The scream was so intense, people said,
that of course Bii was affected, no matter the state of his heart.
12. See the concept of “conditional relevance” for further discussion (Schegloff
2007:20).
13. In regard to the ebb and flow of sports, I asked one man what he might say to
calm a friend down who is feeling “hot-​hearted.” He offered: “Don’t worry about
whether or not you are going to make the shot, who could possibly make every
shot in sports?”
14. Of course, players do at times deny that a heckle is getting to them—​sometimes
explicitly—​by directly engaging with the heckler.
15. In Lao, “bòò1 khòòj5 bòø-​son3 caj3 khòòj5-​qaq1 khòòj5 tang4 caj3 khaw5 tang4 caj3
tii3 lêqø-​kaø-​lêêw4 khòòj4 bòø-​son3 … son3 caj3 ñang3.”

Di s r u p t i n g A g e n t s , Di s t r i b u t i n g   Ag e n c y   [ 267 ]
268

16. These “cartoons” are not meant, in any sense, to be accurate portrayals of the
people involved; in fact, their main appeal, besides the ease with which they can
be read, is that they preserve anonymity as the drawings look quite different
from the people they represent.
17. Although each player has a teammate, neither of their teammates talk during
the course of this segment.
18. That is, Taa treats Phuumii’s actions as medicine, as should become clear below.
19. This is not unusual in sports, as players are generally expected to focus their
attention on the game; cf. Goffman’s (1981a: 134–​135; 1981b:112) notion of an
“open state of talk.”
20. Nor is this discourse-​oriented response cry generally used in situations in which
one is scolding or chastising another’s actions, but instead, as Nick Enfield
writes, it tends to be used as “a news receipt which expresses disappointment or
concern” (2007:313).
21. Players often shout commands at the balls, addressing them with directives
like “Stop,” “Go,” and so on. For some similar themes, see Benjamin Smith’s
discussions of marbles and bad luck among Aymara children (2010:230;
­chapter 20 in this volume).
22. In Lao, “cit2 caj3 … ñung3 ñaak5 leej2 nòq1 laaw2 bòø-​saang1 juu1 lêq1.”
23. In Lao, “saj1 jaa3 khaw2 bak2 haajø-​haaj4 lêq1 tuaø-​qèèng tùùn1.”
24. I do not have space to address this question here, but merely point the reader to
the introductory chapters by Enfield and Kockelman of this volume and some
popular arguments that deal with the possibility of nonhuman agents (Gell
1998; Latour 2005).

REFERENCES

Ahearn, Laura M. 2001. “Language and Agency.” Annual Review of Anthropology

30: 109–​137.
Ahearn, Laura M. 2010. “Agency and Language.” Society and Language Use 7: 28.
Cassaniti, Julia. 2009. “Control in a World of Change: Emotion and Morality in a
Northern Thai Town.” Ph.D. dissertation, University of Chicago.
Duranti, Alessandro. 2004. “Agency in Language.” In A Companion to Linguistic
Anthropology, edited by A. Duranti, 451–​473. Malden, MA: Blackwell.
Enfield, N. J. 2007. A Grammar of Lao. Berlin/​New York: Mouton.
Gell, Alfred. 1998. Art and Agency: An Anthropological Theory. Oxford: Oxford
University Press.
Goffman, Erving. 1978. “Response Cries.” Language 54: 787–​815.
Goffman, Erving. 1981a. “Footing.” In Forms of Talk, edited by E. Goffman, 124–​159.
Philadelphia: University of Pennsylvania Press.
Goffman, Erving. 1981b. “Response Cries.” In Forms of Talk, edited by E. Goffman,
79–​122. Philadelphia: University of Pennsylvania Press.
Goodwin, C. 1994. “Professional Vision.” American Anthropologist 96(3): 606–​633.
Hill, Jane H., and Judith T. Irvine. 1993. Responsibility and Evidence in Oral Discourse.
Cambridge: Cambridge University Press.
Kockelman, Paul. 2007. “Agency.” Current Anthropology 48(3): 375–​401.
Latour, Bruno. 2005. Reassembling the Social: An Introduction to Actor-​Network Theory.
Oxford: Oxford University Press.

