Вы находитесь на странице: 1из 310

The Anthropology of Intentions

How and to what extent do people take into account the intentions of others?
Alessandro Duranti sets out to answer this question, showing that the role of
intentions in human interaction is variable across cultures and contexts.
Through careful analysis of data collected over three decades in US and
Pacific societies, Duranti demonstrates that, in some communities, social
actors avoid intentional discourse, focusing on the consequences of actions
rather than on their alleged original goals. In other cases, he argues, people do
speculate about their own intentions or guess the intentions of others, includ-
ing in some societies where it was previously assumed they avoid doing so.
To account for such variation, Duranti proposes an “intentional continuum,”
a concept that draws from phenomenology and the detailed analysis of face-
to-face interaction.
A combination of new essays and classic re-evaluations, the book draws
together findings from anthropology, linguistics, and philosophy to offer a
penetrating account of the role of intentions in defining human action.

a l e s s a n d r o d u r a n t i is Distinguished Professor of Anthropology at the


University of California, Los Angeles.
The Anthropology of Intentions
Language in a World of Others

Alessandro Duranti
University Printing House, Cambridge CB2 8BS, United Kingdom

Cambridge University Press is part of the University of Cambridge.


It furthers the University’s mission by disseminating knowledge in the pursuit of
education, learning and research at the highest international levels of excellence.

www.cambridge.org
Information on this title: www.cambridge.org/9781107652033
© Alessandro Duranti 2015

This publication is in copyright. Subject to statutory exception


and to the provisions of relevant collective licensing agreements,
no reproduction of any part may take place without the written
permission of Cambridge University Press.
First published 2015
Printed in the United Kingdom by Clays, St Ives plc
A catalogue record for this publication is available from the British Library
Library of Congress Cataloguing in Publication data
Duranti, Alessandro.
The anthropology of intentions : language in a world of others / Alessandro Duranti.
pages cm
ISBN 978-1-107-02639-1 (Hardback) – ISBN 978-1-107-65203-3 (Paperback)
1. Anthropological linguistics. 2. Language and culture. 3. Intention. I. Title.
P35.D857 2014
306.440 089–dc23 2014023513
ISBN 978-1-107-02639-1 Hardback
ISBN 978-1-107-65203-3 Paperback
Cambridge University Press has no responsibility for the persistence or accuracy of
URLs for external or third-party internet websites referred to in this publication,
and does not guarantee that any content on such websites is, or will remain,
accurate or appropriate.
Contents

List of figures page ix


Acknowledgments xi

1 Rethinking anti-intentionalism 1
1.0 Introduction 1
1.1 Reopening a dialogue that never took off 2
1.2 Self and other 5
1.3 Ethnopragmatics 6
1.4 Themes, issues, and intellectual connections 7
1.5 Title and expectations 8

2 Intentions in speaking and acting: the Standard Theory and its foes 11
2.0 Introduction 11
2.1 The Standard Theory: Grice, Austin, and Searle 11
2.1.1 The speech act of promising 12
2.1.2 Searle’s five types of speech acts and the notion of “direction of fit” 14
2.1.3 Searle’s model of intentionality 16
2.1.4 Searle’s notions of Network and Background: capturing implicit cultural
knowledge 17
2.2 The role of intention in defining “action” 19
2.2.1 Unintended consequences and the practical engagement with the world 20
2.3 Michelle Rosaldo’s critique of speech act theory 22
2.4 Brentano’s and Husserl’s use of intentions 25
2.5 Criticism of Husserl’s use of intentionality as the foundation of meaning-making 27
2.5.1 Husserl’s “passive synthesis” and “streaming living present” in the
unpublished manuscripts 29
2.6 A different theorist of speech as action: Ludwig Wittgenstein 30
2.7 The meaning of intention and intending 31
2.7.1 Translating intention: an exercise in Samoan ethnopragmatics 33
2.8 Not giving up on intentions 38
2.9 Alternative “western” ways 40

3 The avoidance of intentional discourse: a Samoan case study 43


3.0 Introduction 43
3.1 Avoidance of making explicit intentions and motivations 44
3.1.1 The 1978–79 project: Samoan language acquisition and socialization 45
3.1.2 The Laboratory of Comparative Human Cognition: neo-Vygotskian
perspectives 46

v
vi Contents

3.2 A brief introduction to the Samoan fono as a speech event and a social drama 48
3.3 Unintended consequences: Loa’s case, April 7, 1979 51
3.3.1 Loa’s defense 60
3.3.2 The relevance of positional roles 62
3.4 Absolute liability, cultural preference, and empirical evidence 63
3.4.1 Causes, reasons, and sources 64
3.5 Conclusions 67

4 The invention of promising in the Samoan translation of the Bible 69


4.0 Is promising a human universal? 69
4.0.1 Promise and promising in English 71
4.1 The invention of promising 72
4.1.1 Searching for promising in Samoan texts and dictionaries 75
4.2 Samoan translation of English promise in dictionaries 77
4.3 Challenges in the translation of the Bible into Samoan 80
4.3.1 Translations for the Christian notions of “soul”and “spirit” 84
4.4 Translations of promise in the Bible 88
4.4.1 Differences between translations of promise in the Old and
New Testaments 90
4.5 Promises in non-Christian Samoan texts 92
4.6 The case of the word māvaega 95
4.7 Promises, oaths, and public announcements 96
4.8 Conclusions 98

5 Intentionality and truth, revisited 101


5.0 Introduction 101
5.0.1 Anthropological thinking and writing 101
5.1 Truth as correspondence between the mind and the world 104
5.2 Intentionality as a relation between the mind and the world 107
5.3 The reconstructionist model of interpretation 107
5.4 Some problems with goals, plans, and intentions-to-do 109
5.5 Code-related limits on speaker’s intentions 111
5.5.1 An example of how a genre constrains what a speaker can say 114
5.5.2 Audience co-participation in shaping intentions 118
5.5.3 Culture-specific constraints on introspection and responsibility 120
5.5.4 Multiplicity of goals 121
5.6 Problems with the truth-functional model 123
5.6.1 Truth and turn-taking 125
5.7 Truth and affect 128
5.8 Truth and metacommunication 130
5.9 Truth and action 131
5.10 Conclusions 133

6 Speaker intentions and the role of the audience in a political


campaign in the US 135
6.1 The discovery of what we mean 135
6.2 The Capps-for-Congress Campaign 137
6.3 When the audience has a different take 137
6.3.1 Controlling unintended humor 138
6.3.2 Can one attack without “being mean”? 144
6.4 Conclusions 148
Contents vii

7 A dialogue on intentions 151


w i t h t e u n a . v a n d i j k a n d t . ja s o n t h r o op
7.0 Introduction 151
7.1 November 1, 2004: from van Dijk to Duranti 151
7.2 November 2, 2004: from Duranti to van Dijk 152
7.3 November 2, 2004: from van Dijk to Duranti 154
7.4 November 8, 2004: from Duranti to van Dijk 158
7.5 November 9, 2004: from van Dijk to Duranti 163
7.6 November 16, 2004: from Throop to Duranti (and van Dijk) 167
7.7 November 16, 2004: from van Dijk to Throop and Duranti 171
7.8 Conclusions 173

8 Opacity of other minds: local theories revisited 175


8.0 Introduction 175
8.1 Levels of argumentation 176
8.2 Opacity of other minds as a problem 179
8.3 Contextual variation and inconsistencies in the ethnographic accounts 179
8.4 Examining spontaneous interaction 180

9 Intentions and their modifications: a lesson from Husserl 187


9.0 Introduction 187
9.1 Husserl’s intentionality and the relationship with the modification of attention 187
9.2 Phenomenal modifications: the constitution of linguistic acts 188
9.3 Intentional modifications 189
9.3.1 Instigating intentional modifications in jazz students 192
9.4 The “natural attitude” 197
9.5 The “theoretical attitude” 198
9.5.1 The theoretical attitude in talk to and by children 201
9.6 Modifications and their role in socialization 206

10 A sense of the other: from intentionality to intersubjectivity 209


10.0 Introduction 209
10.1 Beyond individual intentions in human action 210
10.2 Searle’s notion of “collective intentionality” 212
10.2.1 The “Background” and the “Horizon” 215
10.2.2 People in the park 217
10.2.3 The paradox of collective action: more planning, less pro-social behavior 218
10.3 Socially distributed cognition and extended minds 219
10.4 Following rules 221
10.5 Spontaneity and improvisation 222
10.5.1 Naturalizing intentions 223
10.5.2 Embodied intentions: jazz improvisation 223
10.6 Husserl’s theory of perception as anticipation of the viewing of an Other 228
10.7 Our everyday understanding of others 230
10.8 Conclusions 231

11 The intentional continuum 233


11.0 Introduction 233
11.1 The meaning of intentions and their uses 233
11.1.1 Matching theory with some ethnographic and linguistic data 235
11.2 The “intentional continuum” 238
viii Contents

11.3 Hypo-cognition of intentions and hyper-cognition of perlocutionary effects 240


11.4 The performance of interdependence 241
11.5 Individuals, social coordination, and social control 242

Appendix A: Transcription conventions for English examples 243


Appendix B: Transcription conventions and abbreviations used
in the Samoan examples 245
Notes 248
References 267
Index 293
Figures

2.1 The same Subject (A) can be directed toward the same
Object (G) through two different intentional acts: looking
and worrying. page 26
3.1 Orators (on the left) and a chief (on the right) during a fono
(photo by A. Duranti, Falefā, `Upolu, Western Samoa, 1988) 49
3.2 Map of section of Anoama`a East where Falefā is located, in
the Atua district, of which Lufilufi is considered the capital and
historical center (adapted from Krämer 1902: 704–705) 52
3.3 Three types of fono in Falefā, each defined by the number
of participating subvillages (from Duranti 1981a: 40) 54
4.1 Title page of 1887 edition of the Tusi Paia, the Samoan translation
of the Bible, a collaborative effort by George Pratt, Henry Nisbet,
and others 83
5.1 Truth as correspondence between mind and world 105
5.2 Interpretation as correspondence between a proposition and
a state-of-affairs 105
5.3 Two possible relations between a proposition and the
state-of-affairs it describes 107
5.4 The same Subject (A) can be directed toward the same Object (G)
through different intentional acts, e.g., admiration, fear 108
6.1 Walter Capps speaking to a group of supporters in Paso Robles,
California, November 15, 1995, saying “How do I know that?” 139
6.2 Walter Capps speaking in San Luis Obispo, November 15, 1995,
saying “How do we know we’re gonna win?” 141
6.3 Walter Capps speaking at Hancock College, Santa Maria,
November 15, 1995, saying “I think the reason
we’re gonna win . . .” 143
6.4 Walter Capps speaking on the campus of the University of
California, Santa Barbara, November 15, 1995, saying
“this super-charged political rhetoric that . . .” 149
9.1 The same Subject (A) is directed toward the same Object (G) at
time t1 in the intentional act of admiration and at time t2 in the
intentional act of fear 191

ix
x List of figures

9.2 The rabbit-duck figure discussed in Wittgenstein (1958: 194) 191


9.3 George Bohannon (on the right) interacting with the students in
his Jazz Combo
Class on February 10, 2003, saying “Hearing what J. J. has played
on his tune” 194
9.4 While listening to the song “Cain and Abel” with his students,
on April 23, 2003, Sherman Ferguson points out that the drummer
in the recording, Jeff “Tain” Watts, is “interactive with [the other
players]” 195
9.5 While driving, Walter Capps turns slightly to the left to look
at the side of the road 200
10.1 Dr. Bobby Rodriguez, trumpet player, conductor and instructor,
demonstrates with left hand and left foot how to keep time while
playing 227
Acknowledgments

First, among my intellectual partners, comes Elinor Ochs. The few papers we
co-authored do not do justice to the many ways in which our work is inter-
twined and interdependent. Whether or not we are aware of it, we keep
thinking-with and thinking-along one another even when we do not talk about
what we are reading or writing, which is rare. Our life journey together has
always been rich in discoveries and rewarding in surprises.
On most Saturday mornings I go on a bike ride along the beach with my
friend Chuck Goodwin. We always stop at the same café in Venice and discuss
as many topics as we can fit within an hour or so, including the just observed
intentionality and intersubjectivity of bikers and pedestrians on the bike path.
I am very grateful to Chuck for his interactive listening and generous intellect.
Over the last few years I have been fortunate to teach with Jason Throop
with whom I share a love for philosophy. Our seminar on intersubjectivity has
allowed us to read or reread some of our favorite authors and discuss them with
anthropology students whose questions and comments keep us honest in our
interpretations and speculations.
My interest in creativity and collaboration, two topics that intersect with
intentionality and deserve a book of their own, has been nourished by my
interactions with Kenny Burrell, the Director of the UCLA Jazz Program and
legendary guitar master, and with all the other extraordinary musicians who
have come to make music and talk about it in the course “The Culture of Jazz
Aesthetics” that Kenny and I have taught four times since 2002.
I have also greatly benefited from sitting in two of John McCumber’s
courses at UCLA. The exposure to his deep knowledge of European continen-
tal philosophy has enriched my understanding of the roots of some of the
problems I have been writing about, some of which I discuss in this book.
The transcripts of the Samoan recordings are the product of many collabor-
ations across time and space, starting in Samoa in 1978–1979 and continuing
over the years in Samoa (in 1981, 1988, 1999, 2000) as well as in the US. Most
recently, I have relied on email to ask for help and advice from my friend Rev.
Fa`atau`oloa Mauala and his two daughters Oikoumeni and Rossana, who
have been most generous with their time and their insights. Ultimately,

xi
xii Acknowledgments

however, the Samoan utterances and their translations that you will find in this
book are my responsibility. I have done my best to be faithful to what people
said and how they have said it in the recordings. And I have tried to incorpor-
ate as much as possible of the cultural glosses that the people of Falefā or
Samoans in the US offered while listening to the recordings with me or on their
own. But any linguistic token is always open to yet another interpretation and
I am sure that even more variants from those I heard are possible in both the
transcription and the translation offered in this book. Since all of the original
data presented here are now digitized as either audio or video files, it might be
possible for others in the future to continue where I have left off in the
interpretive process.
The project on the successful 1995–1996 political campaign of Walter
Capps (1934–1997) was made possible by the collaborative support of his
extended family and members of his campaign staff. More specific acknow-
ledgments are found in Chapter 6.
I am very thankful to Rachel George, who read with care and a critical spirit
several drafts of each chapter, always asking challenging questions while
helping me make my style as accessible as possible. I also benefited tremen-
dously from comments on earlier drafts of one or more chapters by Ken Cook,
Anna Corwin, Hadi Deeb, and Dan Zahavi. Alice Mandell and Josiah Chappel
carefully inspected the original biblical texts in Hebrew and Greek respectively
to help me understand certain lexical choices made by the missionaries in
translating the Old and New Testament into Samoan. The staff at the Mitchell
Library, New South Wales, helped me identify and let me have full access to
the original letters and other manuscripts by George Pratt, Henry Nisbet, and
other missionaries who went to Samoa in the eighteenth century and wrote in
their journals or to one another about their Bible translation work.
Many other people have contributed to this book by making me aware of
issues and different perspectives. Among them I am particularly grateful to Don
Brenneis, Penny Brown, Mike Cole, Aurora Donzelli, Steve Feld, Bill Hanks,
Larry Hyman, Marco Iacoboni, Elizabeth Keating, Steve Levinson, Marcyliena
Morgan, Sherry Ortner, Mariella Pandolfi, Justin Richland, Joel Robbins, Alan
Rumsey, Manny Schegloff, Bambi Schieffelin, Bradd Shore, Teun van Dijk, and
Scott Waugh. A special recognition goes to two very original thinkers and
fearless intellectual explorers who are, together with their mother, my dearest
supporters: David Ochs Keenan and Marco Leonard Ochs Duranti.
Finally, I am grateful to the staff members in the office of the deans of the
UCLA College of Letters and Science whose help in running the division of
social sciences has made it possible for me to continue to be engaged in my
own research and writing over the last five years.
This book is dedicated to the memory of my parents, Rossana Biccheri and
Ivio Duranti, who supported me in every enterprise, including my intellectual
pursuits, always with unconditional love and visible signs of appreciation.
1 Rethinking anti-intentionalism

1.0 Introduction
In this book I revisit the anthropological critique of analytic philosophers’
theories of meaning and action based on speakers’ intentions. On the basis of
the empirical investigation of oral communication, face-to-face interaction,
and written texts, I argue that both anthropologists and analytic philosophers
overstated their case and that the salience of intentions cannot be decided once
and for all because it actually varies across cultural contexts. As we will see,
in some cases speakers avoid any kind of discourse about intentions, focusing
on the consequences of actions rather than on their alleged original goals.
But in other cases, speculation about intentions is present even in societies
where people have been said to avoid reading the mind of others. My goal
is to support an ethnographic and interactional perspective on intentions as
cognitive, emotional, and embodied dispositions always embedded in an
intersubjective world of experience. To provide such a perspective, I review
previous arguments made by linguistic anthropologists and return to some of
the fundamental concepts and claims of speech act theory as elaborated
by John Searle over the last half century. In addition to relying on the data
from three research projects – one in Samoa and two in the US – I also draw
from a number of theoretical perspectives, including Edmund Husserl’s
phenomenology, in which both intentionality and intersubjectivity play a key
role. The transcripts and written texts I analyze in some detail in the chapters to
follow will demonstrate that language – broadly defined – is a great resource
for us to understand how particular speakers conceptualize, perform, and
understand social action. Whether or not they believe or act as if intentions
matter, by using language social agents inhabit a world of others that is
constraining, empowering, and inevitable. It is our task as analysts of human
endeavors to examine which linguistic expressions make a difference in
defining actions and assigning responsibility. Sometimes our previously
conceived analytical categories provide us with a useful framework to make
sense of new information. Other times, the fit is not there. An anthropological
perspective must honor the universal without forgetting the particular.

1
2 Rethinking anti-intentionalism

This book is an attempt to apply such a principle of investigation to a


complex set of issues and a challenging set of data.

1.1 Reopening a dialogue that never took off


Thirty years ago I wrote a working paper (Duranti 1984) in which I argued that
the role of intentions in communication is overrated. Starting from a specific
case I had documented during my first fieldwork experience in (then
“Western”) Samoa, I discussed a protracted interaction during a meeting of
the village council (fono) in which speakers’ intentions did not seem to
be as important as they are portrayed to be in commonsense accounts
of interpretation as well as in well-established definitions of meaning in
analytic philosophy (see Chapter 2). The two slightly revised versions of the
1984 working paper that were later published (Duranti 1988, 1993b) contained
assertions that were interpreted as strongly anti-intentionalist by a number of
scholars (e.g., Bogen 1987; Goldman 1993; Nuyts 1993; and van Dijk, see
Chapter 7 in this book). Although I still stand by some of the points I made in
my original account of the Samoan case, in this book I return to the data
collected in Samoa (some of which I present here for the first time) and
supplement them with the data from other projects in the US to clarify my
position and extend my argument in new directions. My goal is to review some
of the main issues of a three-decades-old debate and reach for a perspective
on intentions and reading other minds that relies on anthropological and
phenomenological methods. This perspective reframes previously discussed
issues within a culturally informed theory of intentionality that includes inter-
subjectivity as a key dimension of human understanding and acting in the
world. The simultaneous adoption of what we might call, respectively, a
context-specific and a universalistic stance on intentionality, intersubjectivity,
and agency might at first appear problematic or even contradictory. I hope that
the chapters that follow will guide readers toward the opposite conclusion,
namely, that it is possible to integrate cultural contextualism (a term that has
less baggage than “cultural relativism”) with the need for some general
and generalizable notions, intentionality being one of them. The idea of
an “intentional continuum,” which I present in Chapter 11, is an attempt
to provide such an integration by acknowledging that there are variations in
levels and degrees of intentional awareness and engagement across any human
individual and collective action.
One of the reasons for me to return to the anti-intentionalism I espoused
in my earlier work is to save its more valuable aspects while redirecting its
critical roots toward a positive contribution to the study of human agency
(Duranti 2004) and intersubjectivity (Duranti 2010). I am aware of the fact
that researchers interested in human development and human cognition are
Reopening a dialogue that never took off 3

suspicious of any critical stance toward individual intentions as the basis


for interpreting and defining human action. This is because human beings
are typically distinguished from other species for being intentional subjects.
If there are issues regarding intentions in the literature on human cognition,
they usually concern different types or levels of intentionality, as has been
the case in the literature on primate behavior (e.g., Premack and Woodruff
1978; Cheney and Seyfarth 1990; Jacob 2010). In this literature, the basic
assumption is that (a) intentions are what gives meaning to actions, utterances
included, and (b) humans have more sophisticated kinds of intentions
(e.g., intentions about other people’s intentions) when compared to other
species. The bulk of this book is a discussion and refinement of the first
assumption. I will not have anything to say about the second assumption,
which deals with the differentiation of levels of intentionality across species,
even though some parts of this book might be relevant to those who
are interested in establishing or even measuring degrees of intentionality in
non-human species.
The chapters to follow show that overall my stance regarding speakers’
intentions as used in the literature on speech acts remains critical or at best
cautious. This is made particularly explicit in Chapters 2, 5, and 10. At the
same time, I believe that a theory of human action that takes cultural contexts
seriously cannot entail a complete rejection of intentionality as a human
faculty. As I will discuss in some detail in Chapter 8, when we look at verbal
interaction searching for whether speakers engage in the activity of reading the
intentions of others, we find those signs even in communities, like Samoa, that,
as discussed in Chapter 3, have been described – and in many contexts are – as
reluctant to publicly engage in introspection. The issue, as always in the
difficult domain of cross-cultural comparison, is to be able to capture subtle
differences among the many similarities due to the common human brain and
shared conditions of human adaptation. The additional challenge is one of
methods. Sociocultural anthropologists like to alternate between observing and
asking questions. Biological anthropologists and evolutionary psychologists
like to design experiments. I like to look and see (Wittgenstein 1958: §66)
in the sense that I like to record what people tell each other in all kinds
of situations and then see how their language reveals their thoughts, stances,
beliefs, feelings, and aspirations. This is a humbling method, which has taught
me to be skeptical of sharp dichotomies, including the dichotomy between
mind-readers and non-mind-readers or between those who seem focused
on what a specific Other might have intended to do and those who act on
typification (“any person in this kind of situation is likely to do, say, think,
feel x”). The chapters to follow reveal a variety of stances, strategies, and local
conceptualizations that might be difficult to summarize with a “yes” or “no” as
to whether or not intentions matter for a theory of social action as meaningful
4 Rethinking anti-intentionalism

action. To evaluate the relevance or applicability of such a variety, within and


across cultural contexts, we need to understand the origins of the objections
that a number of anthropologists raised to the ways in which intentions were
(and to some extent still are) being used in the literature on meaning and
especially in speech act theory and related approaches to speech as action.
As we shall see, some of the differences between speech act theorists’ and
anthropologists’ view of meaning have to do with the type of data that are used
to make certain theoretical claims: imagination vs. ethnography, make-up
examples vs. recorded verbal interactions, English speakers vs. speakers of
another, typically non-Indo-European language. There are also differences
in how one writes and argues. The philosopher John Searle likes and often
succeeds at practicing a style of writing that values clarity. For him this
translates into a preference for simplification over complexification and for a
professed dislike of ambiguity. Most anthropologists, on the other hand, pride
themselves on the ability to identify and represent complex, often ambiguous
domains of human interaction and are suspicious of simplification, especially
when dealing with non-western cultural traditions. If one adds the fact
that Searle usually presents stereotypical examples of human behavior,
while linguistic anthropologists tend to focus on the minute details of actual
communicative exchanges, the challenge of establishing a dialogue is hard
to overcome. In this book I try to counter what I see as a consensual
construction of incommensurability by revisiting claims, reviewing arguments,
and introducing some new characters in the story. With this overall goal
in mind, I discuss specific cases where speakers’ utterances may be interpreted
by recipients in ways that are independent of speakers’ intentions (see in
particular Chapters 5 and 6). I also argue that certain properties of verbal
communication constrain or guide speakers’ messages – in form and content –
by making it more (or less) difficult for speakers to express certain meanings
and thus perform certain social acts (see Chapters 5 and 6).
One of the possible conclusions of the analyses presented in this book is that
the reconstruction of a speaker’s state of mind might not tell us what we need
to know to understand the force of their words, that is, what a person’s
utterances accomplish in the social world (I am borrowing the concept of
“force” from J. L. Austin 1962). Another conclusion is that speakers have
contextually variable access to and authority over the illocutionary force of
their own utterances. Thus, sometimes speakers seem to fully control what
they mean or the direction of the ongoing interaction. Other times the audience
feels or seems empowered to assign particular interpretations to what someone
said or did without concern for what that person might have “meant” or
“intended.” In Chapters 5 and 6 I will argue that in some cases speakers
are led toward either expressing or accepting certain meanings because
of the particular type of communicative system or communicative devices to
Self and other 5

which they have access (e.g., the genre or the lexical choices they are expected
to use to convey their opinion). More generally, the data presented in several
of the chapters of this book suggest that there are cultural preferences for
engaging in particular interpretations of speech acts.1 I use “cultural” here to
cover both the (typically unconscious) dispositions acquired by individuals
over the course of socialization processes – what Bourdieu (1977, 1990) tried
to capture with the use of the term habitus – and the means (verbal or
otherwise) through which humans express to others and to themselves what
they are (or were) up to.

1.2 Self and other


Since the beginning of the anthropological discussion about intentions in
the 1980s, the terms of the debate have changed and so has the type of
evidence that scholars have used for supporting their claims. But the central
anthropological concern has remained the same, namely, the implications of
local theories of interpretation on any universal theory of mind or of linguistic
communication.
At first, linguistic anthropologists focused their criticism on analytic
philosophers’ models of “speech acts” (e.g., promises), understood as basic
units of human conduct. Over time, however, both the theoretical concepts and
the range of ethnographic materials brought to bear on the debate expanded.
The analysis of particular speech acts (or lack thereof) in one community
(e.g., in the Philippines, Samoa, or Morocco) in relation to local notions of
self (or person) and social action has been extended to include the effects
of social change (e.g., cultural contact through colonization or missionization)
on individuals’ and communities’ adoption or transformation of new practices
(e.g., praying, confessing, translating and interpreting foreign texts).
The theoretical debate has also been widened by new models and methods
recently introduced in a number of fields, including evolutionary anthro-
pology, developmental psychology, and neuroscience. The evidence for sub-
conscious and language-independent understanding of others’ actions or for
actions uncovered by recent research on mirror-neurons has, for example,
helped to broaden the spectrum of philosophical approaches potentially useful
for thinking about how humans make sense of each other’s actions. From
an almost exclusive concern with countering the individually based view of
intentionality and truth-value proposed by analytic philosophers, a space has
been opened in anthropology and other social sciences – across observational
and experimental approaches – for thinking about intentions as embedded in
interactively established modes of thinking, feeling, and doing. It is not by
accident, then, that the notion of “shared intentionality” (or “we-intentions”)
has acquired some popularity as a substitute for the older, less familiar, and
6 Rethinking anti-intentionalism

more nuanced notion of intersubjectivity. In returning to my own and others’


earlier arguments about the use of intentions as explanatory devices for human
behavior, I engage with Searle’s notion of “collective intentionality” at both
a theoretical and empirical level (see Chapter 10). I argue that even though
his notion of “we-intentions” is an important step in the recognition of the
intersubjective quality of human understanding, it ends up reifying distinctions
that do not quite capture how people communicate with one another.

1.3 Ethnopragmatics
Some twenty years ago I used the word “ethnopragmatics” to promote a
blending of ethnography and pragmatics for studying the ways in which
language is both constituted by and constitutive of social interaction and the
social order (Duranti 1993b, 1994). That approach was born out of my
intellectual engagement with Samoan ways of speaking, an engagement that
has continued over time in almost everything I have studied, including
improvisation, as shown by my comparison of Samoan orators and American
jazz musicians (Duranti 2008b). It should not be surprising, then, that there is
plenty of Samoan ethnopragmatics in this book, whether I look at matai (chiefs
and orators) arguing in a fono or I examine the Samoan translation of Bible
stories. The Samoan examples are important for me among other reasons
because I believe that there is no other way of doing anthropology than starting
from the anthropology of a particular place and a particular group of people
who cannot but speak to one another in particular ways, for which they are
accountable practically, morally, and aesthetically (Duranti 2004). This belief
can translate to different methods of data collection. In my case, over the last
thirty-five years I have favored the audio or audio-visual recording of what
people say to one another not just on one occasion but over some extended
period of time.2 This means that the Samoan as well as the English speakers
who are quoted in this book are anything but anonymous characters, even
when only initials of their names or pseudonyms are used. In most cases, they
are people I knew personally or interacted with on a number of occasions,
often for years or decades. It is the combination of these shared experiences
with the recordings of spontaneous interactions across all kinds of situations
that I use to make my claims. I certainly do not consider my method of inquiry
the best or the only one that others, including my students, should adopt.
At the same time, it is important to understand the differences among the
methods currently available in the study of human interaction.
Psychologists, economists, and game theorists have their clever experiments
to test their hypotheses. Philosophers have their argumentative styles filtered
through an ancient hermeneutical tradition. Ethnographers have their very
personal experience of participating in the flow of social life they are trying
Themes, issues, and intellectual connections 7

to capture in their fieldnotes. As a linguistic anthropologist, I have combined


participant-observation with the frequent use of recording devices that have
produced hundreds of hours of sounds and images amenable to repeated
listening and viewing. What was said on a particular occasion by a Samoan
chief, a candidate for the US Congress, or a jazz musician continues to live in
the (now digitized) original recordings and the (repeatedly revised) transcripts.
By no means do I believe that such transcripts are “natural” objects that anyone
can see and interpret in the same way. As I have explained elsewhere (Duranti
2006b), they are cultural artifacts and as such they are not easy to read for
someone who is not initiated into the cultural tradition where transcripts
are produced, exchanged, and appreciated. I am thus aware of the fact that
the sometimes lengthy examples of spoken or written language found in the
chapters to follow – especially when they include the Samoan text and one or
two lines of the English glosses – are hard to read through. But I hope that they
have a redeeming side, namely, the opportunity for patient readers to examine
the type of evidence I collected and relied upon.

1.4 Themes, issues, and intellectual connections


Regardless of whether it is a revised version of a previously published paper or
a new contribution to our understanding of intentions in human affairs, each
chapter is here ordered to reflect the evolution of my engagement with ideas
and data. Thus, for example, Chapter 4 is a new chapter on promising in
Samoan that comes right after the extensively revised version of a paper that
was written some thirty years earlier (Duranti 1984). The two are one after the
other because Chapter 4 provides the type of evidence that I should have
provided in 1984 but I could not, in part because linguistic anthropologists at
the time were not yet engaged with the cultural implications of the translation
work done by missionaries.
Chapter 6 is about speeches in a political campaign in the US and as such it
connects with one of the main issues discussed in Chapter 3, namely, the
limited amount of control that speakers in political arenas have on how others
interpret their actions including their speech. Chapter 7 documents the moment
in 2004 when, prompted by a question by discourse analyst Teun van Dijk,
I decided to return to write about intentions after some years spent writing
about the linguistic encoding of agency (Duranti 1994, 2004) and absorbing a
phenomenological perspective on intentionality. This approach comes to the
fore in the last part of the book where I introduce Husserl’s work and build on
some of his insights on meaning-making (Chapter 9) and intersubjectivity
(Chapter 10).
Some of the chapters are dedicated to the issue of intentions in broad
theoretical terms. The second chapter tries to recapture the original intellectual
8 Rethinking anti-intentionalism

climate that motivated my own earlier interest in speakers’ intentions


and presents the critical-historical background for the chapters to follow.
It contains a brief introduction to Searle’s model of intentionality because his
approach was the main target of the anthropological critique of how analytic
philosophers were studying language as action back in the 1980s. It also
shows that the critique was a missed opportunity for both linguistic anthro-
pologists and philosophers to have a real debate. Searle took twenty-four
years to respond to Rosaldo’s criticism; linguistic anthropologists, in
turn, did little to empirically test Searle’s view of how people talk and act –
Levinson’s (1983) reanalysis of speech acts as parts of conversational
sequences (e.g., adjacency pairs) is an exception. Anthropologists also ignored
Searle’s (1990) proposal for collective intentions, where the individual finally
meets the Other. I discuss Searle’s notion of “we-intentions” in Chapter 10
where I compare it to Husserl’s notion of intersubjectivity and question
Searle’s understanding of how people do things together.
This book is both past and future oriented. It critically reexamines one
thread of intellectual history in linguistic anthropology while showing ways
to move forward, building on insights from phenomenology and drawing
examples from recordings of actual events where utterances can be shown to
reveal more than the speakers themselves might be able to remember or
acknowledge.

1.5 Title and expectations


This book is a combination of previously published essays, which I have
revised and updated, and new essays. Even though the chapters can be read
separately and not necessarily in the sequence in which they are presented,
there is a story that runs through the entire book. It is partly a personal story –
the development of my own thoughts about intentions and mind-reading – and
partly the story of an interdisciplinary engagement between my field, linguistic
anthropology, and other intellectual traditions, starting with analytic philoso-
phy and then continuing with phenomenology and interactional perspectives
on human action.
Each chapter is about a different topic, but there is among them a
common theoretical concern: the role of reading the intentions of others
in the interpretation of their actions, utterances included. This concern, in
turn, extends to another key issue in contemporary debates about mind and
society, namely, cross-cultural differences in people’s ability or willingness to
speculate about their own intentions or the intentions of others. To address
this issue, I will use the analytical tools of my discipline, linguistic anthropol-
ogy, combined with insights from other fields, including Edmund Husserl’s
writings.
Title and expectations 9

In reviewing a few key contributions to what I am calling “the anthropology


of intentions,” I encountered a recurrent problem in contemporary academia,
namely, the separation and isolation produced by hyper-specialization. Even
though this is not the place for me to analyze the causes and consequences of
this problem, I will try to point out some of the missed opportunities for
building on the work of others and possible directions for interdisciplinary
engagement. I believe that it is important to reflect on the fact that it took over
twenty years for the anthropological critique of speech act theory to be noticed
in print by its main target, John Searle. It is also instructive that linguistic
anthropologists’ position in the 1980s on the conceptualization of social action
was not too far from what was being argued roughly at the same time by
Hubert Dreyfus, Searle’s colleague at Berkeley and harsh critic of analytic
philosophy (see Chapter 10). Not surprisingly, given his long intellectual
engagement with Heidegger’s Being and Time, Dreyfus’ criticism of the
analytic approach, especially of Searle’s treatment of intentionality, was, in
turn, a replay of Heidegger’s veiled attacks in the 1920s on the centrality of
intentionality in the type of phenomenology practiced by his mentor Edmund
Husserl (see §2.5). Precursors and parallelisms do not end there. Husserl’s
response to Heidegger’s proposal, once he carefully read Being and Time after
their personal and professional falling-out (Husserl 1997), was to label it
“anthropological” (Husserl 1981) because it was too preoccupied with the
details of human existence in an “already given world” (Husserl’s August 3,
1929 letter to Georg Minsch in Kisiel and Sheehan 2007: 397) and not enough
with its a priori foundations as revealed by transcendental phenomenology
(e.g., Husserl 1969; Bernet, Kern, and Marbach 1993). As suggested by a
number of more recent interpreters of this old debate between Heidegger and
Husserl (e.g., Moran 2000a; Zahavi 2001a), the differences between the
two philosophers might not have been as pronounced as they (and their
students) made it appear. Informed by such historical and critical analyses of
past debates, I will address the apparent incommensurability between analytic
philosophers and linguistic anthropologists by proposing contextualized
interpretations of face-to-face interactions and cross-linguistic and cross-
cultural comparison of the encoding of English terms like intention, intent,
and intending.
Finally, I want to make it clear that in this book I am not covering
everything that has ever been written about intentions by anthropologists or
philosophers. What I offer here are linked episodes of a particular story of
engagement with the issue of intentions in human interaction seen through
the lenses of anthropological perspectives that take seriously the role of
language as a human faculty and languages as the historical instantiations
of that faculty. The method is a brand of linguistic anthropology that favors the
recording of spontaneous interaction and the analysis of what people actually
10 Rethinking anti-intentionalism

said to one another on a given occasion. Like all stories of intellectual


enterprises and academic arguments, the story of the anthropological critique
of intentional readings of human action is an ensemble of observations,
realizations, discoveries, arguments, misinterpretations, doubts, and gaps.
In recounting for contemporary readers different parts of the story, I have also
become aware of the forgetting that took place at different points of the
discussion and of the missed opportunities for a real dialogue. The story that
will emerge from the chapters to follow is a retelling and as such it is
a reframing of an issue – or series of issues – from the point of view of
the teller. Any question, doubt, or critique of my telling would, in turn, be
welcome, like all signs of recognition.
2 Intentions in speaking and acting: the Standard
Theory and its foes

2.0 Introduction
The anthropological critique of the use of intentions for explaining human
action focuses on a small group of authors, who helped establish what could
be considered the “Standard Theory” of interpretation in linguistics, the
philosophy of language, and the cognitive sciences in the twentieth century.
In this chapter, I review the Standard Theory, with special attention to
Searle’s use of intentions in his model of linguistic as well as non-linguistic
acts, Rosaldo’s criticism, and other authors’ earlier accounts and criticism of
intentionality as a key feature of meaning. I also return to the question of the
universality of the notion of intention by reviewing the history of the term
and then engaging in a cross-linguistic analysis that will reveal hidden
aspects of the semantic field covered in English by intention, intent, and
intending.