[ 268 ]  From Cooperation to Deception and Disruption

  269

Schegloff, Emanuel A. 2007. Sequence Organization in Interaction. Volume 1: A Primer in

Conversation Analysis. Cambridge: Cambridge University Press.
Silverstein, Michael. 1993. “Metapragmatic Discourse and Metapragmatic Function.”
In Reflexive Language: Reported Speech and Metapragmatics, edited by J. A. Lucy,
33–​58. Cambridge: Cambridge University Press.
Smith, Benjamin. 2010. “Of Marbles and (Little) Men: Bad Luck and Masculine
Identification in Aymara Boyhood.” Journal of Linguistic Anthropology
20(1): 225–​239.

Di s r u p t i n g A g e n t s , Di s t r i b u t i n g   Ag e n c y   [ 269 ]
  271

AUTHOR INDEX

Abric, J.-​C., 225 Brennan, Susan E., 63

Adams, Douglas, 65 Brewster, Paul, 195
Agnew, J. A., 217 Brighton, H., 162
Ahearn, Laura M., 8, 267 Brockbank, M., 170
Akrich, M., 221 Brooks, Rodney A., 165
Alonso, F. M., 239 Brown, D. J., 144
Amato, Massimo, 104 Brown, Penelope, 71, 75
Anderson, B., 218 Browning, Michael, 181
Arendt, Hannah, 41, 42 Burke, S. M., 146
Augé, Marc, 99, 100
Aureli, F., 11 Call, Josep, 12, 172
Austin, John, 47 Callon, M., 221
Avedon, Elliott M., 195 Carey, S., 171
Carlstein, T., 216
Bachelard, G., 216 Carpenter, Malinda, 172, 173
Bacon, Francis, 16, 17, 18 Casey, E. S., 214, 215, 216
Banks, C., 225 Chalmers, David, 60
Barker, Roger G., 222 Cheney, Dorothy L., 37
Bass, William, 181 Chomsky, Noam, 162
Bayat, Asef, 52 Cirelli, L. K., 175
Beebe, B., 174 Clark, Andy, 60
Behne, Tanya, 172 Clark, Herbert H., 12, 63, 75
Benjamin, Walter, 99 Clutton-​Brock, T. H., 131
Berliner, Paul, 200 Cohen, Emma, 153, 174
Berndt, Catherine H., 9 Cornish, H., 162
Berndt, Ronald M., 9 Costall, A., 184
Bingel, U., 147 Couper-​Kuhlen, Elizabeth, 74, 75, 79
Biro, S., 170 Cremer, S., 132
Blau, Peter M., 42 Cresswell, T., 216
Boecker, H., 142 Crown, C. L., 174
Bot, A. N. M., 132 Csibra, G., 170
Bourdieu, Pierre, 8, 194, 228 Curl, Traci S., 75
Bowles, Samuel, 75
Bramble, D. M., 143 Davis, A. J., 141
Bratman, Michael E., 236, 239, 244 Davis, M., 217
Brembeck, Helene, 196 de Jong, Eduard, 111
272

Devine-​Wright, P., 217 Heath, C., 144

Dingemanse, Mark, 67, 74, 79 Heinze, J., 132
Dittmer, J., 218 Henrich, J., 136
Drew, Paul, 74, 75, 79 Heritage, John, 91, 126
Du Bois, John W., 206 Hill, G. F., 113
Duncan, J. S., 217 Hill, Jane H., 265, 267
Duranti, Alessandro, 8, 267 Hölldobler, B., 132
Durkheim, E., 143 Honig, Bonnie, 42
Horst, Heather, 52
Ehrenreich, B., 152 Hudson, Michael, 113
Einarson, K. M., 175 Huffine, Edwin, 185
Eisenberger, N. I., 147 Hull, Matthew, 43
Ekkekakis, P., 146 Humphrey, N., 147, 245
Elyachar, Julia, 193, 196 Husserl, E., 213
Enfield, N. J., 49, 50, 53, 63, 74, 79, 80, Hutchins, Edwin L., 12, 221,
87, 95, 129, 162, 202, 203, 206, 246, 223,247
267, 268
Erasmus, Charles J., 195 Irigaray, L., 216
Evans-​Pritchard, Edward E., 205, Irvine, Judith T., 265, 267
206, 208 Ivens, A. B. F., 132

Fantacci, Luca, 104 Jaffe, J., 174

Fawcett, C., 174 Jakobson, Roman, 196
Feldstein, S., 174 James, J. E., 151
Ferllini, Roxana, 182 James, William, 18
Floyd, Simeon, 79 Janata, P., 154, 155
Foucault, M., 25, 73, 74, 76, 218, 224 Jasnow, M. D., 174
Fritz, T. H., 156 Jefferson, Gail, 74
Jefferson, Jon, 181
Garfinkel, H., 96 Johnson, M., 164
Gell-​Mann, Murray, 59, 60, 62, 63 Johnson, R. A., 132
Gell, Alfred, 13, 37, 268 Johnson, S., 171
Gergely, G., 170 Joyce, Christopher, 181, 182
Gibson, J. J., 221, 225
Gilbert, Margaret, 60, 238, 239, 244 Keenan, E. O., 249
Gilsenan, M., 248 Kelso, J. A. S., 162, 163
Gintis, Herbert, 75 Kirby, S., 162
Goffman, Erving, 75, 80, 254, 255, Kirschner, S., 153, 155
260, 268 Kitzinger, Celia, 75
Goodwin, Charles, 265 Klonowski, Eva-​Elvira, 185
Goodwin, Marjorie H., 196 Kockelman, Paul, 4, 7, 8, 49, 50, 53, 120,
Graeber, David, 101, 104, 105, 110 182, 187, 194, 200, 246, 265
Grau, Carles, 59 Koelsch, S., 152, 154, 155
Grice, H. Paul, 75 Koos, O., 170
Gross, S., 133
Guajardo, J. J., 173 Laidlaw, James, 13
Larkin, Brian, 50, 52
Hägerstrand, T., 216 Latour, Bruno, 13, 188, 195, 196,
Han, Clara, 104, 105 221, 268
Haney, C., 225 Lazzarato, Maurizio, 103
Hart, Keith, 109, 111 Leroi-​Gourhan, E., 221

[ 272 ] Author Index
  273

Levinson, Stephen C., 62, 71 Rao, Rajesh P. N., 59, 60, 62, 63

Lieberman, D. E., 143 Ratnieks, F. L. W., 135, 136
Linton, Ralph, 12 Reason, J. T., 224
Liszkowski, U., 173 Reddy, V., 173
Lollis, S. P., 174 Robbins, Bruce, 52
Rohlfing, K. J., 163, 165
MacWhinney, B., 162 Rosenblum, K., 96
Malchukov, Andrej Lvovich, 80 Ross, H. S., 174
Manzo, L. C., 217 Roth, A. S., 240
Maples, William R., 181
Markova, G., 173 Sack, R. D., 214, 215
Mashaw, Jerry L., 47 Sacks, Harvey, 61, 63, 74, 88, 
Massey, Doreen, 214 89, 96
Mayr, Ernst W., 21 Sadock, Jerrold M., 75
McGarity, Thomas O., 43 Sawday, Jonathan, 183
Melis, A. P., 146 Saxe, R., 171
Miles, L. K., 144 Schegloff, Emanuel A., 73, 74,
Milgram, Stanley, 12, 225 75, 267
Miller, Daniel, 52 Schmid, H. B., 235, 239, 240
Miller, David, 3, 4, 5, 6, 7 Scott, Helenus, 112
Morley, D., 217 Searle, John R., 12, 75, 235,
Moscovici, S., 225 236, 244
Murphy, Keith M., 196 Seyfarth, Robert M., 37
Murphy, M., 141 Shimizu, Y. A., 171
Mustafa, Mentor, 10 Siewierska, Anna, 80
Silverstein, Michael, 267
Nelms, Taylor C., 111 Simondon, G., 221
Newman, D., 218 Simone, AbdouMaliq, 52
Nicolelis, Miguel, 60, 64 Skinner, Mark F., 187
Nomikou, I., 163, 165 Skoyles, J., 147
Norman, D. A., 221, 225 Small, C., 156
Smith, Benjamin, 268
Ok, S-​J., 171 Smith, K., 162
Owen, David, 182 Smith, S., 217
Sober, Elliott, 75
Paasi, A., 218 Soja, E. W., 216
Passeron, J.-​C., 228 Sperber, Dan, 248
Pea, R. D., 221 Star, Susan Leigh, 50
Peebles, Gustav, 103, 115 Stivers, Tanya, 129
Peirce, Charles S., 18, 182, 184, 187, Stover, Eric, 181, 182
194, 200 Strassmann, J. E., 132
Peterson, Mark Allen, 51 Strudler, A., 249
Pettit, P., 234 Sullivan, P. J., 143
Piaget, Jean, 192, 196 Sutton-​Smith, Brian, 195
Pomerantz, A. M., 89, 91, 96 Swartz, Lana, 111
Poulsen, M., 132 Szymanski, Margaret H., 80