2.1 The Standard Theory: Grice, Austin, and Searle


The three major authors associated with the standard theory are: H. P. Grice,
for whom it is the reliance on intentions that makes linguistic meanings
different from all other kinds of phenomena to which an interpretation can
be given; John Austin, who included intentions in the felicity conditions
for speech acts; and John Searle, who made intentions a central component
of his speech act theory, his theory of mind, and, in combination with the
notion of “we-intention” (see Chapter 10), his theory of society.
Like all theories, what I call here the “Standard Theory” is anything but
monolithic. There are differences across authors, periods, and fields where it
is employed. But for the purpose of our discussion, I will treat it as a largely
shared view of intentions among many formal linguists, analytic philoso-
phers,1 and cognitive scientists. One justification for such a unified treatment
is that the anthropologists who criticized speech act theory as being too
“western” have not been so much concerned with the details of the model or
of its formalization as much as with the assumption that the perspective on the

11
12 Intentions in speaking and acting

individual subject (Self, or Ego) developed in (late) modernity could provide


the grounds for a universal model of human action.2
The Standard Theory claims that there is a particular and common type of
“meaning” in human actions that is realized through conventional “signs”
(with linguistic utterances being the most commonly discussed) that must
be distinguished from non-conventional “signs” (e.g., those commonly based
on recurrent associations). For example, the meaning of the utterance “Fire!”
shouted by a person is a sign that is different from the meaning of some
smoke going up in the sky, which could also be interpreted as a “sign of fire.”
Grice (1957) calls the first type of meaning (by a person shouting “Fire!”)
“non-natural” and the second type (the smoke) “natural.” “Non-natural” corres-
ponds to what other authors have called “conventional” or “arbitrary” even
though the latter two terms are not synonymous. For Grice non-natural meaning
has two properties. The first is that the person who produces the message
intends to make the hearer/addressee/recipient knowledgeable about something
(in Grice’s terms “to induce . . . a belief in an audience”). The second property
is that the speaker wants the hearer (or addressee, etc.) to recognize that the
message was intended to have the particular meaning that the speaker is trying
to convey. Here is the definition of non-natural meaning found in Grice (1957):
[F]or A to mean something by x . . . A must intend to induce by x a belief in an audience,
and he must also intend his utterance to be recognized as so intended. (1957: 441)
Grice’s definition is sufficiently general to be used to account for any attempt to
communicate, irrespective of the sharing of a code, whether English or American
Sign Language. Even in a case in which a tourist might address a policeman
in a language that the policeman does not know, the intention to communicate
would make the act an example of the communication of non-natural meaning.3
Grice’s focus on the intending to induce a belief that x by saying x reproduces
a standard concern of the time in analytic philosophy with assertions or state-
ments (which are about beliefs). This would not suffice for philosophers like
Austin and Searle who were conceiving of speaking as performing all kinds
of acts besides asserting or informing. But the concern with the recognition of
the speaker’s intentions remained a key feature of speech act theory, even though
some amendments were proposed by authors like Peter Strawson (1964) and
Searle (1965), both of whom were concerned, albeit for different reasons and in
different ways, with the relative weight of intentions as opposed to conventions.4

2.1.1 The speech act of promising


In the earlier formulations of speech act theory, “intention” or the act
of “intending” is one of the felicity conditions (for Austin) or conditions
of satisfaction (for Searle). Thus, when Austin (1962) speaks of the act of
The Standard Theory 13

promising, with “I promise to . . .” being one of the first examples he gives


of an explicit “performative,” he points out that
In the particular case of promising, as with many other performatives, it is appropriate
that the person uttering the promise should have a certain intention, viz. here to keep
his word: and perhaps of all concomitants this looks the most suitable to be that which
“I promise” does describe or record. (Austin 1962: 11)

Austin makes this point in order to clarify that, even though we might
informally speak of a “false” promise, being a performative, “I promise
to . . .” cannot be defined in terms of truth or falsehood – at most with “bad
faith.” Instead, “I promise” entails “I ought” (1962: 51) and as such it is part of
a whole set of acts, “commissives,” which “commit the speaker to a certain
course of action” (Austin 1962: 156).
Searle ([1965] 1971) combines Austin’s informal account of the act of
promising with Grice’s definition of non-natural meaning. In so doing, Searle
makes intentions a key aspect of promising (and, more generally, of speaking),
as shown in conditions 6, 7, and 8. This is shown in (1) below, which
reproduces the main points of Searle’s description of the conditions of
satisfaction for a promise.
(1) Given that a speaker S utters a sentence T in the presence of a hearer H,
then, in the utterance of T, S sincerely (and non-defectively) promises that
p to H if and only if:
1. Normal input and output conditions obtain.
2. S expresses that p in the utterance of T.
3. In expressing that p, S predicates a future act A of S.
4. H would prefer S’s doing A to his not doing A, and S believes H would
prefer his doing A to his not doing A.
5. It is not obvious to both S and H that S will do A in the normal course of
events.
6. S intends to do A. [. . .] (6 * S intends that the utterance of T will make
him responsible for intending to do A) [. . .]
7. S intends that the utterance of T will place him under an obligation to
do A.
8. S intends that the utterance of T will produce in H a belief that conditions
(6) and (7) obtain by means of the recognition of the intention to produce
that belief, and he intends this recognition to be achieved by means of
the recognition of the sentence as one conventionally used to produce
such beliefs. (Searle 1971: 48–51, passim)

In the 1965 article, Searle expands on each of these conditions, for example
clarifying that by “normal input and output conditions” he is referring to the
fact that the speaker must be intelligible and the hearer must be able to
understand what is being said (Searle 1971: 48). But he also avoids discussion
of the vast literature on promises since Aristotle and all the way to Hume and
14 Intentions in speaking and acting

beyond. Thus, there is no mention of the fact that the speech act of promising
does not bind unconditionally or absolutely (Scanlon 1990: 214) and it might
be used for acts that cannot be promises, such as I promise not to bore you
with my talk, which Raymond Hickey (1986) interprets as a wish turned
into an “optative” (for the relevance of these issues in political speeches,
see Hill 2000; Duranti 2007: chapter 7). In contemporary usage, the verb
promise can also refer to an act in the past, like when speakers use it to
vow for the truth of something they just said (A: I have no time on Tuesday.
B: Really? A: I promise).
For Searle, the act of promising is achieved by an act of recognition of
the intention (condition 8) but it is also an act that entails responsibility
(see condition 6*) and obligation (see condition 7) – whether contractual
and/or moral (as we shall see below, in Searle’s 1983 theory of intentionality,
responsibility is explicitly separated from intentionality).
The importance of intentions and their relevance to any meaningful act grew
over time. In the earlier classification of speech acts (Searle 1976: 4), intention
is identified as one particular type of “psychological state” shared by
commissives like “promises, vows, threats and pledges.” As such, intention
is contrasted with desire, that covers “requests, orders, commands, askings,
prayers, pleadings, beggings and entreaties” and pleasure that is shared by
congratulations, felicitations, and welcomes (Searle 1976: 4). As we shall see,
later, when writing about intentionality for all kinds of acts, not just the ones
realized through speech, Searle uses a more general notion of intention, which
he borrows from Franz Brentano and Edmund Husserl, without however
engaging in a discussion of these authors’ specific points or contributions
(see §2.4). This lack of intellectual engagement with previous treatments
of the subject he is discussing is not unusual for Searle and is particularly
conspicuous in the case of the phenomenological tradition (see Chapter 10).

2.1.2 Searle’s five types of speech acts and the notion of “direction of fit”
One key element of the speech act model that Searle exported to his model of
intentionality is “direction of fit,” an expression that had been used by Austin
but Searle modifies to capture the relative weight that language or the
(real or imaginary) world have on each other. To understand how this works
in Searle’s model of speech acts first and later in his general model of
intentionality, we need to start from the fact that for him there are five and
only five possible types of speech acts:
1. Representatives (or assertives)
2. Directives
3. Commissives
The Standard Theory 15

4. Expressives, and
5. Declarations.
The concept of the direction of fit further distinguishes among these five types
(Searle 1976). Representatives, which include descriptions (it’s raining) as
well as acts like accusations and complaints, are speech acts that have a word-
to-world direction of fit, that is, the “words” or linguistic expressions describe
an already existing state of affairs (Searle 1976: 3–4). For example, for Searle
when I say it’s raining, I am using English to inform you of something that is
supposed to be independent of my will, desire, or imagination. On the other
hand, utterances with the illocutionary force of requests, commands, vows, and
promises have a world-to-word direction of fit. In these cases I am trying to
make the world adapt (that is, change) to the kind of world that my words
describe. Thus, in requesting students to write a paper on a given subject, I am
using language to make other people do something they might not have
otherwise done. The “world,” that is, theirs and mine, will change as a result
of my linguistic act of requesting a term paper.
For Searle, there are also speech acts that have a double direction of fit and
other ones that have no direction of fit. Belonging to the first group are
so-called declarations, that is, acts that make the world adapt to the description
of a state of affairs. “The performance of a declaration brings about a fit by the
very fact of its successful performance” (Searle 1976: 14). Acts that have no
direction of fit are the so-called “expressives,” like apologies and excuses.
In performing an expressive, the speaker is neither trying to get the world to match
the words nor the words to match the world, rather the truth of the expressed prop-
osition is presupposed. Thus, for example, when I apologize for having stepped on your
toe, it is not my purpose either to claim that your toe was stepped on or to get it
stepped on. (Searle 1976: 12)
The direction of fit has some potential relevance to an anthropological theory
of language because it deals with whether language is trying to catch up with
the way the world is or it tries to change the world. As such, the concept of
the direction of fit is a potential link with linguistic relativity, understood as the
influence that the lexicon or the grammar of a language has on the ways in
which its speakers view the world (Whorf 1956; Lucy 1992a, 1992b). A main
difference between Searle’s model and a model inspired by Whorf’s writings is
that for Searle only certain types of speech acts do have the effect of changing
the world (e.g., commands, vows) whereas for Whorf (and most linguistic
anthropologists) the speakers’ view as well as their surrounding social world
can be affected by linguistic taxonomies and the use of certain expressions in
all kinds of linguistic acts, including assertions and expressives. Thus, for
example, Stephen Levinson and his research team at the Max Planck Institute
for Psycholinguistics looked for and found some (Whorfian) effects on how
16 Intentions in speaking and acting

languages represent space and spatial relations (e.g., Levinson 1997, 2003).
One could also easily argue against the idea that “expressives” (see above)
have no direction of fit. The example of apologies Searle gives in the quote
above is misleading. When I apologize for having stepped on your toes I am
actually trying to affect your feelings or disposition toward me as well as
toward any discomfort you might have in your foot. I am reassuring you that it
was not an act of aggression on my part and that I might be blamed for
being distracted or clumsy but I should not be blamed for being a mean-
spirited person.
Searle (1983) used the concept of direction of fit to distinguish between
intentional states like believing, seeing, and listening, which have the mind-to-
world direction of fit, and intentional states like wishes, desires, or “intention
to do something” (e.g., raising my arm, going to the garage, driving to school),
which have the world-to-mind direction of fit. This dichotomy means that for
Searle either the world out there causes something in me (like when I perceive
something) or I cause something to happen in the world (in the case of actions
that I perform or that I ask someone else to perform).5

2.1.3 Searle’s model of intentionality


Searle (1983) conceptualizes intentionality in terms of four components:
1. priori intention (“the whole action,” e.g., raising my arm)
2. intention(s) in action (e.g., getting the arm to start moving up in a
particular way)
3. bodily movement (e.g., the arm moves in particular ways)
4. the action (e.g., my arm goes up).
Causation provides the link between intentions and action.6 Thus, the prior
intention causes “the whole action,” which is typically the combination of
one or more intentions in action, which, in turn, cause (in the example above)
the bodily movement(s). The “intention in action” seems similar to what
some cognitive psychologists call “subgoal,” even though the latter might be
more explicitly connected with situational cues, especially in the psychological
literature on self-regulation. For Searle, intentions in action seem to include
all kinds of small or shorter actions (or activities) that are necessary in order to
perform the whole task as defined by “prior intentions.” For example, in order
to go to the office to work, we need to get dressed, get in the car, turn on
the motor, and drive to the office; to eat lunch while at work we might need
to go to the cafeteria and order food.7
Unlike other philosophers and cognitive scientists who include plans
and planning in their model of the relationship between intentions and
actions (e.g. Bratman 1980, 1999; Tuomela 1984; Grosz and Sidner 1990;
The Standard Theory 17

Hobbs 1990), Searle (1983) did not use the notion of plan, but his distinction
between “prior intentions” and “intentions in action” (Searle 1983: 84–85)
could be interpreted as a way of dealing with different levels of planning. To
the extent to which prior intentions are conscious and tend to cover a “whole
action,” they seem to imply “a plan” – even a vague one – to do something,
whereas the intentions-in-action, which are used by Searle for unconscious
and often subsidiary acts, imply more spontaneous ways of acting.8
Most analysts of spontaneous human interactions would find Searle’s
examples not only sketchy and stereotypical but also empirically inadequate
as accounts of what people actually do. Take the “action” of going from home
to the office. This would be defined by a (general) prior intention – that covers
the whole action – and by a series of intentions in actions. Some of these
intentions in action could be performed automatically and unreflectively, but
some other ones might be performed with great care, that is, while paying
close attention to what is being done and how (such a possible distinction, as
far I can tell, is not included in the model). For example, the intention in action
of getting the car out of the garage – which would be part of getting to the road
in order to drive to the office – might itself have an embedded intention in
action to drive in such a way as to avoid hitting anyone or anything that might
happen to be on the side of the car or behind it. The way in which the
neighbors or their guests parked their car in the street or the traffic of the other
moving cars on the street in front of our house will also affect how we perform
the intention in action of driving out of our garage and onto the street. It would
seem that at any given time something that is a sub-activity of a higher goal
(or “prior action”) might become a major activity in its own right, worthy
of attention. Furthermore, how does Searle’s model allow for unforeseen
circumstances generating new intentions in action? How do we know whether
such intentions in action are unconscious or not? Drawing on Heidegger’s
famous discussion of “equipment” (Zeug) in Being and Time, we could say
that any tool we routinely use in a habitual manner can suddenly come to the
forefront of our concern when it malfunctions or when it is missing (Heidegger
1962: 95–107). As he expanded his theory to include any kind of human
action regardless of whether language was used, Searle (1983) tried to account
for this type of implicit or embodied knowledge through his notions of
“Network” and “Background.”

2.1.4 Searle’s notions of Network and Background:


capturing implicit cultural knowledge
Both the “Background” and the “Network” – as in “a Network of Intentional
states” – were introduced by Searle in his book on intentionality to account
for all kinds of “capacities and social processes” (1983: 147) that are necessary
18 Intentions in speaking and acting

for human action to occur at all. Searle gives the example of someone who has
“the intention to run for President of the United States.” He points out that
this “intention” is related (or “refers”) to other intentions, which, in turn, have
other conditions of satisfaction, each with its own set of beliefs and desires.
This means that “in any real life situation, the beliefs and desires [associated
with one particular intentional state] are only part of a larger complex of still
other psychological states; there will be subsidiary intentions as well as hopes
and fears, anxieties and anticipations, feelings of frustration and satisfaction”
(Searle 1983: 141). This is what he calls the “Network of other beliefs and
desires” (1983: 141). Similarly there is knowledge that we have of our world
that is similar to and yet different from “belief” in the sense in which the
word is used by Searle and most analytic philosophers. This knowledge is
necessary and yet not explicit. To characterize it, Searle (1983: 143) uses the
term “preintentional.” The totality of this type of information (which is both
“knowing-that” and “knowing-how”) constitutes the “Background,” a domain
for which rules, in any traditional sense, cannot be evoked to explain how
people know what they know.
Echoing descriptions of expertise found in Hubert and Stuart Dreyfus
(1980),9 Searle gives the example of learning how to ski. Even though the
instructor gives a set of explicit instructions, which Searle considers examples
of “representations” (see endnote 13), over time, “the skier gets better; he no
longer needs to remind himself of the instructions, he just goes out and skis”
(Searle 1983: 150). The mentioning of Heidegger’s notion of “equipment”
a few pages later (1983: 154) would suggest that this might be a place where
Searle is reaching out for a different type of explanation, one in which practice
wins over mental “representations.” After all, Dreyfus and Dreyfus’ (1980)
model of expertise is informed by Heidegger’s view of tools (in Being and
Time) as having the primary property of being “ready-at-hand” (that is, entities
whose existence is primarily defined by their use) instead of “present-at-hand”
(that is, entities that are just there for us to reflect on them without engaging
with them in the ways in which we engage with tools or with other human
beings). In Being and Time, humans are described as existing in a different
way from the ways in which things and tools exist. Humans are first and
foremost thrown into a world that occupies them with moods and concerns
before they can have “representations” or “essences.” We are first “there”
(“da”) and then we think and reflect about what we are doing and which
categories can describe who we are and what we are dealing with. Such
a pragmatic engagement is explicitly evoked by Searle with an unusual
reference to tools understood in “a Heideggerian vein” (1983: 154), that is,
as “the referential totality of ready-to-hand equipment” (1983: 154).10 For a
few pages Searle’s unique blend of mentalism and analytic philosophy seems
to let the voices of alternative philosophies leak through. But it is only
The role of intention in defining “action” 19

a fleeting moment.11 Order is quickly restored, at least until 1990, when,


as we shall see in Chapter 10, Searle introduces “collective intentionality.”

2.2 The role of intention in defining “action”


As we shall see later, anthropologists did not appreciate a number of Searle’s
assumptions and concepts. One was the exclusive use of individual intentions
to characterize social action. Another was the notion of the autonomous self
that seemed to be implied by the model. A third problem was the treatment of
linguistic communication as caused by mental states. Searle’s perspective was
criticized for not acknowledging the ambiguous, dialogical, intertextual quality
of human discourse and, more generally, the ubiquitous indexical value of
language as a situated and situating human activity that might not be controlled
or controllable by the individual (Duranti 1984, 1993a, 1993b; Vance 1995).
As we shall see in Chapter 10, Searle partly moderated his reliance
on individual intentions through the later introduction of the concept of
“collective intentionality” (Searle 1990), but anthropologists, myself included,
have continued to refer only to his earlier work (for a recent exception, see
Harr 2013). I will try to remedy this lacuna in Chapter 10.
The most obvious problem with Searle’s 1983 model from an anthropo-
logical perspective is his thesis that there is human action if and only if what
occurred matches the conditions of satisfaction as defined by the intentions
of the person who performed the action.
There are many states of affairs without corresponding beliefs and many states of
affairs without corresponding desires, but there are in general no actions without
corresponding intentions. (Searle 1983: 82)

According to this view of what constitutes “action,” falling down by acciden-


tally tripping on a banana peel is not an “action.” But ordinary language – or at
least ordinary English – often does not distinguish between an action that a
person meant to do and an action that happened by accident. Thus, for
example, the English utterance he smashed down a fence could describe an
action that a driver did on purpose or something that happened after he had
lost consciousness due to a heart attack.12 The fact that the same utterance, he
smashed down a fence, could describe both was sufficient for Searle to reject
descriptions altogether as a criterion for the interpretation of action. To illus-
trate this point, in his book Intentionality Searle uses the plot of Sophocles’
tragedy Oedipus Rex. Since Oedipus intended to marry Jocasta and did not
know that she was his mother, for Searle the proposition “Oedipus married his
mother” is true but problematic (or misleading) as a description of the meaning
of the event because it might be understood as saying that Oedipus intended
to marry his mother (we are here in the domain of what logicians call “opaque
20 Intentions in speaking and acting

reference”). Thus, for Searle, Oedipus married his mother does not describe
“the action” because for him “action,” as we saw, is what the actor intended to
do (Searle 1983: 101–102), and we know that Oedipus could not intend
to marry his mother when he married Jocasta because he did not know that
she was his mother. This is the same as saying that an act or series of acts
counts as an “action,” in Searle’s sense of the term, if and only if it turns out
to be performed as it was “represented” in the intention(s) of the person who
performed it.13
Since Searle’s model is not set up to explain events that are not included
(or “represented”) in the mind of the social actor in terms of “prior intentions,”
it is not surprising that Searle wants to keep intention and responsibility quite
distinct and independent (Searle 1983: 103), even though responsibility did
play a role in his earlier definition of the speech act of promising, as shown in
(1) above, and he seemed to accept it as equivalent to a speaker’s commitment
(e.g., Searle 1991: 100, in response to Alston 1991). This means that Searle’s
model of action as developed in his 1983 book does not cover “unintended”
actions, a set of circumstances that, as discussed below, are instead of great
interest to social scientists. In the case of Oedipus Rex, for example, the model
would not have much to say about the fact that the entire city of Thebes
must suffer because of what Oedipus did even though, as Oedipus himself
says in Oedipus at Colonus, scene IV, he did not know that he was marrying
his mother (and she did not know that he was her son). Similarly, albeit in
less tragic ways, Searle’s theory does not have much to say about a case
I discuss in Chapter 3 where a Samoan orator is blamed for the unintended
consequences of having announced a visit by the newly reelected MP that
never happened – a visit that was anticipated as including gifts for the
members of the village council.

2.2.1 Unintended consequences and the practical


engagement with the world
The exclusion of unintended consequences of individuals’ actions is a problem
for many social theorists, from Karl Marx’s characterization of the social
conditions for alienation and exploitation under capitalism all the way to
various versions of what has been named “practice theory” (Ortner 1984).
In the case of Marx, individuals might either not understand or not control
their intentions due to false consciousness and to the material, i.e., economic,
conditions of their existence.
Theorists like Bourdieu and Giddens extended Marx’s perspective to
include an appreciation of the temporal and spatial situatedness of everyday
existence that draws from a number of philosophical sources. Such sources
included Heidegger’s hermeneutics of the human condition or Dasein in Being
The role of intention in defining “action” 21

and Time (e.g., Giddens 1979: 54).14 As shown in the following quote, for
Giddens “intentionality” does not imply that people are conscious of what
we might call the telos of their actions.
The reflexive monitoring of conduct refers to the intentional or purposive character of
human behaviour: it emphasises “intentionality” as a process. Such intentionality is a
routine feature of human conduct, and does not imply that actors have definite goals
consciously held in mind during the course of their activities. That the latter is unusual,
in fact, is indicated in ordinary English usage by the distinction between meaning or
intending to do something, and doing something “purposefully,” the latter implying an
uncommon degree of mental application given to the pursuit of an aim. When lay actors
inquire about each other’s intentions in respect of particular acts, they abstract from a
continuing process of routine monitoring whereby they relate their activity to one
another and to the object-world. (Giddens 1979: 56)

In this perspective, the interpretation of each other’s intentions – which is, as we


saw above, a condition for Grice to have non-natural meaning and for Searle to be
able to know whether what happened was meant – is produced after the fact and
as such it is considered by most social scientists a rationalization of actions that
took place within a series of socially constrained and social meaningful events. As
already pointed out by Husserl in the first volume of Ideas (Husserl 1931, 1982:
§§27–28), in the “natural attitude,” that is, the taken-for-granted stance of every-
day life (which we could easily call “the cultural attitude”), we know and act in a
non-conceptual way. Our practical engagement means that most of the time we act
as citizens, spouses, parents, children, friends, employers, employees, coworkers,
team players, band members, opponents, passengers, drivers, or strangers, etc.,
without having to think that we are doing so, even though others might see us as
embodying those roles and might react accordingly. Husserl stated very clearly
that this everyday way of being in the world precedes any theoretical engagement
with the world – Heidegger radicalized this point in Being and Time.
As I will show in Chapter 9, such a theoretical attitude is typically realized by
adults as well as children through the faculty of language. From an early age,
language gives us the ability to have a “theoretical stance” toward our experi-
ences as well as toward the institutions within which we operate. But does
reflexivity give us control of our own intentions and, thus, of the meaning of
actions? Not for authors like Bourdieu. The fact that individuals can “stop and
think” about their actions does not mean that they have control of their destiny.
There is in this position a declared affinity with Marx’s view of the independ-
ence of “objective” social relations from individual consciousness (Bourdieu
and Wacquant 1992: 97) and an echo of the famous line in The German
Ideology, where Marx wrote: “Life is not determined by consciousness, but
consciousness by life” (Tucker 1978: 155). In the realm of artistic production,
this means that “we must . . . ask, not how a writer comes to be what he is, in a
sort of genetic psycho-sociology, but rather how the position or ‘post’ he
22 Intentions in speaking and acting

occupies – that of a writer of a particular type – became constituted” (Bourdieu


1993: 162). Bourdieu’s answer will be: through a double perspective – that of a
“field,” that is, the context of a given occupation, and that of the habitus, that is,
a set of dispositions acquired through socialization that make it possible or
impossible to participate in a given field (Bourdieu 1977). The habitus carries
us through our daily life and makes us feel we are inventing a life that is in
many respects limited by history and external forces.
The individual, the subjective, is social and collective. The habitus is socialized
subjectivity, a historic transcendental, whose schemes of perception and appreciation
(systems of preferences, tastes, etc.) are the product of collective and individual history.
Reason (or rationality) is “bounded” not only . . . because the human mind is generically
bounded . . . but because it is socially structured and determined, and, as a consequence,
limited. (Bourdieu 2005: 211)

In this perspective, even when acknowledged, individual intentions are not


directly relevant to what is ultimately produced or achieved. An author could
write a book with the goal of showing the need to radically change or even
abolish a given field of study, but that action might in fact contribute to the
reproduction of that field of study and make it even stronger.
Some anthropologists took a similar stance toward the conceptualization of
social action in speech act theory, but they used cross-cultural comparison
as their instrument of criticism. Intentionalist models of human action as the
product of (supposedly) autonomous selves were defined as too “western” and
as such unable to make sense of non-western ways of acting and interpreting
(e.g., Rosaldo 1982; Moermon 1988; Rosen 1985, and the essays in Rosen 1995).
As I will discuss in Chapter 8, some of the more recent investigations of the
cultural dimensions of “theory of mind” continue in the same critical vein (but see
Danziger 2006 and the articles in Danziger and Rumsey 2013 for more nuanced
positions). In the rest of this chapter, I will limit the discussion to the relevance
of speakers’ intentions in the interpretation of language as a social activity.
Do intentions explain the force that utterances have in social life? Among
the linguistic anthropologists who explored this question Michelle Rosaldo
was the first who most directly engaged with Austin’s and Searle’s earlier
writings, claiming not only that speech act theory could not explain the ways in
which the Ilongots in the Philippines used language, but that the implicit notion
of self or social actor was not even adequate for English-speaking communities.

2.3 Michelle Rosaldo’s critique of speech act theory


One of the strongest attacks on Searle’s earlier model of speech acts was
presented by Michelle Rosaldo in a posthumously published article, where
she criticized Searle for analyzing speech acts like promising in ways that lead
Rosaldo’s critique of speech act theory 23

us to think of meaning as originating from inner experience instead of


the public sphere where the community interacts (1982: 211). Deploying
anthropologists’ classic trope of doubting the universality of western practices,
Rosaldo used her ethnography among the Ilongots in the Philippines to
criticize Searle for generalizing from his own limited worldview and for not
paying attention to the cultural context of the speech act:

In brief, by generalizing culturally particular views of human acts, intentions, and


beliefs, Searle fails to recognize the ways that local practices give shape at once to
human actions and their meanings. Ignoring context, he discovers that linguistic action
can be classified in universal (and essentially subjective) terms, but in so doing he
projects misleading patterns on our categories of speech. (Rosaldo 1982: 212)

Rosaldo saw Searle as presenting speakers and hearers as “autonomous selves”


whose actions were based on individual choice and individual intentions and
not enough on social expectations “constrained by the relationships and
expectations that define their local world” (Rosaldo 1982: 204). More gener-
ally, Rosaldo’s argument was an attack on the epistemological and ontological
foundations of analytic philosophy as a theoretical approach that posits the
relevance of an “inner self” for explaining behavior. Her underplaying of
intentions and sincerity could be read as a behaviorist argument against the
relevance of inner disposition, including individual will and intentions. But the
Ilongots favor cooperation, sharing of goods, and an ethos that does not
conform to the practices of the type of society presupposed and championed
by Hume in A Treatise of Human Nature, Book III, where promises are
practices needed to keep the social system running. Where Hume sees reason
as the engine of the social order and doubts that equality could ever exist or last
(Miller 1976: 168–169), Rosaldo championed passion as the social engine of
Ilongot actions (Rosaldo 1980). She claimed that the Ilongots could function
fine as a society without promises. She also explicitly rejected an interpretation
of her critique as “behaviorist” (Rosaldo 1982: 227) and stressed the theoret-
ical importance of the “force” of language as something that shapes individ-
uals’ motives and activities and as such participates in the constitution of
society and the ways in which people (philosophers included) think about
social systems and local forms of collaboration and production (Rosaldo
1982: 228).
Reread today, Rosaldo’s argument – like so much of what we write in
academia – appears as largely directed to the converted. In contrast to an
earlier attempt by Ruth Finnegan (1969) to mend and yet build on Austin’s
theory of speech as action, Rosaldo did not seem interested in trying to extract
something useful for anthropologists out of the analytic philosophers’ specu-
lations about how people think, act, and feel. A similar lack of interest in a
dialogue is found in the delayed response by Searle who, as we shall see,
24 Intentions in speaking and acting

completely missed Rosaldo’s main points, namely, that it was wrong to build a
theory about how utterances can be (or become) social acts solely or almost
exclusively on speakers’ sincerity and intentions. Social life and language,
Rosaldo argued (and tried to show through some examples of Ilongot speech
acts and their interpretations), build off each other in ways that draw from and
simultaneously reconstitute the social order, people’s views of it, and local
speaking styles. Whether or not a given utterance would count as a promise
would depend on the social relations between the parties involved. For certain
people promises might be ill conceived or inappropriate. Here is the gist of her
argument in her own words:
What Searle forgets, and yet to me seems clear, is that the good intentions that a promise
brings are things we only offer certain kinds of people, and at certain times. Introspec-
tion suggests, for example, that promises to one’s child are typically didactic and
tendentious. A promise to, or from, a candidate for public office is apt to prove neither
sincere nor insincere but in equal measure suspicious, significant, and grand. Sincere
promises to my colleagues are typically no more than that: sincere commitments. To a
high administrator, my promises may seem peculiar. And I cannot escape a sense of
awkwardness in imagining a promise to my spouse. There are, in short, quite complex
social “rules” that circumscribe the happy “promise” – although our ideology of
promising leads one to focus not on these but on the “‘inner” orientations and commit-
ments of the speaking self. Moreover, it would appear that Searle, by focusing on the
promise as a paradigmatic act of speech, himself falls victim to folk views that locate
social meaning first in private persons – and slight the sense of situational constraint
(who promises to whom, and where, and how) that operates in subtle but important
ways in promising, and in yet more salient ways in the case of a directive, like
“commanding,” or such apparently expressive acts as “congratulating,” “greeting,”
and “bidding farewell.” The centrality of promising supports a theory where conditions
on the happiness of a speech act look primarily not to context, but to beliefs and
attitudes pertaining to the speaker’s private self. (Rosaldo 1982: 211–212)
Unfortunately, instead of realizing – or admitting – that Rosaldo’s argument
was also an attack on his model of the relationship between person and
context, Searle interpreted it exclusively as a counterexample to disprove the
universality of promising as a speech act.
When I published a taxonomy of the five basic types of speech acts (Searle, 1979 [same
as Searle 1976]), one anthropologist (Rosaldo, 1982) objected that in the tribe that she
studied, they did not make very many promises, and, anyway, how did I think I could
get away with making such a general claim on the basis of such limited data? But the
answer, of course, is that I was not offering a general empirical hypothesis but a
conceptual analysis. These are the possible types of speech acts given to us by the
nature of human language. The fact that some tribe does not have the institution of
promising is no more relevant than the fact that there are no tigers at the South Pole is
relevant to a taxonomy of animal types. I am discussing the logical structure of
language and getting the categorization of possible types of speech acts. (Searle
2006: 26–27)
Brentano’s and Husserl’s use of intentions 25

This is “crosstalk,” to use the term made famous by the homonymous BBC
documentary by John Gumperz about cross-cultural miscommunication.
Searle completely missed Rosaldo’s theoretical objections just like she had
implicitly given up on modifying Searle’s formal model of communication.
The lack of attention to the cultural context made any real dialogue impossible
for Rosaldo as well as for the rest of us who were considering similar issues.
There is no question that Searle’s speech act theory was light on context, as
shown in (1) above by his use of expressions like “normal input and output
conditions” (condition 1) and “the normal course of events” (condition 5).
What is normal in one situation is not normal in another. What is normal in a
given situation in one community might not be normal in another community.
Social scientists could have accepted the limitation of Searle’s vague descrip-
tions and work to “fill in the blanks.” But that is not the way it went. Rosaldo
and others, myself included, opted for holding Searle accountable for what we
considered at the time a “western” psychologizing of Austin’s original insights
about what makes a speech act work or misfire15 (see also §10.2.2). The focus
on speakers having to be sincere and on the need to have certain intentions,
which, in turn, had to be recognized by the hearers, struck many of us at the
time as too inward oriented as a theory of social action, placing the responsi-
bility for both encoding and decoding on the speakers having certain beliefs,
feelings, and goals, as well as the will and ability to express those beliefs and
feelings.
At the time we did not realize that we were in part reproducing a much
earlier confrontation about the use of intentions in the definition of meaningful
human experience. It was a confrontation between two major protagonists
of twentieth-century German philosophy, Edmund Husserl and Martin
Heidegger.16

2.4 Brentano’s and Husserl’s use of intentions


There are two main notions of intention in the philosophical literature, and
Searle utilizes them both. One coincides with the folk psychological sense
of “intention” and the common use of the English noun intention and the
English verb intend as it was used in speech act theory (see above). The other
is the broader notion of intentionality as “directedness,” which draws from
the phenomenological tradition. Searle (1983) distinguishes between the two
by writing “intention” and “Intention” (and “Intentionality”) respectively.
He clarifies that “intentions in the ordinary sense” are “just one form of
Intentionality along with belief, hope, fear, desire, and lots of others . . .”
(Searle 1983: 3).
The broader and more foundational sense of intention (“Intention” for
Searle) comes from the Scholastic tradition (see §2.7) and was reintroduced
26 Intentions in speaking and acting