Queller, D. C., 132 Taylor, Charles, 19

Taylor, J., 141
Radcliffe-​Brown, Alfred R., 96, 142 Thomas, Peggy, 181
Raichlen, D. A., 143 Thorsteinsson, E. B., 151

Author Index  [ 273 ]
274

Tomasello, Michael, 12, 75, 146, 153, Wagner, Sarah E., 185

155, 172, 173, 244, 245 Wallot, S., 173
Tomlinson, Gary, 152 Wan, S. J., 175
Trainor, L. J., 175 Wenseleers, T., 136
Trimper, John B., 60 Wheeler, W. M., 132
Tschinkel, W. R., 131 White, R. A., 141
Tuan, Y.-​F., 213, 215, 217 Whiten, A., 245
Tuomela, R., 12, 236 Wilson, David Sloan, 75
Tzelnic, T., 171 Wilson, Edward O., 131, 132
Wiltermuth, S., 144, 153
Uexküll, J. von, 221, 225 Woodward, A. L., 173
Ugelvig, L. V., 132 Woolf, Virginia, 216
Urton, Gary, 195
Young, Antonia, 10
van Zweden, J. S., 135
Vinkhuyzen, Erik, 80 Zimbardo, P., 225
Vygotsky, L. S., 196, 221 Zwicky, Arnold, 75

[ 274 ] Author Index
  275

SUBJECT INDEX

accountability and agency, 228

causal, xii, 36, 37 environments, 28
in deception, 247, 249 installations as, xiv, 222, 223
distributed, xii, 10, 13, 104, 119 physical objects, 5, 19, 31,
economic, xii 225–​226, 244
and flexibility, 7 agreement, 238, 239, 241
moral, xii, 15, 21, 22 agriculture, 132
potential, 6 aition, 15
shared, 12, 13 Albania, 10
actions algorithms, xiv, 21, 28, 199,
accountability for, 17, 91, 94, 95 203–​205, 208
actions modifying actions, 73, 75 altruistic behavior, 72–​74, 249
in child development, 164, 166 Andes, 191
and consequences, 225 animal behavior, 65, 131, 132, 234, 245
coordinating, 152 animals. See monkeys, vervet;
in distributed agency, 67 nonhuman: primates
emplaced, 214–​215, 217, 218 animatorship, 81–​85
and interpretations, 255 anomalous behavior, 30
joint actions, 20, 169–​175, 234–​237 anthropology, 50, 105, 166, 267
physical, 119, 148 anthropology, forensic, 181–​189
and practical agency, 207 anticipation, 7, 8
public, 200 Apple, 110, 115
punishable, 9 ApplePay, 114
purposeful, 18 Aristotle, 15–​19, 213, 215
requests as actions, 79, 80 artifacts
adaptation, 134 and agency, 13, 119
adaptive agents, 21, 29 and flexibility, xii
biological/​environmental, 131, material and cultural, 60, 194,
160, 163 196, 221
social/​emotional, 174 physical, 224
social actions, 87, 129 textual, 203
adaptiveness, 21, 134 assistance, 67, 73, 81, 135
administrative states, 41, 42, 43, 47 offering assistance, 94, 95
adolescents, 71. See also children public, 217
affiliation/​affiliative behavior, 71, 72 See also help
affordances, 225, 247 authoring, 12, 13
actions as, 164 avoidance, 70, 71, 84, 88, 208
276