Subject A Object G

Figure 2.1 The same Subject (A) can be directed toward the same Object
(G) through two different intentional acts: looking and worrying.

in European philosophy by Franz Brentano (1874) and adopted by Edmund


Husserl, who in 1884–1886 had attended Brentano’s lectures in Vienna.17
Husserl refined Brentano’s use of intentionality (Mohanty 2008: 47–50) and
made it a basic concept of his theory from Logical Investigations (Husserl
1900, 1901, English translation in Husserl 1970b) all the way to The Crisis
of the European Sciences, the book he was trying to complete toward the end
of his life (Husserl 1970a). Intention is the aboutness of our interpretive acts,
that is, the property that they all share of being about something, whether
visible or invisible, hearable or silent, external or internal, physical or
psychological.
We understand under Intentionality the unique peculiarity of experiences “to be the
consciousness of something.” . . . perceiving is the perceiving of something, maybe a
thing; judging, the judging of a certain matter; valuation, the valuing of a value; wish,
the wish for the content wished, and so on. Acting concerns action, doing concerns the
deed, loving the beloved, joy the object of joy. In every wakeful cogitatio a “glancing”
ray from the pure Ego is directed upon the “object” of the correlate of consciousness for
the time being, the thing, the fact, and so forth, and enjoys the typically varied
consciousness of it. (Husserl 1931: 223)
According to this definition, the notion of intention captures the fact that we
can attend to the same object or referent (e.g., a person, a thing, an idea) or
the same state of affairs (e.g., an event or a situation) in many different
ways, e.g., as an entity that we love, hate, need, want, admire, feel curious
about, despise, miss, judge, recognize, remember, feel longing for, and
so on. Intentionality implies a potentially unlimited set of relations (from
psychological to the kinetic, from theoretical to emotional) between the
Subject/Ego/Self and the same object/referent/entity. For example, as shown
in Figure 2.1, Subject A can relate to Object G in terms of “looks at” or in
terms of “worries about.”
Let us take the example of us connecting intentionally to a tree that is in
front of us. We can look at it and in so doing notice, take in information about
Criticism of Husserl’s use of intentionality 27

its long and twisted trunk or about the way the sun shines through the slightly
moving leaves; and/or we can worry about it, if the gardener has just told us
that the tree needs to be chopped down. In each case the tree does not change.
In fact, we keep assuming that it is the same tree over time even though
our sensations differ from one moment to the next because the light changes,
we change location, or our eyes keep moving (Husserl 1931: §§41–43; see also
Pritchard 1961; Pelaprat and Cole 2011). The notion of intention for Husserl
is meant to capture the manifold ways in which we as human beings can relate
to or be directed towards objects or events in the external world or in our own
mental world, giving them particular meanings.
As members of society, we do our noticing and interpreting in an unre-
flective manner, while engaged in what Husserl calls “the natural standpoint”
(Husserl 1931: §30).18 To get to the bottom of such a meaning-giving
process, we need to “bracket” this “natural” (or rather “cultural”), taken-
for-granted way of being in the world through a momentary “suspension” or
epoché (Husserl 1931: §§31–32) of our everyday assumptions, including
the assumption of the reality of the world around us and its materiality
(this “suspension” is meant to be more radical and thus more revealing than
the Cartesian doubt). Through the epoché Husserl (1931) performs a number
of “reductions” of our mental activity that reveal the fundamental role of our
consciousness in meaning-making, thereby isolating the “pure Ego” as the
“residuum” of such reduction and the “essence” (noema) of any intentional
experience.

2.5 Criticism of Husserl’s use of intentionality


as the foundation of meaning-making
Husserl’s use of the notion of intentionality with its implied conceptualization
of the dynamics between Subject and Object was not immune to criticism.
In fact, some of the arguments presented by anthropologists regarding the
use of intentions by speech act theorists echo earlier criticism of Husserl’s use
of intentions by Martin Heidegger and those who were either directly or
indirectly influenced by him.
The unsolvable theoretical disagreement between the two philosophers
came out in full force right after the 1927 publication of Heidegger’s Sein
und Zeit, during the failed collaboration on the article “Phenomenology” for
the Encyclopedia Britannica (Husserl 1997; Biemel 1977). But Heidegger’s
critique of Husserl’s most basic assumptions and concepts, including that
of intentionality, had been building for years, as shown by an analysis of his
pre-1927 lectures (Kisiel 1993; Sheehan 1993: 77–84). It is also apparent to
a close reading of Being and Time, which Husserl had not done before
accepting it for publication in his journal. In the book, Heidegger
28 Intentions in speaking and acting

problematizes Husserl’s understanding of both the “I” and “the world” in


several ways, the most obvious of which is the replacement of the notion of
“Ego” or “I” with the notion of Dasein understood as the human entity that
is historically contextualized and always projected towards the future,
including the anticipation of being-toward-death. One of the main points
of Being and Time is that we first and foremost exist in a world of everyday
activities where intentional acts and the distinction they imply between
Subject and Object are derived rather than foundational or constitutive.
Given that being-with-others is what characterizes us as humans (Heidegger
1962: 152), our experience in the world does not start from an isolated,
independent “I” that at some point “encounters” people and things. Rather,
the autonomous, individual (thinking, perceiving, knowing) “I” of traditional
western philosophy is a derived theoretical construction and “Knowing
oneself [Sichkennen] is grounded in Being-with” (1962: 161). Furthermore,
this foundational, literally a priori, coexistence is characterized by concern
for others as well as for the objects (e.g., tools) around us that are first and
foremost understood in terms of their use or, in Heidegger’s language, their
“in-order-to” (um-zu) existential character (Heidegger 1962: §15). But
Dasein’s transcendence, that is, the human existential condition of “Being-
already-alongside-the-world” (Heidegger 1962: 88) is hidden from us, and
we end up believing that the starting point of philosophical analysis should
be an observing Subject that can know or see the world. For Heidegger, it is
the favoring of epistemology – and the choice of perception as the most
fundamental way of “knowing” – that gives us the illusion that we are
routinely confronting, observing, getting to see or hear a world of static,
objectively present, atemporal entities whereas in fact our knowledge and
perception have been made possible by the special way of being-with
and being-along that distinguishes human existence.
If epistemology is a derived rather than a primary way of being in the world,
the methods and goals of traditional metaphysics needed to be reformulated
(Heidegger 1962, 1984, 1985, 1988). In philosophy, this meant to return to
the study of logic as the study of logos (Heidegger 1962: 47–49, 55–58, etc.;
1984; Dastur 2007). This revision of the meaning of “logic” included
a critical rethinking of Husserl’s approach, which, in Heidegger’s view,
was too epistemological and missed the opportunity offered by the phenom-
enological method to investigate the “mode of being” of the Subject/Ego/I
(see Biemel 1977: 297).
Heidegger’s critique of Husserl’s work as a problematic continuation of
a faulty western metaphysics had a direct or indirect impact on poststructuralist
approaches to human action, especially in France.19 Thus, Bourdieu’s (1977)
influential notion of “practice” – which had a major impact on socio-cultural
anthropology in the US (Ortner 1984) – can be read as a combination of
Criticism of Husserl’s use of intentionality 29

Husserl’s and Heidegger’s ideas (Throop and Murphy 2002), even though his
notion of habitus20 can also be traced back to other sources (Hanks 2005).

2.5.1 Husserl’s “passive synthesis” and “streaming living present”


in the unpublished manuscripts
The posthumous publication of the second and third volumes of Ideas
(Husserl 1980, 1989), as well as the vast number of lecture notes and other
manuscripts saved from the Nazis right after Husserl’s death in 1939 give us a
fuller picture of Husserl’s theoretical preoccupations over his entire career and
indicate that his position on intentionality was not exactly the way in which
Heidegger portrayed it (see on this topic Hopkins 1993; Moran 2000a; Zahavi
2001a). We learn, for example, that, in addition to being interested in the
process by which our consciousness gives meaning to sensations, feelings,
objects, and people, Husserl was also very interested in how the objects “out
there” exert a particular affective “pull” on our consciousness (e.g., Husserl
2001: 196–197). In his lectures in the 1920s, Husserl spoke about how
intentionality is constantly at work within a “streaming living present” (see
Donohoe 2000, 2004) that is pre-phenomenal, namely, prior to any specific
meaning-giving or meaning-fulfilling act. This is the domain of “passive
synthesis” where multiple objects coexist before one of them becomes
prominent and provokes an “awakening of an intention directed toward it”
(Husserl 2001: 198). This orientation toward an object may further develop
into a type of attentiveness that can become conscious interest. Through
Husserl’s published and unpublished manuscripts we come to appreciate the
simultaneous ongoing work of mind and world and the variations undergone
or displayed by the objects of our intentions (Husserl 1931). Through a
phenomenological analysis, which includes a discovery of phenomenal and
phenomenological modifications of our intentional acts (see Chapter 9),
we constitute some sounds into words and other sounds into music. More
generally, we discover the mutual constitution of self and world and the
fundamental role played by the Other in the constitution of the Ego from
birth to death (Husserl 1960, 1989; Donahoe 2004: 103; and see Chapter 10 in
this book). All of this is to say that, seen through the large body of
the Nachlass, which has been now made available almost in its entirety,21
Husserl’s ever-evolving project22 provides us with a more nuanced notion of
intention, which does not completely coincide with Heidegger’s hermeneutics
of Dasein’s affective involvement in the world, but shows meaning-making
to be a temporally unfolding process that must include an investigation
of both the working of the human mind and the particular ways in which
the world – other humans included – makes itself manifest to us, awaking our
attention and, thus, instigating our interpretive work.
30 Intentions in speaking and acting

2.6 A different theorist of speech as action: Ludwig Wittgenstein


In a way, Searle was an easy target for linguistic anthropologists, more of a
straw man than someone with whom to have a real argument. There were other
philosophers, no matter how decontextualized or naïve their points about
social interaction and non-western societies, whose reflections about action
would have been closer to the perspective that Rosaldo was promoting. One of
them is Ludwig Wittgenstein, a contemporary of Austin, who started out
within the analytic philosophy tradition, as a student of Russell and Moore at
Cambridge, but later criticized their limited understanding of how language
is actually used in activities.23 Wittgenstein was skeptical of traditional
philosophical discourse, and “intention” (for which he alternated between
two German words, Absicht and Intention) was at times one of the targets of
his critique (e.g., §§658, 659 in Philosophical Investigations).24
One of his observations corresponds to an anthropological understanding
of intentions as dispositions or goals that are made possible by the very activity
they seem to regulate:

An intention is embedded in its situation, in human customs and its institutions. If the
technique of the game of chess did not exist, I could not intend to play a game of chess.
In so far as I do intend the construction of a sentence in advance, that is made possible
by the fact that I can speak the language in question. (Wittgenstein 1958: §337)

A weak interpretation of this passage is that intentions can only be understood


within the frame of reference established by the contexts in which they are
expressed. A strong interpretation is that it is the code – broadly defined – that
makes certain intentions possible in the first place. What a chess player can
“intend to do” with a bishop will be constrained by the rules that define the
range of its moves, that is, the fact that it can only move diagonally and
without ever “jumping over” its own or the opponent’s pieces. If we took the
chess example as a metaphor for language use, we would have to say that
speakers would only be able to express intentions that make sense in the code
they have at their disposal. Among contemporary philosophers, this strong
version of linguistic relativity has been proposed by Karl-Otto Apel (1991)
who saw the range of available speech acts (in a given language) as determin-
ing the range of our intentional states.25 I will test this hypothesis by first
examining the meanings of the Latin terms from which the English intend and
intention are derived (through Old French) and then by looking for correlates
of such concepts in Samoan, a non-Indo-European language that will play a
key role in several of the chapters to follow. As we shall see, both languages
present some challenges for rendering English intention, intent, and intend.
The Latin terms tend to suggest embodied movement rather than an exclusive
mental attitude, and Samoan offers a range of possible expressions that go
The meaning of intention and intending 31

from affective disposition to willful planning but do not include the


affect-neutral mental state or attitude of intending that is taken for granted by
Grice, Austin, and Searle.

2.7 The meaning of intention and intending


As I will discuss in Chapter 5, within dominant paradigms of western philoso-
phy to be concerned with the problem of truth and falsehood has typically
meant to be interested in the relationship between the mind and some entity
that might or might not actually exist in the world but is made present by
certain thoughts or dispositions like will or desire. In reviewing the history of
this philosophical problem in antiquity, from Gorgias’ On not being to
Augustine’s De trinitate and beyond, Victor Caston (2001) urges us to separate
the issue of terminology from the conceptualization of intentionality as an
essential human faculty. I will return to this issue in Chapter 11, but for now let
me say that we should not quickly put aside terminology, because we can learn
from it. For example, the Latin noun intentio and the corresponding verb
intendere from which English intention and intend derive (through Old
French) have a range of meanings and associations that are more amenable
to an understanding of intentionality rooted in embodied action and inter-
action. When we look at classical texts in Latin, we find that intentio can mean
‘intention’ or ‘purpose,’ but it can also have a string of other, only vaguely
related meanings that include that of a bodily extension or effort toward some
object, place, action, or effect (see Anzenbacher 1972; Depraz 1995: 47–49;
Caston 2008; Aucouturier 2012: 8–11). This is made clear by looking at the
Latin quotations provided by Henry Nettleship (1889: 491–494), who starts his
long list of meanings of the entry intendo – the first person singular of
the present tense of the verb intendere used as the citation form in Latin
dictionaries – with the gloss ‘to stretch, point, towards or against’ and
continues with a set of related and revealing meanings, including ‘to fix one’s
attention upon’ as in intendere animum, mentem ‘to direct one’s thoughts or
mind to’ (p. 491), ‘to stretch or extend over or throughout something’ (p. 492),
to ‘turn’ (as in intendere pugnam ‘to turn the conflict in a certain direction’),
to ‘extend’ as in intendere iter, viam ‘to extend (or continue) a journey’, and
‘to strain, to exert the voice’ (as in omnes sonorum tum intendens tum
remittens persequetur gradus ‘the whole range of sounds, now straining
it, now releasing it’).26
These meanings have been lost in the English cognates, which are devoid of
any physical or emotional connotation. It is just such an indexically poor
notion of intention that is used in the contemporary literature on human
cognition and human action, where the common English term intention
is used together with other emotion-free notions such as “plan” and “goal,”
32 Intentions in speaking and acting

as done by Bratman (1989), for example, whose definition of intention as a


plan of action in pursuit of a goal27 is adopted by Michael Tomasello
et al. (2005). For Dennett (1987), an intentional system is a system that can
be assumed to act rationally on the basis of certain beliefs.
Despite some differences among authors, cognitive scientists tend to share
the view that a specific sense of “intending” as a human quality or activity can
be isolated and used as a general and generic label for acts of consciousness
with varying degrees of awareness. This type of folk psychology understands
intention in a sense similar to the dictionary definition: “a determination to
do a specified thing or act in a specified manner” (Webster, Second College
Edition, 1978: 733).
I have elsewhere referred to this understanding of “intention” as the
“narrow” definition (Duranti 2006a) to be contrasted with a “broad” definition,
which would be close to Brentano’s and Husserl’s notion of intention and
would allow for all kinds of relations between human consciousness and
“the world.”
The understanding of intentions in analytic philosophy – and to some extent
in European or American law – seems to be based on a rather generic sense of
intending as a state of mind devoid of or separable from emotions, embodi-
ment, and history, even though such attributes or dimensions may cause or
explain how what we might call “intention” in English came to be (for a review
of the use of the notion of intention in German criminal law, see Taylor 2004).
Can we gain from investigating in some detail whether or not the common
meaning of the English “intention” and “intending” finds translations in
other natural-historical languages? It is strange that, given the importance of
linguistic labels and metaphors in scientific discourse, cognitive scientists or
philosophers writing in English have not shown interest in whether the notion
of intention is something easily translatable in other languages. But even
linguistic anthropologists, who have criticized psychologists or philosophers
for being too “western,” have not engaged in the cross-linguistic analysis of
the taken-for-granted scientific vocabulary of the human sciences. Some subtle
differences in the available words for such key theoretical concepts like mind,
mean, meaning, and intention are found even among European languages. For
example, German and French do not have a separate word corresponding to
the English noun mind (from Old English gemynd ‘memory, remembering,
state of mind, purpose, intention’) or the Italian mente (from Latin mens). The
academic field of “the philosophy of mind” is translated by German scholars as
die Philosophie des Geistes – where Geist also means ‘spirit’ (as in Hegel’s
famous Phänomenologie des Geistes ‘Phenomenology of Spirit’ or ‘Phenom-
enology of Mind’) – and, in a similar fashion, the French use philosophie de
l’esprit where esprit has connotations similar to the German Geist. When
we move outside of Indo-European languages, the rendering of such concepts
The meaning of intention and intending 33

gets more challenging because it is more difficult to separate the “mental” from
the physical or the emotional (the mind–body divide is not lexically encoded in
the same ways across languages). To provide an example of such challenges,
in the next section I explore whether one may be able to convey the notion
of “intention” or “intending” in a non Indo-European language, Samoan.

2.7.1 Translating intention: an exercise in Samoan ethnopragmatics28


In Samoan, it is difficult to find a term that can closely and fully capture
the English intention. The translations found in the dictionaries are at best
approximations and always with connotations that do not belong to the English
intention or intent. Rev. George Pratt’s 1893 Dictionary (the third and
last edition that he authored) gives manatu as the Samoan translation of
“intention.” But elsewhere in the dictionary, the word manatu is glossed
‘a thought’ (as a noun) or ‘to think, remember, consider’ as a verb (Pratt
1893: 14, 20, 35, 207). In all of my recordings of informal conversations and
formal meetings, manatu (mostly typically pronounced /magaku/, see Appen-
dix B) is used and translated by bilingual Samoan-English speakers with the
meaning ‘think’ (as in /magaku `oe e makagā?/ ‘do you think it looks bad?’29)
or ‘thought’ (e.g., /lo`u ā magaku/ ‘my very thought’30 or /so mākou magaku
legā/ ‘some thought of ours’).
Not surprisingly, given Pratt’s key role in the translation of the Bible into
Samoan (see Chapter 4), manatu is found in the Samoan Tusi Paia as one of
the translations of English intent – the word intention is not used in the King
James version – together with filifiliga ‘deliberation’ (Pratt 1893: 159) or
‘decision’ and the generic phrase le mea . . . ai (lit.) ‘the thing for which . . .’
which is still found in conversations to mean ‘the reason for which’ as in `o le
ā le mea `ua le fia . . .? ‘what is the reason you do not like . . .?’ The word
manatu is also found as one of the Samoan translations of purpose, like in the
passage “And, behold, I purpose to build an house unto the name of the LORD
my God . . .” (I Kings 5:5).31 Other translations of purpose in the Bible include
loto (see below), mafaufauga ‘consideration, thought’ (Pratt 1893: 199),
pule ‘command, order, decision,’ and finagalo (when speaking about God),
an honorific term (`upu fa`aaloalo) appropriate for talking about a chief’s
will (Pratt 1893: 159), wish, or decision, which can function as either a noun
or a verb.
In the 1911 Pratt’s Grammar and Dictionary of the Samoan Language, an
edition revised and considerably expanded by J. E. Newell, we find a second
translation of “intention,” namely, fa`amoemoe, which is also given as the
Samoan verb corresponding to English ‘1. to expect, to hope. 2. To lay wait
for, 3. To prepare for war’ (Newell 1911a: 103). The entry fā’amoemoega,
with the nominal suffix –ga, is translated as ‘hope.’ This semantic extension
34 Intentions in speaking and acting

of fa`amoemoe is confirmed in George Milner’s more recent dictionary (1966:


147) where we find fa`amoemoe translated as both ‘hope’ and ‘purpose.’
The expression fa`amoemoeina, that is, fa`amoemoe with the addition of
the suffix –ina (which can give it a sense close to our passive voice) is
translated as ‘be expected, be intended for, meant for.’
The word fa`amoemoe is built on the word moe ‘sleep,’ whose reduplica-
tion, moemoe, gives a sense of continuous sleep (‘go on sleeping’ in Milner
1966: 146). The addition of the prefix fa`a–, which here works in the sense of
‘in the way of,’ suggests a state of daydreaming or wishing that something
may happen in the future, hence the English translation of fa`amoemoe with
‘expectation, hope.’
Milner’s 1966 Samoan Dictionary does not have “intention” as a separate
entry in the English–Samoan part. But we find it in the English–Samoan
portion as a paraphrase of the entry “will,” which Milner (1966: 462) translates
with loto and a number of semantically related terms:
Will n. (determination, intention): loto; (po[lite]): finagalo; (desire, consent): malie;
(testament, last wish): māvaega; (will-power): fati (lana).

However, neither manatu – as we saw above – or loto have the distinct


character of the English intend or intent. Manatu encodes the notion of a
generic mental activity similar to the English verb think – which can be
extended to remembering by adding the suffix –a (producing manatua) –
and loto seems closer to English attitude, disposition, inclination, or feeling.
It can be used to refer to the character of a person, as a whole, as in the
expression e le iloa le loto o lena ali`i in Schultz (1953), which was translated
in English by Brother Herman as ‘I don’t know the views of that person’
(emphasis added in both Samoan and English).
Gerber (1985) interprets the Samoan loto as both ‘depth’ (as in i loto o le
sami ‘in the depths of the sea’) and as ‘the repository of some of a person’s
thoughts and feelings’ (Gerber 1985: 135). She adds that “The feelings that
occur here, lagona i le loto ‘feelings in one’s depths’ are roughly equivalent to
“emotions” in English” (1985: 135). But then she goes on to distinguish
between the feelings in the loto from the feelings in the body, lagona
(Gerber 1985: 135–136).
As we shall see in Chapter 4, in the nineteenth century the missionaries and
their Samoan consultants used loto as a noun for ‘soul’ and, at times, ‘heart’
(in a non-literal, metonymical sense, given that the body part is called fatu in
Samoan), a translation that is only in part reflected in the Samoan–English part
of Milner’s dictionary where loto is glossed ‘will’ as well as ‘[h]eart, feeling
(as opposed to mind and soul)’ (Milner 1966: 112).
The excerpt below, taken from a speech given in a meeting of the village
council (fono) in Falefā, shows the use of the term loto (here pronounced /loko/)
The meaning of intention and intending 35

in conjunction with fa`amoemoe ‘hope, expectation’ as grammatical subject of


the verb laloga ‘feel.’ Such pairings of two terms – called “couplets” by students
of poetics – with closely related but not necessarily identical meanings are
common in Samoan oratory. I will leave loto (/loko/) untranslated even though
‘heart’ or ‘soul’ would work in English.

(2) (Fono April 7, 1979)


465 e laloga fo`i lo`u loko ma lo`u fa`amoemoe,
TA feel also my ? and my hope
‘my loto and my hope feel’
466 . . . `o so`u fa`agoagoaga i ia kūlaga (. . .)
‘some sadness of mine about those issues’

The conceptual connection – but by no means synonymy – between loto


and fa`amoemoe is supported by the long list of compounds that the entry
loto has in dictionaries compiled by Pratt and Milner almost a century apart.
The lists include loto vaivai ‘a weak-spirited man’ (Pratt 1893: 8; see also
Mageo 1991: 408–409) and loto-leaga ‘jealous, envious’ (Milner 1966:
112) (lit. ‘loto-bad’) or ‘evil disposed’ or ‘bad disposition’ in Pratt
(1893: 14, 25; see also Mageo 1991: 415; 2010: 124). In fact, it seems that
almost anything can be added to loto to form a compound indicating
a particular type of disposition or inclination. For example, loto-`āiga
(lit. ‘loto-family’) means ‘loyal, devoted to one’s family and kin’; loto-
nu`u (lit. ‘loto-village’) ‘to love one’s country’ (Pratt 1893: 191), loto-pua`a
(lit. ‘loto-pig’) ‘bestial, without thought’ (Pratt 1893: 191), loto-fiafia (lit.
‘loto-happy’) for ‘jolly.’32
Some of the expressions that belong to this set are found in my own
recordings of Samoan conversations, prayers, and formal speeches. They
include loto maulalo, which could be translated as ‘humble disposition,
humble demeanor,’33 loto alofa, corresponding to something like ‘loving
heart’ or a disposition to be generous, and loto fesoasoani ‘helpful heart’ or
disposition to help others (this is found in prayers). Even used by itself, loto
carries a similar meaning of disposition to act or to feel in a certain way. This
is the way loto was used, for example, by Chief F in an exchange here partly
reproduced in (3). His line comes at the end of a semi-serious back-and-forth
between the two of us that had started with Chief F stating: /lelei Sāmoa i
Amelika/ ‘Samoa is better than America.’34 This was at the time a fairly
typical conversational opening during a first encounter between a member of
our research team and adult Samoan speakers. Chief F goes on to explain that
Samoa is better than America because there is no war (tau) and there are no
fights (misa) (/e lē kau. e lē misa/). When I counter with a purposefully
provocative move that in the village where I was living there were lots of
fights (/misa kele i Falefā/), Chief P, our host, shows surprise and Chief T,
36 Intentions in speaking and acting

who had accompanied me to the village where the exchange is taking place,
contradicts me by saying ‘no, not much fighting’ (/leai lē misa tele/). Rather
than giving up, in line 279 I reassert my position this time with /e malosi le
lima/ literally ‘the hand is strong.’ This is a colloquial metonymical expression
used to characterize people who easily get into physical fights.35 Chief T’s
laughter in line 280 is probably due to the fact that he is both amused and
surprised by the fact that I knew such a colloquial expression and used it at the
right time. Chief F partly accepts my claim by providing a different framing
for it. His utterance in line 281 repeats the same predicate (malosi ‘strong’)
and the same syntactic structure (Verb þ Subject) of my turn in line 279. This
is an elegant rhetorical move whereby he rephrases what I said in terms of
Samoan temperament. It is the loto (here pronounced /loko/) that is strong, not
(only) the hand. Possible translations include ‘the will is strong’ or ‘the spirit
is strong.’

(3) (“The watch” – conversation of researcher, A, talking with three Samoan


chiefs in the village of Falealili, while sitting on mats inside a traditional
Samoan house; recorded in 1979)
279 A; e malosi le lima.
Pres strong Art hand
‘they have fist fights.’ (lit., ‘the hand is strong’)
280 T; ((LAUGH)) huhuhu
[
281 F; malosi le loko. (0.7) `ā? loto (pronounced /loko/)
strong Art loto
‘the will/spirit is strong, (0.7) isn’t it?’

All of these uses have characteristics that differentiate loto from the English
intention and intend:
1. rather than referring to planned actions, they often express dispositions or
inclinations that allow for different resolutions or manifestations;
2. they may or may not express ways of being of which a person is conscious;
3. they tend to have an affective, emotional meaning; and
4. they tend to imply embodied attitudes or practices (i.e. they are usually
expressed and interpreted as a combination of verbal and kinesic
behavior).
Do these observations about the lack of a close translation of the English
terms intention or intend imply that Samoans do not have the notion of
intention in the specific English sense? This is a question that should be
approached empirically. For one thing, we know that individuals or groups
of speakers can and do adopt new words with new meanings that eventually
make it into the repertoire of the larger speech community (Morgan 2014). For
The meaning of intention and intending 37

example, Samoan borrowed words like time (Samoan taimi) and duty (Samoan
tiute) from English and natura ‘nature’ from Latin36 (see Chapter 4 for a brief
discussion of loanwords and lexical innovations introduced in the Samoan
translation of the Bible). Languages can also use existing words for a new
meaning. This is what I think the Bible translators chose to do when they
translated intent or semantically related words like purpose. Something similar
seems to have also happened in the juridical domain of the western-inspired
Samoan court or the new Samoan Constitution. Thus, in 2000, I found in a
newspaper article the word loto used in a context that suggests the attempt to
convey English intent or intention. The article appeared on p. 5 of The Weekly
Samoa Post, April 17, 2000. It was inside a report on a much publicized
murder case that involved two government ministers accused of having
convinced the son of one of them to cold-bloodedly execute another minister.
In the article, loto was used in the context of the phrase le moliaga o le fasioti
tagata ma le loto i ai, which could be translated in English as ‘the accusation
of killing a person (or persons) with the intent (loto) to (do) it.’
In the 2008 Constitution of the Independent State of Samoa, which is
available on the internet in both Samoan and English, the word “intentionally”
is used in the English text of Article 5. The corresponding Samoan text of Article
5 is provided in example (4) below, accompanied by my interlinear gloss. This
example shows that the Samoan text utilizes a modification of the term provided
in Pratt’s dictionary, namely, fa`amoemoe, even though here the word is
modified with the suffix –ina, which gives it a passive-like meaning.

(4) (From the 2008 Samoan Constitution; Article 5)


E leai se tagata o le a aveesea lona ola ma le faamoemoeina, [. . .]’
Pres no a person Fut take-awayþsuffix his life with the expecting
‘No person shall be deprived of his life intentionally, [. . .]’

The key phrase in question here is ma le faamoemoeina corresponding


to ‘intentionally’ in the English version of the Samoan constitution. The
expression is composed of the conjunction ma ‘with,’ the article le ‘the,’ and
the verb fa`amoemoe (spelled without the inverted apostrophe in the prefix
fa`a-) modified by the suffix –ina (Milner 1962; Duranti 1981a; Cook 1996).37
One possible literal translation of the Samoan sentence would then be ‘no one
shall (be) with the expectation that his life be taken away.’ This translation is
supported by the translation of fa`amoemoeina with ‘be expected, be intended
for, meant for’ in Milner’s dictionary (1966: 147).
It is not a trivial detail that these last two Samoan proposals for conveying
the meaning of what English expresses as intent, intention, or intentionally are
found in the context of an imported western practice, that is, a court trial, and
of a western-style written document (the Constitution), where individual
38 Intentions in speaking and acting

responsibility must be ascertained and evaluated by taking into account an


individual’s state of mind (e.g., motives and plans) and individual rights that
are meant to be protected in general, that is, universal terms. This is an
imported or borrowed practical logic that seems distinct from the ways
Samoans had thought and acted in the past and, to some extent, still did in
1978–1979 when I recorded the political-judicial councils (fono) I discuss in
Chapters 3, 4, and 5. In those arenas, when a case is being argued, the
emphasis is more on finding out whether or not someone did something
and discussing the effects or consequences of a person’s actions. In such
discussions the primary concern is with the consequences of someone’s
actions. Particular attention is given to ways to restore social harmony between
parties and to punish those who have caused embarrassment for the village
council or people of high status. A person’s intentions are not discussed.
Statements that can be interpreted as explanation for someone’s behavior tend
to be about social roles, relations between individuals or communities, the
influence of alcohol (see §3.4.1), or the effects of particular life experiences
(e.g., whether someone has served the community or has grown up and gone to
school in a foreign country).
What are the consequences of this discussion for the model of intentionality
proposed by Searle (1983)? Could we say that, were we to start from Samoan
(or other similar languages) instead of English, our theory of intentionality
would be different? Or would Searle dismiss the Samoan data discussed above
as irrelevant as he did regarding Rosaldo’s claim that the Ilongots do not
have the institution of promising (see §2.3)? Would we want to say that
in some speech communities the concept of intention is hypo-cognized,
that is, hidden or not recognized? Or should we just give up on using the
(English) notion of intention because it is too culture-specific? I will start
to address some of these questions in the next section and then return to them
in Chapter 11.

2.8 Not giving up on intentions


As I discuss in some detail in Chapter 3, my first fieldwork experience in
Samoa prompted me in the 1980s to argue that intentions are overrated in
speech act theory and that a certain amount of social interaction takes place and
is interpreted, at least in a place like Samoa, without participants having much
of a concern for what a particular person meant to do. As I will discuss in
Chapter 3, my line of reasoning was consistent with other accounts of Samoan
interactional practices (Mead 1928, 1937; Shore 1982; Ochs 1982). It also
resonated with the critique of intentionalist theories of actions by the cultural
and linguistic anthropologists mentioned earlier (see also Du Bois 1987, 1993;
Moermon 1988). My own account and those of other ethnographers who dealt
Not giving up on intentions 39

with similar questions shared at the time the assumption that interpretive
practices are culture specific and fundamentally incommensurable. In rejecting
the notion of intention as used by (mostly English-speaking) speech act
theorists and some cognitive scientists, such a relativist stance either rejects
altogether the idea that intentionality is a relevant factor in human action or
takes an agnostic position. The dissatisfaction with the analyses of social acts
based on intentions (Heritage 1990/91), the difficulty of finding an adequate
translation for the English intention in languages like Samoan, and the fact that
in a number of societies members avoid engaging in explicit mind-reading
(see Chapters 3 and 8) might lead some researchers to avoid talking about
intentions altogether. This result would, however, be unfortunate, especially
for cross-cultural comparison. Regardless of the type of data collected
(e.g. introspection, interviews, audio-visual recordings of spontaneous inter-
actions, written texts of various kinds, results of experiments), all researchers
need concepts that are general enough to be used across situations. We need to
avoid producing analyses that appear ad hoc. One solution is to characterize
intentionality as a graded phenomenon, subject to recognition, oblivion, or
manipulation. The notion of an “intentional continuum,” which I will propose
in Chapter 11, is a more radical solution for a number of reasons, including
the fact that it allows for ideology to play a role at the metadiscursive level.
We can then, for example, distinguish between forms of social organization
that recognize the “intentions” of some social actors – in particular social roles
within particular social occasions – but not of others.
To claim that intentionality is graded also means to question or even reject
the dichotomy accepted by many authors, Searle included (see above), between
two different types of intentionality, one roughly corresponding to Brentano’s
and Husserl’s notion – as the “aboutness” of our mental activities (which Searle
writes “Intentionality” with the capital “I”) – and the other (“intentionality”
with the small “i”) coinciding with the ordinary use of the verb intend in
English in utterances like I intend to take you out for dinner (e.g., Chisholm
1957, 1967; Dreyfus 1982; Fisette 1999; Searle 1983; Voltolini and Calabi
2009). Although linguistic encoding should not be seen as definitive evidence
of the relevance of the concept of intentionality in a given speech community, it
should not be ignored either. The language (as a code) and discourse (as a
practice) of intentionality provides us with important hints about the local
possible understandings of human experience, which we expect to be related
to particular forms of social organization as well as daily preoccupations,
including the exploration, reproduction, and evaluations of moral and aesthetic
values. Whether or not an individual’s goals, ends, or purposes in doing
(or saying) something will be recognized or considered relevant to the inter-
pretation of that individual’s actions might depend on a number of emotional
dispositions, social arrangements, and interactional practices that will have to
40 Intentions in speaking and acting

be detected in each case. An anthropology of intentions can also help us


to clarify whether the model of intentionality that we identify for a particular
group is really “their” (local, folk) model or the projection of a series of
practices that “we” also have but do not recognize in our theories.