Aymara language, 191, 192, 196, 268 and social development, 175

Azande, 205, 247 See also adolescents; babies; infants;
toddlers
babies, 161, 163, 164, 165, 173. See also choice
children; infants; toddlers and accountability, 7
behavioral sciences, 72, 145, 163 adjusting, 248
biology, xi, 12, 225 and agency, xv, 28, 43, 226
Bitcoin, 110, 111, 112, 114. See also and ants, 134
digital currencies and brute force, 60
blame as human act, 20, 103, 208
apportioning, 45, 49, 101, 128, and presence of mentality, 18
248–​249 Christianity, 102. See also God; religion
inheriting, 9 citizenship, xiii, 51, 249
as an interpretant, 6 coercion, xi, 131, 136, 164, 222, 244, 245
sharing, 11 cognition, 18, 152
blind spots, 17 distributed, 12, 221
blockchain, 112 human, 170–​174
bodies, 112, 201, 222 social, 175
body parts, 144 coincidence, 205
collective, 218 coins, 109, 113
constraints of, 60 collaboration, 60
and control, 12 collective action, 67, 44
dead, 181–​188 collective agency, 72, 134, 169–​177,
“docile bodies,” 224 216, 234, 244
individual, xiv, 10, 59, 63 collective behaviors, 20
and minds, xi, 143, 225 collective mass, 44
and motion, 51, 155 collectives, 72, 234
physical, 13 commodities, xii
and selves, 129 communality, 141, 145
women’s, 216 competition, 34, 112, 244, 245, 246
Boston, 200 for animatorship, 81, 84
British, 245 in music, 201, 202
British Museum, 113 in sports, 146
bureaucracy, 41, 42, 44 complex networks, 64, 151
computerized algorithms. See algorithms
causality, 35, 36, 166, 254, 266 conflict, 43, 44, 135, 155, 184, 237, 245
causal processes, 27, 30, 33–​37, 255 contention, 71, 240
cause-​effect, 34 contingency, 22, 27, 100, 170, 199–​209
See also accountability: causal life contingencies, 104
cellular slime molds, 64 vs. necessity, 36
Chachi, 68, 71, 72, 73 control
Cha’palaa, 68, 70, 71, 72, 75 agents and agency, 37, 124, 125, 224
child-​rearing, 71, 103, 145, 166 and the body, 119, 122, 123, 55
children, 102, 169 controlled behavior, 7
and agency, 71, 196 and deception, 247
and collaborative activity, 146, 173 institutional, 51, 222
and family, 225 joint, 11, 235
grandchildren, 10 and language, 62
parental responsibility, 103, 106 neural activity, 59, 60
and play, xiv, 192, 196 physical, 12

[ 276 ] Subject Index
  277

situational, 264, 265 derivative agents, 18, 19, 22, 26

social, 223 Descartes, 25, 213, 215
convergence, 102, 244 digital currencies, 110, 111. See also
conversation Bitcoin
conversation analysis, 70, 73, DNA, 21, 185
74, 75, 87 dualism, 128
everyday, 61, 71, 121, 126, 127, 128 dyadic activity, 104, 152, 162
as joint action, 247 dyads, 71, 72
repair, 63
structures, 62 Ecuador, 68, 71, 73
cooperation, 233–​242 Egypt, 49, 51, 52, 53
agency and, xi, xii, 27 eliciting/​elicitation, 88–​93, 165
in ants, 131–​150 embodiment, xii, 119, 215, 247
by consent, 60 embodied agency, 142–​158, 226, 227,
human capacity for, 74, 85, 245–​251 228, 237
joint action in, 233–​242 embodied competencies, 223, 224,
practices, 68, 152 225, 226
principles, 72 emplacement, 213, 214, 215, 216, 218
shared goals, 65 enculturation, 29
social, 74, 134, 136, 145, 245 endorphins, 142, 143, 153, 154
corpses. See bodies: dead entitlement, 6, 11, 75, 227, 228, 235, 238
creativity, xii, xiv, 199 environment, 21, 27, 36, 37, 243
crises, xii, 114, 200 adaptation to, 131
cross-​disciplinary approach, xiii built/​urban, 216, 222
cultural evaluation, xii, 6, 13, 22 environmental psychology, 217
cultural scripts, 165, 166 material, 79
culture Environmental Protection Agency, 43
as collective creativity, xii envorganisms, 27
constraints of, 166 Erasmus, 195
cultures, 22, 51, 79, 145, 249 error, 12, 31, 246
and group activity, 148, 152, 155, 156 ethics, xv, 99–​107, 249
human capacity for, 245 ethnography, 71, 102, 106
inheritance of, 21 ethnographic contexts, 199–​200,
material, 60, 196, 221–​228 205–​207
popular, 60, 218 ethnographic knowledge, 50
and social interaction, 85 ethnographic methods, 104
See also enculturation eusocial organisms, xiv, 64, 131–​137
evaluation, xii, 6, 13, 22
dance/​synchronized movement, 142, everyday interaction, xv, 53, 68, 73,
143, 145, 151–​158 141, 249. See also
dark economy, 111 conversation: everyday
Dasein (being in the world), 215 evolution, 27, 30, 64, 145, 169
debt, 99–​107, 111, 114 and agency, 29
deception, 60, 62, 243–​251. See also of cooperative systems, 132, 136
control: and deception; lying cultural, 162, 165
decision-​making evolutionary accounts, 21, 72
collective, 134, 145, 234 evolutionary anthropology, 142
socially constructed, 42, 45 human, xiii, 143, 148
democratic state, 41, 112, 225 natural selection in, 22
deontic meaning, 8, 81–​85 exchange, 62, 105, 109, 114, 145