2.9 Alternative “western” ways


Over the years I have come to see the other side of the debate about intention-
ality, namely, I have realized that some of the criticism of intentionalism
within anthropology was empirically and theoretically questionable.
From an empirical point of view, the fact that in some communities people
seem to refuse to speculate about states of mind and might subscribe to a
theory of the “opacity of other minds” (Rumsey and Robbins 2008) does not
mean that they never or rarely engage in reading the minds of others. As
I discuss in Chapter 8, when we look in some detail at the ethnographic
evidence, it appears that even in those communities for which a claim of
some kind of “opacity of other minds” has been made there are contexts where
people do guess what one was, is, or will be thinking or feeling. Anthropol-
ogists need to resist the pressure to create sharp dichotomies (e.g., between
the West and the rest of the world) in order to be heard by other social
scientists or by society at large. The first step in modulating differences and
accepting a discourse of preferences and tendencies, which foregrounds con-
text without necessarily embracing relativism, is to examine our own intellec-
tual history. Not only have we not been “modern” for very long, as Bruno
Latour (1993) reminded us, but in the very heart of “western modernity,”
namely, twentieth-century Euro-American epistemology and ontology, a wide
range of positions can be found about the human mind and social action.
A brief examination of the history of western ideas should raise doubts about
whether there is such a thing as “the western” theory of interpretation and
whether such a theory is, by definition, built on individual intentions. A good
number of Euro-American philosophers and social scientists over the last
century or so have not given speakers’ (or social actors’) intentions the central
role they have in speech act theory and especially in Searle’s writings. For
example, American pragmatists Charles S. Peirce, William James, John
Dewey, and George H. Mead offered models of human cognition that are
grounded in practical action, in which the knowing-how is just as important as
knowing-that. Rather than looking at intentions in the mind, pragmatists
tended to look at consequences and effects of human actions (see Chapter 4).
Stated in this way, we could call Samoans pragmatists or say that there is
something in the pragmatists’ view of action that resembles Samoan
ethnopragmatics, even though such a move might obscure the specific
contributions that a study of Samoan ways of speaking can offer to us.
Alternative “western” ways 41

Similarly, as shown by the English translation of writings by Lev


S. Vygotsky, of M. M. Bakhtin, and V. N. Vološinov, more nuanced, dialogic
views of meaning-making were developed and proposed in Europe in the first
part of the twentieth century. It was in particular Bakhtin’s view that the
language we use is populated with the intentions of others that alerted me in
the early 1980s to the possibility of rethinking meaning-making through
Samoan ways with words (Duranti 1984, 1988; and Chapter 3). More recently,
I have come to extend this perspective to linguistic relativity, in particular in
rethinking how speakers may be not just constrained or guided by grammatical
structures or the lexicon – which is the original Whorfian position – but also by
particular narrative genres (Duranti 2011). I will return to this idea in Chapter 5.
But there are also new trends that converge on similar positions, this time from
the experimental sciences. At the turn of this century, neuroscientists using
experimental methods first developed in Parma, Italy, argued that empathetic
reactions and non-reflexive, pre-conscious, pre-rational interpretations of
others’ actions are very common in both primates and humans (e.g., Gallese
2000, 2003, 2006; Rizzolati, Fogassi, and Gallese 2000; Rizzolati and
Sinigaglia 2006; Iacoboni 2008; Iacoboni et al. 2005; Liew, Han, and
Aziz-Zadeh 2011). These findings do not rule out the occurrence and relevance
of rationalization of our own and others’ behavior but do suggest that rational
choice, including the assumption of specific intentions behind – and therefore
understood as causes of – our actions, coexists with and sometimes might be
replaced by unmediated neural reactions and instinctive interpretations that
only later, post hoc, can be given rational accounts.
Of course, we can easily go the other way and overemphasize spontaneity
even in contexts in which in fact monitoring of actions and interactions is at
work in visible ways if we are willing to look. For example, in a study of
musicians’ behavior in front of an audience, Berger (1999) argued that, in the
middle of what appears as a most spontaneous and improvised performance,
the musicians on the stage may exhibit different levels of reflexivity about their
own actions depending on a variety of factors including not only the type
of music they are playing (with each genre being syncretically linked to
particular types of public behavior during performance), but also their fellow
musicians’ behavior as well as their audience’s reactions or expectations. This
type of phenomenologically inspired study where behavior is linked to
tradition-specific dispositions – the habitus of the “field,” in Bourdieu’s terms –
suggests that the notion of intention may be more effectively reconceptualized
by reconnecting it with other key notions such as attention, empathy, and
stance (see, among others, Arvidson 2003, 2006; Depraz 2004; Hollan and
Throop 2008; Berger 2009; and Chapters 9 and 11). It also suggests that,
together with the goal-directed actions that can be interpreted in terms of what
I called earlier “the Standard Theory,” we should be also prepared to examine
42 Intentions in speaking and acting

the role that improvisation plays not only in the art forms where it is
recognized and practiced (e.g., improv theater and many musical genres
around the world), but in everyday life (Sawyer 2001; Duranti and Black
2012). I will explore this perspective in Chapter 10, where I use a phenomeno-
logical understanding of intersubjectivity to question the sharp distinction
between I-intentions and we-intentions. Even in actions that appear strictly
controlled by an individual, other human beings may play a role, even when
they are not physically present.
3 The avoidance of intentional discourse:
a Samoan case study

3.0 Introduction
In this chapter I return to a case study first presented in a paper I delivered at
the 1983 Meetings of the American Anthropological Association, later distrib-
uted as a technical report (Duranti 1984) and then slightly revised as a journal
article (Duranti 1988), and as a chapter of an edited book (Duranti 1993b).
What I present here is a revision of the original paper that takes advantage of
some additional inspection of the Samoan data collected in 1978–1979 and
frames the issues in the broader context of this book.
Overall, the data presented in this chapter support previous claims that
Samoan interpretive practices are fundamentally non- or even anti-
introspective and as such tend to privilege a reading of a person’s actions in
terms of their effects on social relations and on the public face of particular
institutions, groups, and positional roles. Samoan epistemology, that is, shows
similarities with a pragmatist interpretation of human action whereby the truth
of a statement tends to be evaluated in terms of its effects. Even when people
seem to be searching for causes, reasons, or sources (māfua) of wrongful
behavior, discussions about “causes” usually have a brief discourse life as
they tend to be quickly dismissed or ignored by the participants in the
interaction (see §3.4.1 and Chapter 5). For example, we will see that even
though lying is considered by the matai or ‘titled holders’ a violation of trust
and something to condemn, it usually does not evoke a discussion about what
the liar was trying to achieve by not telling the truth, with the exception of
stereotypical explanations such as being under the influence of alcohol regard-
less of whether there was empirical evidence that alcohol was involved. This
suggests that in Samoa, even though there is language that could be used for
speculating about one’s emotions and motivations (see Gerber 1985), there is a
cultural dispreference for engaging in such discursive practices and a recourse
to typifications of standard cause-and-effect connections as opposed to an
individual-specific analysis that might reveal unexpected and potentially unset-
tling results. This is particularly true in political or juridical arenas like the
Samoan fono where the style of speaking (or speech genre) used is designed to

43
44 The avoidance of intentional discourse

be opaque or indirect, a quality of verbal performance frequently used in


Pacific societies to help protect speakers from offending each other or from
provoking retaliation (Bloch 1975; Brenneis 1984, 1986, 1988; Besnier 1989,
2009; Watson-Gegeo and White 1990; Duranti 1994: chapter 5; Throop
2010a: chapter 5)
The more extensive and detailed analysis of the transcripts presented here
also uncovers something new, namely, nuances and differences among partici-
pants that were missed in previous analyses. Key participants in the village
council are shown to take a different stance on the events discussed or the
criteria relevant for assessing responsibility. For example, kin relations
(between the defendant and the other major culprit in the alleged violation)
are brought up as an aggravating circumstance by one high-ranking member of
the council but dismissed by another. This difference of opinions is common in
Samoa, but it is sometimes missed without a detailed examination of what
people actually said (see also Chapter 8).

3.1 Avoidance of making explicit intentions and motivations


While we were in Samoa in 1978–1979, Elinor Ochs, Martha Platt, and I were
struck again and again by what we perceived as a recurring reluctance among
our Samoan neighbors, friends, and consultants to engage in speculation about
people’s motivations and intentions. This reluctance had been written about by
Margaret Mead in her Coming of Age in Samoa (1928) and by Bradd Shore in
his 1977 dissertation, a copy of which we had taken with us.1 Mead had
introduced the lack of interest in inner states of others through the description
of the meaning of the word musu.
There is one word musu which expresses unwillingness and intractability, whether in
the mistress who refuses to welcome a hitherto welcome lover, the chief who refuses to
lend his kava bowl, the baby who won’t go to bed, or the talking chief who won’t go on
a Malaga [organized group journey]. The appearance of a musu attitude is treated with
almost superstitious respect . . . Once this attitude has appeared, a Samoan habitually
gives up the struggle without more detailed inquiry and with a minimum of complaint.
This fatalistic acceptance of an inexplicable attitude makes for an odd incuriousness
about motives. (Mead 1928: 122–123)
The Samoan word musu expresses an embodied mood as much as a state of
mind and, as made clear by Mead in the above quote, the musu condition is
something that must be reckoned with. It is made manifest by a non-doing or
even a “shutting down,” which typically expresses itself (and is interpreted) as
a refusal to engage in any negotiation (see also Mageo 2010: 126–127). The
contexts in which musu may be experienced and its implications are captured
by some of the songs and stories collected at the end of the nineteenth century
by Dr. Augustin Krämer, surgeon of the German Royal Navy turned
Avoidance of explicit intentions and motivations 45

ethnographer (see Chapter 4), but the translations (in German) sometimes miss
the specific meaning of the Samoan expression. Thus, in “The Song of the Bad
Girl” (‘O le pese o le teine leaga), the word musu is used to describe the fact
that the “bad girl” (teine leaga) is resisting the advances of the man who sings
the song. In this case, Krämer translates at one point musu with the German
Ablehnen ‘refusal, rejection’ (or ‘refusing, rejecting’) and another time with
Ausweichen ‘evasion’ (or ‘evading, dodging’). Both translations are adequate
if they are meant to capture the fact that the male singer is treating the girl’s
behavior as something that he expects and, in fact, is used to (`ua masani ai)
(Krämer 1903: 347). But when translating a story about a stubborn heroine,
Sina, who tells her parents that she refuses to consider the marriage proposals
of high-status men,2 Krämer (1902: 136) renders musu with the negative of
wollen ‘want’ (Krämer 1902: 133; see also Krämer 1902: 204, and the English
translation in Krämer 1994: 166). The generic verb wollen and its English
counterpart want cannot quite convey the force of a mood that should not be
questioned or speculated about.3
Beyond the common problems of translations (a focus of Chapter 4), this
brief discussion of the meaning of musu reveals an aspect of Samoan ethno-
psychology that is particularly relevant for the topic of this chapter and, more
generally, for the book as a whole. As will be made evident from the case to be
discussed below, Samoans have strong feelings about responsibility and obli-
gations. Individuals and groups may be criticized, punished, or expelled from a
community for not having matched the expectations associated with their
positional role in society, but they will not be forced to explain themselves
in terms of their motivations or intentions.4 At times, they may be asked to
mention what they did do – this is a concern with “truth” which I will discuss
in Chapter 5 – but little talk and time will be spent examining an individual’s
actions beyond what is known from public behavior and consistent with the
community’s conventional wisdom about human nature in general.

3.1.1 The 1978–1979 project: Samoan language acquisition


and socialization
One of our challenges was to further test the claims made by Mead and Shore
about Samoan cultural practices. We engaged in this task by examining our rich
corpus of verbal interactions among family members, friends, and acquaint-
ances, in informal situations as well as among high-status individuals in formal
gatherings. The outcomes of our inquiry came out in stages. The first installment
was a paper Elinor Ochs wrote as a contribution to a special volume commemor-
ating a decade of the journal Language in Society (Ochs 1982). In that article,
written in 1981 while Elinor and I were postdoctoral fellows at the Australian
National University (ANU),5 she argued that Samoan caregivers do not use
46 The avoidance of intentional discourse

so-called “baby talk.” This claim was based on the observation that in the 18,000
pages of handwritten transcripts of household interactions with children that
Elinor and Martha had audio-recorded in Samoa, there was no trace of the
special lexicon, morphological modifications, or simplified syntactic construc-
tions that are commonly found in the speech of mothers and other caregivers to
children in the US and other places where “baby talk” had been documented
(Ferguson 1964, 1977). In addition, when compared with previous studies of
language acquisition and adult–child discourse in the US, the data recorded in
1978–1979 revealed some other intriguing differences. While Samoan care-
givers were like those in the US in responding to children’s intelligible utter-
ances in a topic-relevant manner (e.g., agreeing, disagreeing, continuing to
discuss a particular topic), they were different in one particular type of discourse
activity: they did not “expand” children’s utterances, that is, they did not try to
put in words possible guesses of what a child might have tried to say while
producing an incomplete or unclear utterance. Samoan adults and older children
did not do what the adult speaker does in example (1):
(1) (Examples of “expansions” of a child’s utterances; from a recording made on
May 29, 1990 by then UCLA student Debbie Heick)

Child; mo- ycle


Adult; motorcycle? EXPANSION
Child; RIDE IT?
Adult; oh you wanna ride the motorcycle. EXPANSION

Ochs (1982) presented the absence of such verbal exchanges in the data
collected in Samoa as empirical evidence of a lack of concern for a child’s
intentions in doing something or causing something to happen.
Expansions are interpretations of children’s intentions and reflect middle-class Anglo
caregivers’ assumptions that children can and do control and guide their actions towards
some goal. Further, they manifest middle-class caregivers’ perceptions of very small
children as social persons . . . traditional Samoan caregivers do not share all of these
assumptions. Infants and very young children are generally not treated (1) as socially
responsive beings (cooperative); and (2) as being in control of their actions. (Ochs 1982: 91)
This argument was reframed in a broader comparative perspective in an article
that Elinor Ochs and Bambi Schieffelin wrote in 1981 while we were at ANU
(Ochs and Schieffelin 1984).6

3.1.2 The Laboratory of Comparative Human Cognition:


neo-Vygotskian perspectives
A second paper on the lack of interest in speakers’ intentions in Samoan
discourse came out of my experience as a postdoctoral fellow at the University
of California, San Diego, where in 1983–1984 I was a member of the
Avoidance of explicit intentions and motivations 47

Laboratory of Comparative Human Cognition (LCHC), directed by Michael


Cole, a cross-cultural psychologist who was influenced early on in his career
by the socio-historical approach pioneered by psychologists Lev Vygotsky,
Alexander Luria, and A. N. Leontyev (also known as “the troika” of Soviet
psychology). My participation in the weekly seminars at LCHC and the
frequent informal meetings and conversations with LCHC members, Michael
Cole and Peg Griffin in particular, opened up for me a new intellectual
universe. I became familiar with Vygotsky’s theory of the role of semiotic
mediation in human development, Luria’s writings on the influence of
schooling on cognition, and Leontyev’s speculations about the notion of
activity and its relevance to the formation of human consciousness. The
English translations of the works of these three authors then available together
with the rich intellectual context of LCHC gave me a new way of looking at
some of the data I had collected in 1978–1979 in Samoa. I was particularly
intrigued by Vygotsky’s idea that intra-psychological (individual, inner,
hidden) processes – including remembering – are first inter-psychological, that
is, they exist first as the by-products of interactions between individuals and
only later are internalized. I was even more excited about the neo-Vygotskian
perspective that was being discussed in those years at LCHC. Ed Hutchins was
a member of LCHC and Jean Lave’s research was often mentioned during our
meetings (see Chapter 10). In addition, it was the beginning of computer-
mediated communication, and our conversations on campus would often spill
over into short and long messages exchanged at 300 baud that had a much
wider range of participants. The neo-Vygotskians were pushing for something
quite radical at the time, namely, the idea that tools – including symbols like
language but also computers and writing as a cultural activity – are more than
mediation devices that allow the exchange of ideas and learning. They make
possible certain ways of thinking, feeling, and acting, favoring (or “affording”
as we learned from James Gibson’s ecological approach) certain activities or
modes of interaction over others. This was a version of constructivism that
paid attention to both the material and ideational conditions of an activity.
I saw parallels between the developmental theories championed by Michael
Cole, Peg Griffin, Esteban Diaz, and Luis Moll, among others, and an anthro-
pological understanding of conflict-management as practiced by Samoan
matai, who used traditional oratory as a public vehicle for airing out problems
and reproducing the social system that they considered necessary for finding a
viable solution.
It was at this time that I was also introduced to the writings of Mikhail
Bakhtin and Valentin N. Vološinov. Their influence is apparent in my 1984
paper. I tried to reframe Samoan oratory in dialogical terms, arguing that no
individual was seen as able to control the meaning of what was being said.
Through the lenses of dialogism – and with some help from conversation
48 The avoidance of intentional discourse

analysis (Sacks, Schegloff, and Jefferson 1974) – I became interested in the


contexts in which the format of long speeches in the fono, which I had
analyzed in terms of “macro-turns” (Duranti 1981a), broke down to become
almost a conversation (Duranti 1990, 1992a, 1994). I return to this topic in
Chapters 5 and 6. In this chapter, however, I have decided to leave out Bakhtin
and dialogism in order to sharpen the argument and leave room for more
details of the case.
While a postdoctoral fellow at LCHC, I also became aware of Husserl’s
writings, after I heard his name mentioned by Hugh (Bud) Mehan, a former
student of Aaron Cicourel’s and another member of LCHC at the time, who
had discovered phenomenology via ethnomethodology (Mehan and Wood
1975). Over the years and especially in the last decade, Husserl has become
a frequent companion of my reflections and an important figure in my teaching
and writing, as shown in later chapters.