Subject Index  [ 277 ]
278

Exodus, 105 hardwiring, 134
explicit negotiation, 82, 103 help, 68, 153, 215, 221
accepting, xi
face, 70, 71 requesting, xii, 62, 79
Facebook, 51, 52, 53, 54, 115 See also assistance
family, 73, 131, 186, 227 high-​trust societies, 249. See also trust
and death, 184, 185 Hitler, 245
and interaction, 61–​62, 82–​83, Hobbes, 218
225–​226 homo sapiens, 21
internal dynamics, 75 honor, 19, 248
and payback, 10 human body. See bodies
feud, 10 humanity, 110, 113, 155, 244
financial crisis, 114
fission, 11 imagination, 18, 29, 102, 191
food, 18, 132, 133, 134, 146, 225 imitation, 6, 60
foraging, 132, 133, 135, 141, 152 immediate and mediate spheres, 36
force fields, 35 impositions, xiv
formal cause, 15, 17 improvisation, 199–​209
Foucault, 25, 73, 74, 76, 218, 224 independent action, 119, 227, 236
framing, 33, 99, 105, 200, 102 indigenous law, 10
and accountability, 6, 217 inductive reasoning, 30
and agency xiii, xiv, 15–​17, 22, 32 infants, xiv, 131, 143, 163–​167,
free will, xv, 223 169–​176. See also babies;
friendship, 63 children
fusion, 11, 13, 113, 128, 129 inferential agency, 33, 35, 36, 37,
75
games, 191–​197 infrastructure, 49–​55, 111, 114,
gaze, 120, 128 223, 227
gaze patterning, 122, 123 and agency, 35, 64–​65
in mother-​child interaction, 161, and interactions, 32
163–​166, 174 and the internet, 112
generativity, xv installations, xiv, 221–​228
geographic theories, 213 instigation, 26, 28, 36, 37
Ghana, 61 instincts, 21
goal-​directed behavior, 7, 12, 170, 171 institutions, 32, 67, 222, 223–​226, 245
goals instrumental acts, 19
assistance in, 67 instrumental agency, 18, 19, 22,
and installations, 223, 226 33, 36, 37
and intentions, 145 intentionality, xi, xiv, xv, 225
interactional, 88, 94 and agency, 233
others’ goals, 170, 171, 227 shared, 12, 60, 234–​235, 244
shared, xi, 13, 65, 169, 175 internet, xii, 51, 52, 53, 111, 112
God, 21, 102. See also Christianity; interpretants, 28, 34, 206, 207, 257
religion anticipation, 8
Google, 110, 115 definition, 25, 26, 27, 87
government, 10, 45, 105, 112 non-​anticipated, 4, 8
grammar, 79–​86 Peirce, 184, 194, 200
Greece, 245 sign-​object relations, 201, 202,
grounding, 63, 239 204, 205
group exercise, xiv, 141–​150, 156 and subprehension, 5–​6
group size, 11, 152 interpreting, 12, 43, 195, 225

[ 278 ] Subject Index
  279

interrogation, 64, 68, 69, 80 material culture. See culture: material

intuition, 15, 16, 244 membership
iPod, 208 group, 73, 144, 205, 206
in group exercise, 151, 156
joint action, xiv, 11, 65, 169–​177, hunting groups, 145
233–​242 members of a society, 73
and commitment, 12, 64 and punishment, 136
nonverbal, 70 workgroups, 43–​44, 46
and tools, 62 meta-​agents, 17, 20, 21
joint commitment, xi, 238–​242, 244 meta-​analysis, 146
money, 115 meta-​consciousness, 64
negotiated, 60, 64 meta-​pragmatics, 267
See also joint action meta-​representation, xiii, 248
joint enterprise, 9 meta-​representational capacities,
Judaism, 105 243
judgments, xii, 20, 46, 49, 186, 192 metaphor, 18, 106, 120, 128, 182–​183,
186, 249
Kant, 243, 244, 249 micro-​level context, 67, 73, 74, 161
kinship, 9, 63, 73, 100, 151 mind meld, 64
knowledge, 33, 103, 207, 247, 248 minds
acquiring, 201 vs. bodies, xi, 143, 225, 256
and beliefs, 246 reading, 245
common, 145, 237 money, 18, 45, 109–​116
ethnographic, 50 as credit, 17
externalization of, 214 gambling, 256, 257, 258
and power, 16, 17, 18, 21 monkeys, vervet, 26, 28, 37
multiple individuals, xiii, 9, 10, 11, 67,
Lajamanu, 9, 10 152, 154
language music, 145, 151–​158, 199–​209
and agency, xv, 59–​96 mutual obligation, 233, 235, 238–​239,
compositionality, 162 240, 245
as a human capacity, 18
languages, xiv, 120, 156 natural selection, xii, 16, 21, 22, 27, 28,
music as a language, 200 35, 136
resources, 121 natural speech, 68, 71. See also
law courts, 13, 41, 45–​56, 183, 185 conversation
laws, xiii, 7, 41, 45, 104, 227, 255 Nazi Germany, 245
learning, xv, 28, 165, 225 negotiation, 63, 65, 80, 85, 249, 256
in children, 164, 165, 169, 174 and compromise, 192
intergenerational, 162 and distributed agency, 67–​77
learned behavior, 37 explicit, 82, 83
Leviticus, 105 joint, 62
linguistic diversity, 31, 85, 152, 218 negotiability, 59, 62, 64, 65,
litigation, 42, 45, 46, 47 196
luck, 109, 195, 196, 266, 268 Netherlands, The, 111
lying, 243. See also deception Newton, 213, 215
non-​intentional actors, 16, 21, 22
manipulation, 85, 128, 243, 246, 248 nonhuman
Marx, 20, 21, 50, 73, 75 agents, 196, 199, 200, 228
Mastercard, 111 entities, 202, 203, 205, 206, 207, 208
material causes, 15 primates, 151, 152, 245

Subject Index  [ 279 ]
280

nonhuman (Cont.) primates. See monkeys, vervet;

as a social category, 192 nonhuman: primates
normative accountability, xii privatization, 51
norms, xii, 7, 192, 223, 240–​241, 249 programming, 29
Northern Territory (Aus.), 10 promises, 102, 103, 105, 239, 244
lying promise, 243
obligation, 13 prosocial behavior, 68, 72, 73, 74
debt, 104 punishment, xv, 10, 65, 136, 145, 218
deontic, 8
joint, 11 rationality, xiii, 21, 170
moral, 6 recipients, 88, 91, 94–​95, 169–​171, 248
mutual, 245 reciprocity, 152, 214
social, 80, 233–​242 reciprocal causality, 162, 166
obstacles, 85, 133, 172 recognition, 10, 88, 94–​95
ontology, 29, 31, 36, 234 and trust, 248, 249
ontological ambiguity, 181 recruitment
ontological assumptions, 30, agents, 94, 95
31, 36, 37 in ants, 133, 134
outcomes, 7, 147, 162, 172, 175, 206 goal directed, 79
ownership, 6, 265 as a request, 74
recursion, 17, 19, 20
parallel actions, 235 reduced agency, xv, 201, 203, 207
parasitic practices, xi regulations, xiii, 43, 165
participant observation, 105 religion, 21, 105, 148, 152, 249. See also
PayPal, 110, 114 Christianity; God
perception, 31, 164, 215 reneging, 12
of pain, 146, 147, 148, 153 representational capacities, xiii, 243
personhood, xv, 13, 102, 120, 128 representations, 33, 146, 223, 249
Peru, 191 auditory, 201, 202, 203
PET scans, 155 mental, 224, 227
phatic labor, 193 social, 225
pheromones, 133, 134 See also meta-​representation
placebo, 147, 148 reputation, 18, 146, 248, 267
planning, 165, 166, 203, 208 requests, xii, 62–​63, 67–​77, 80, 91–​92
and design, 7, 13 resources, 127, 128, 226, 265
vs execution, 5, 205, 206, 207 access to, 226, 227, 228
problems, 204 conceptual, 240
play linguistic, 62, 68, 121, 125
as collective behavior, 20, 143 material, 50, 143, 217
as coordinated activity, 148 symbolic, 217
distributed agency in, 191–​197 responsibility, 99–​107, 124
games, 61, 62, 169, 200, 253–​269 attribution of, 47, 120, 255, 260, 262,
and musical instruments, 29, 156, 265, 266
199–​209 collective, 234
and synchronous movement, 145 displaced, 255
pointing gesture, 9, 25–​26, 37, lack of, 85, 122
172–​175, 261 legal, 45
as a request, 67, 69–​70 physical, 128
politeness, 68, 70, 71, 75, 76 units of, 46
poverty of the stimulus, 162 revenge, 10, 18

[ 280 ] Subject Index
  281

revolution, xii, 20, 49–​55 social units, 43, 64, 135, 136

reward, xiii, 17, 36, 46, 112 socialization, 30, 53, 166, 196
intrinsic, 145 soul, 120, 128
motivating, 148, 155 space
and pain, 143 for action, 191, 192, 195, 201
psychological, 144, 146 bounded, 213
robot, 204 public and private, 213
rupee, 112, 113, 114, 115 and time, 35, 100, 110-​112, 214, 216,
255, 265
sabotage, xiv urban, 50, 53
sanctions, 4, 6, 7–​8, 96, 102 spirituality. See religion
scaffolding, 46, 222 Srebrenica, 185
selectivity, 60, 63, 64, 65 standards, 19, 20, 41
self-​consciousness, xv, 21, 37 state agencies, 41–​47
semiosis, 37, 200 stress, 142, 146
semiotic framework, 200 subjectivity, 42, 182, 215
semiotic mediation, 199–​209 subprehension, 4, 8, 12, 13, 206, 207
semiotic processes, 25, 26, 34, 37 supra-​mind, 237
serendipity, 16, 22 symbiosis, 132
shared attention, 154, 173 synchronized movement. See dance
shared intentionality, xv, 12, 60,
238, 244 technologies, 64
sharing, xiii, 9, 60, 105 and artifacts, xii, 196
attention, 170, 171, 172–​173 and communication, 52, 53, 148
Sherlock Holmes, 187 infrastructure, 51
sieving, 16, 22 new, 3, 6
signs See also artifacts
and agency, 18 teleological processes, 21, 22
as behavior, 4, 202, 203, teleomatic processes, 21, 22
206, 207 teleonomic agents, 21, 22
false, 246 telepathy, 65
in forensics, 184, 185, 186, 187 temporality, 100, 101, 104, 106
interpretants of, 25, 28 asymmetry in, 241
and objects, 188 framing, 99, 105
Siwu, 61 temporal regimes, 102, 103
small talk, 171 toddlers,169. See also children
society tools
agents as, ix, 13, 42, 60, 62 analytical, 162, 166
compound persons, 11 the body as, 62, 247
social bonding, 141–​150, 151–​158, communicative, 174
175, 176 language as, xiv, 63
social commitment, 240, 241, 242 people as, xii, 16, 243, 244,
social cooperation ( see 247–​249
cooperation: social) topics, 33, 163, 165
social identity, 73, 75, 227 topophilia, 217, 218
social insects, 131–​137 transformation, 29, 30, 31, 216, 217
social isolation, 151 in jazz, 201
social sciences, 50, 73, 99, 145, in life cycle, 102
163, 166 spaces transformed, 191, 192
social stratification, 73, 75, 216 transgression, 10

Subject Index  [ 281 ]
282

trust, 153, 239, 240, 241, 245 value-​oriented agents, 19

mistrust, 248, 250 values, 19, 20, 22, 44, 46
truth, 182, 183, 184, 249 violence, 105, 184, 185
Turing machines, 21 Visa, 111, 112
turn (conversational), 63, 74, 87, 174,
261, 262 war, 101, 104, 214
anti-​war protest, 3, 4, 5
unconscious, xii, 17, 20, 30, 147, 155 as collective behavior, 20, 132, 145
United States, 43, 45, 47, 101 WWII, 50
urban action, 216 Weber, 44, 73, 75, 95, 101
utterances, 68, 120, 165, 248, 258, 264 withholding information, 62

[ 282 ] Subject Index
  283
284
  285
286