3.2 A brief introduction to the Samoan fono as a speech event and


a social drama
All the examples of oratorical speech in this chapter are taken from transcripts
of a particular kind of Samoan cultural activity called fono, which I analyzed
first in my dissertation (published as Duranti 1981a) as a speech event, that is,
an extended, multiparty interactional sequence that is largely or predominantly
organized by and through speaking (Hymes 1972a).7 Given that the word fono
can also be understood as a generic term for ‘meeting’ in Samoan society
(Larkin 1971) and, as a verb, can mean ‘to gather, meet,’ I need to clarify that
the kind of fono I discuss in this chapter and elsewhere in this book is the
special convocation of a deliberative assembly of matai – comprising chiefs
(ali`i) and orators (tulāfale) – which, as typical of similar events in other so-
called “traditional societies” (Comaroff and Roberts 1981; Bloch 1975), acts
both as a court that deliberates on crimes or conflicts involving matai and as a
legislative body, which can make, ratify, and abrogate all kinds of policies or
laws (tulāfono) that regulate how individuals and groups must behave in the
community. The meetings I recorded in 1979, 1981, and 1988 were usually
dedicated to discussing a new problem or an existing or potential conflict
among members of the village council (also called fono8) or between one or
more member of the council and people in other polities (nu`u).
Even though the discourse organization of a fono discussion is, in many
ways, unique (cf. Duranti 1981a, 1981b), the speech genres used (Duranti
1983), the social relations among participants, and the types of strategic
interaction found in a fono are also characteristic of other cultural activities
in the social life of a Samoan village. In fact, given the importance of political
relations and actions in everyday life as I experienced it during my stay in
Samoan fono as a speech event and a social drama 49

Figure 3.1 Orators (on the left) and a chief (on the right) during a fono
(photo by A. Duranti, Falefā, `Upolu, Western Samoa, 1988)

Samoa, it is not surprising that Margaret Mead (1930) saw the fono as
emblematic of much of Samoan adult life.
Although only matai are allowed to speak in the fono I observed and
recorded, these events are rather “public” for a number of reasons. The first
is that there are more than a few matai present – anything from fifteen to thirty
in meetings I attended. The second reason is that fono take place in daylight
and in houses that have either no walls or wide and open windows (see
Figure 3.1), which makes the interaction partly visible and audible to people
in the vicinity, especially to the younger and untitled people (taulele`a) who
may be sitting in another section of the same house or in an adjacent space or
construction, ready to serve food or take orders. And thirdly, speakers’ verbal
(and non verbal) acts during a fono can be the object of later discussions or
conversations within and outside the confines of the meetings, thereby reach-
ing people who were not physically present at the time.
From the point of view of its relation to the flow of social life in the
community, fono are typically embedded in what we might call, following
Victor Turner’s (1974) terminology, a larger “social drama.” This term helps
us capture the fact that a fono is a highly antagonistic arena in which different
groups and individuals try to control one another’s actions and in so doing work
at maintaining a position of authority, defending their reputation, or acquiring
50 The avoidance of intentional discourse

more prestige and influence in the community. While stressing the need to
get along and “take care of the relationship” (see below), speakers tend to be
cautious. Thus, they project a humble persona and stay as vague as possible in
their assessments of the problem at hand. At the same time, a speaker may
choose to be forceful and direct in his phrasing and volume of his voice. This
may happen because someone is exercising his prerogative to play a particular
role in the proceedings or because of contingencies that make someone feel
entitled to explicitly accuse someone or defend himself or others.
A Samoan orator can gain prestige and material gratification by speaking on
behalf or in support of a powerful or wealthy party (e.g., a high chief, a
government official, a wealthy guest), but he may also get in trouble and risk
retaliation if the party he represents does something wrong or does not satisfy
the expectations of the other matai. This means that, from the point of view of
the responsibility for what is being said, in Samoa there are contexts in which it
is hard to keep the distinction expressed by Ervin Goffman (1981) through the
roles of “animator” (the one who speaks), “author” (the one who creates the
“text”), and “principal” (the one whose ideas, wishes, etc. are being represented)
(see Chapter 10).9 In a fono, key elements of any accusation or claim involving
such roles are the relationship between the speaker and the party the speaker
represents. The accused’s own understanding of the events or his personal
intentions (or his motivations) may not be mentioned at all in the discussion.
Within a fono, participants do not usually invoke good will as an excuse for
something that went wrong. Even though at times they might claim lack of
knowledge or question the authority of someone to bring a charge or mention a
breach of social norms, there is a tendency, even in making an accusation, to
focus on the consequences or effects of actions. This was noted by Shore, who
wrote:
[W]hen I questioned informants about the relative seriousness of different misdeeds,
their tendency was to base their evaluations on the results for the actor of the action
rather than on any intrinsic quality of the act. (Shore 1982: 182)

In addition, within a fono the local rules of etiquette or respect (fa`aaloalo)


toward people of high status are foregrounded and are often a factor in the
avoidance of direct confrontation over past events even when they are known
by everyone around. There are only a few individuals who have both the social
status and the personal prestige to bring forth charges and in some cases
directly question some other members about what they know or what they
did or will do. In the fono I attended and recorded in 1979, only the two matua
or ‘senior orators,’ Iuli and Moe`ono,10 had the positional roles to call a
meeting and then bring a charge against some party or lay out an issue that
needed discussion and hopefully a resolution – to be arrived at by consensus.
From a structural point of view, the two senior orators acted as mediators
Unintended consequences 51

between the nominal authority of the high ‘chiefs’ (ali`i) – which in the village
of Falefā are collectively referred to as `Āiga (literally ‘families’)11 – and the
larger group constituted by the lower-ranking orators (tulāfale) or speech-
makers (failāuga) whose role was meant to be limited to that of advisors. De
facto, however, the two matua at the time of my first stay in Falefā (in
1978–1979) played a major role in all fono decisions, while the high chiefs
were often physically absent from deliberations or not visibly invested in their
outcome. The aloofness of the holder of the high chief title Lealaisalanoa, for
example, follows a cultural model of the publicly displayed disposition or
stance of a high chief.
Given the fact that each member of the fono as an assembly of matai is the
representative of one or more branches of an extended family (`āiga), special
attention must be paid to what Goffman (1955) called “the face” of the
participants, all of whom are considered worthy of respect (fa`aaloalo)
(Milner 1961; Duranti 1992b). Although typically several members of the
assembly may take the floor on any given issue, their opinion is often left
vague or ambiguous as they deliver speeches full of metaphors and proverbs
that express support for traditional values like getting along or caring for each
other (fealofani), protecting the dignity (mamalu) of the high chiefs, and
respecting the traditional ways of living (fa`asāmoa), while searching for the
truth (sa`ili le mea tonu) or taking care of the relationship (teuteu le vā)
between individuals, families, or villages.
Most typically, the matai gathered in a fono try to deal with the conse-
quences of one of them having partaken in a particular social act that was not
fulfilled, e.g., a public commitment to doing something, or which had an
unfortunate outcome, e.g., a political defeat or a loss of face. In what follows,
I will illustrate how this actually works by analyzing a series of verbal
exchanges within one of the fono I recorded in 1979.

3.3 Unintended consequences: Loa’s case, April 7, 1979


All of the fono I audio-recorded and studied during my first year of fieldwork
in the village of Falefā, on the island of `Upolu, originated from concerns and
conflicts regarding the February 1979 general election (pālota, from English
ballot) for the (then “Western”) Samoan Parliament. One major issue before
the election was whether the matai from Falefā should support, as they had
originally announced, the incumbent Member of Parliament (MP), Fa`ama-
tuā`inu, a matai from the nearby village of Lufilufi (see Figure 3.2) or one of
their own matai.12
The fact that three matai from Falefā had decided to run for office against
the incumbent made things more complicated. In a lengthy discussion of the
second option during the fono held on January 25, 1979, some matai argued
52 The avoidance of intentional discourse

South Pacific
Savaii Ocean

Malua APIA
Faleolo

Upolu

Lufilufi

Falefa

ITU ANOAMAA Sauago Saletele

Falevao

A T
U
A

0 25 50 75 100 km

0 10 20 30 40 50 miles

Figure 3.2 Map of section of Anoama`a East where Falefā is located, in the
Atua district, of which Lufilufi is considered the capital and historical center
(adapted from Krämer 1902: 704–705)

that having three candidates from Falefā would split the votes and let the
incumbent (from Lufilufi) win. Example (1) shows how the point was sum-
marized in the opening speech of the senior orator Moe`ono, one of the
candidates, who also acted as the chairman in the meeting.
Unintended consequences 53

(1) (Fono, January 25, 1979; brief section of Senior Orator Moe`ono’s
introductory speech about the situation)

419 Moe`ono; `a kākou ō ko`akolu, (1.0)


if the three of us run (lit. ‘go’), . . .
420 `ua mālō Lufilufi.
‘Lufilufi will have won’
421 ?; mālie!
‘well said!’

Another example is provided in (2), an extract from a long and rhetorically


complex argument presented after Moe`ono’s speech by the orator Usu, who
is more explicit at building up an argument in favor of two candidates instead
of three and hints to the youngest of the three, the chief Savea Sione,13 that if
he loves the village, he should step aside and let the two matua or ‘senior
orators’ Moe`ono and Iuli (whom he affectionately calls toea`ina ‘old men’)
run. Rather than as an inner feeling, what is here translated with the verb
‘love’ corresponds to the Samoan word alofa (in line 755) that must be
understood as a disposition toward others that needs to be externalized
through actions.
(2) (Fono, January 25, 1979; excerpt from the speech delivered by the orator
Usu, from the subvillage of Saleapaga in Falefā)

751 Usu; lau afiogae. Savea. . . .


‘oh14 your highness, Savea’
752 `ia `a e alofa,
‘well if you love,’
753 vaku le pa`ia o `Aiga,
‘pardon me honorable chiefs,’
754 Savea; lau fekalaiga.
‘you may say so’ (lit. ‘your honorable speaking’)
755 Usu; ia `ē alofa
‘so (if) you love’
756 ia kākou sa`ili mālō.
‘let us all seek victory.’
757 sa`ili mālō.
‘seek victory.’
758 `ā ko`alua lava kākou,
‘if we (have) just two (candidates),’
759 i koea`iga gei,
‘with these two old men’ (i.e., ‘the two senior orators’)
760 vaku ā lo lā mamalu.
‘(may) the two honorable (orators) pardon (me)’
761 kau ke suipi.15
‘we win (the game).’
54 The avoidance of intentional discourse

Subvillages :
1. Sagapolu
2. Saleapaga
fono falefā
fono falelima 3. Gaga ¢emalae
fono falefitu 4. Sanonu
5. Falevao
6. Sauago
7. Saletele

Figure 3.3 Three types of fono in Falefā, each defined by the number of
participating subvillages (from Duranti 1981a: 40)

Despite the apparently reasonable argument that either one or maximum two
candidates should run against the incumbent, three Falefā matai, the two senior
orators (Iuli and Moe`ono) and the chief Savea Sione, stayed in the race and,
predictably, none of them garnered sufficient votes to win the seat in the
Samoan parliament: the incumbent, Fa`amatuā`inu (from Lufilufi), was thus
reelected. I have discussed some of the details of this political scenario
elsewhere (Duranti 1981a, 1990: 476–478). In what follows I will concentrate
on an issue that was raised in a fono held two months after the election and was
connected to the election and was partly affected by the opinions expressed in
examples (1) and (2).
At the fono on April 7, 1979, there were two topics officially introduced in
the beginning part of the meeting (see Duranti 1994 for a discussion of the
difficulty of getting the potentially controversial agenda clarified in the
opening speech). One had to do with the two subvillages of Sauago and
Saletele, in Fagaloa Bay, the residents of which were accused of lackluster
participation in the so-called fono falefitu, that is, ‘meeting (fono) of the seven
(fitu)’ (see Figure 3.3).16
The second topic was about a suit that the young chief Savea Sione had filed
against the MP Fa`amatuā`inu. Moe`ono tried to convince him to drop the suit
because it would affect the relationship (vā) between Falefā, where the meeting
was taking place, and the MP’s village, Lufilufi. The discussion of the first
topic was very animated and confrontational, ending with an apparent recon-
ciliation between Moe`ono, who brought the charges, and the two senior
representatives from the villages in Fagaloa Bay. The second topic was less
confrontational but with no definitive resolution. In fact, the matai agreed to
meet again after nine days (on April 16) at which point they would know what
Chief Savea had decided. On the same day, they would all go to the nearby
village of Lufilufi to meet with the MP on his own ground in hopes of
straightening the situation once and for all (see §3.4).
After the discussion of these two controversial topics, one of the two senior
orators, Iuli, introduced a new agenda item. He proposed to fine the orator Loa
for having announced, a few weeks earlier, that the newly reelected district
Unintended consequences 55

MP, Fa’amatuā`inu (shortened to ‘Inu’ and pronounced /Igu/ in the fono


discussion), was going to come and present some goods to the village matai
gathered in an assembly. Since the MP had not come, Iuli argued that Loa
should be heavily fined, perhaps even expelled from the village. This accu-
sation and the ensuing discussion constituted the central case for my original
discussion of intentions in Samoan discourse and will remain a key piece of
evidence in this chapter.
Here is an excerpt with the first part of Iuli’s speech in which he introduces
the issue.
(3) (Fono April 7, 1979, notebook 3, p. 81)

3362 Iuli: `o le makā`upu gei e uiga iā Loa . . .


‘this topic is about Loa . . .’
3363 Loa; mālie!
‘nicely said!’
3364 Iuli; kusa `o le aso ga pokopoko ai lo kākou gu`u
‘regarding the day our village gathered’
3365 e fa’akali le faipule . . .
‘to wait for the MP . . .’
3366 `o mea fa`apea `o se luma o se gu`u.
‘such things are a humiliation for a village.’
3367 (pause)
3368 Loa; mālie!
‘nicely said!’
3369 Iuli; ma e:- `ua ka`uvalea lo kākou gu`u . . .
‘and- our village is ridiculed . . .’
3370 `ua fiu le kākou gu’u e kakali . . .
‘our village was tired of waiting . . .’
3371 le ai se faipule e sau . . .
‘the MP did not come . . .’ (lit. ‘there was no MP who came’)
3372 `ae se’i gofogofo Loa
‘but Loa went on sitting’
3373 alu amai se mea e kaumafa ma le gu`u
‘(instead of) going to get some food for the village’
3374 (pause)
3375 Loa; mālie!
‘nicely said!’
3376 (pause)
3377 Iuli; `o lo`u lea kalikoguga, (. . .)
‘this is what I believe.’
3378 (ka)kau ga sala Loa. (. . .)
‘Loa should be fined.’
[. . .]

As shown above, Iuli lays out his arguments right away for his request for a
fine to be imposed: Loa created a situation that was a humiliation (luma) for
56 The avoidance of intentional discourse

the village. The matai were ridiculed (/ka`uvalea/, lit. ‘called stupid’) by
having to wait for hours for gifts that did not materialize. In describing what
he sees as Loa’s too casual attitude, Iuli uses the verb nofonofo (here pro-
nounced /gofogofo/), a reduplication of nofo ‘sit’ that implies a prolonged and
repeated action or, rather, in this case, the prolonged absence of action.
According to Iuli, Loa just “sat and sat” instead of getting up and going to
buy some food himself to distribute to the matai who had been waiting for the
MP to come with gifts. In expanding his argument, Iuli adds that Loa and the
MP are related.17 This statement could have two meanings. One is made
explicit, namely, that Loa had more direct access to the MP and thus more
opportunities to get clarification on whether he was going to come or not. The
other is implicit, namely, that, given his family connections, Loa is even more
responsible – the practical logic here being that one is supposed to help family
members and, if necessary, make up for their faults.
(4) (Fono April 7, 1979)

3388 Iuli; `ae kakau oga fai maukigoa


‘but (Loa) should have made sure’
3389 auā e `āiga Loa ma- ma Igu, (. . .)
‘because Loa and- and (the MP) Inu are related’ (lit. ‘are
family’)
3390 `āfai `ua fai age iā ia,
‘if (he) said that to him,’
3391 sa kakau oga makuā ma`oki (. . .)
‘(it) was necessary to make very clear’
3392 gi mea e fa`akaumafa ai le gu`u
‘what to feed the village with’
3393 ga- ga amai lā`ea
‘to bring then’
3394 pe`ā lē sau le faipule(.)
‘if the MP doesn’t come.’

In a last and dramatic part of his speech, Iuli goes as far as proposing a heavy
fine (a whole cow and 100 Samoan dollars) and even expelling Loa (/alu `ese
ma le gu`u/ ‘go away from the village’).
After Iuli has concluded his accusatory speech, Teva, a chief from Fagaloa
(see Figures 3.2 and 3.3), briefly takes the floor to announce that he needs to
leave in order to catch the bus. Before getting up, he provides his opinion on
the case. Referring to Iuli’s speech, he says that either a fine or forgiveness
would be acceptable. This is a respectful way of acknowledging Iuli’s
concern while simultaneously mentioning an option that might avoid the
complications associated with establishing a fine and making sure that some-
one pays it – Teva does not even mention the more severe punishment of
expelling Loa from the village. As Teva is leaving, the orator Fa`aonu`u, who
Unintended consequences 57

was not present when the events recounted by Iuli took place, takes the floor.
He asks for clarification. Is Iuli saying that Loa lied to the village (/`ua fai se
pepelo i lo kākou gu`u/ lit. ‘has a lie been said to our village’)? What did Loa
do? Iuli immediately responds to Fa`aonu`u’s request providing a second and
more detailed narrative account of Loa’s actions including the consequences
for everyone else. It starts with Loa summoning the matai with the news that
the newly reelected MP will deliver (momoli) presents – a conventional
gesture by a newly elected representative. Iuli says that Loa’s announcement
prompts the arrival of matai from all seven subvillages or falefitu (here
pronounced /falefiku/) (see Figure 3.3). After having gathered (line 3470),
the matai become fed up (fiu) with waiting for the MP (line 3473) who does
not show up (line 3474). In the meantime, Loa is described again as sitting
around (/gofogofo/) and then as strolling (savalivalia`i) and going around
(fealualua`i) (lines 3475–3478), two other repetitive and heedless actions that
are meant to demonstrate that Loa does not worry (popole) about what is
happening (/e le popole i le mea lea/). In this second and expanded narrative,
Iuli underscores the gravity of the situation created by Loa by providing
details that are designed to elicit sympathy for those who are waiting,
including the untitled men (taulele`a) who had baked some food (suāvai)
to accompany what the MP was expected to bring (e.g., canned meat and
fish, and a cooked pig), and the matai who came all the way from Fagaloa
Bay (line 3479). In this case, the road (`āuala) is described as ‘difficult’ (/
faigakā/) for the orators and chiefs, who are not directly mentioned but
evoked through the use of the honorific terms maliu and afio, two verbs that
respectively index actions by orators and chiefs.
(5) (Fono April 7, 1979)
3466 Iuli; `o le fā`aliga, lau kōfā . . .
‘as for evidence,18 your honor. . .’19
3467 `o Loa `ua sau kala`i le kākou gu`u
‘Loa came to boast to our village’
3468 e fogo- ma pokopoko
‘to meet- and gather together’
3469 la`a sau le faipule- e `amai20 oga momoli . . .
‘the MP will come- to bring his treat (of food) . . .’
3470 ia` oga- pokopoko lea o le kākou falefiku,
‘so our seven subvillages gather,’
3471 leai se isi e o`o iā Fagaloa ma Falevao
‘there is no one who stays behind in Fagaloa and Falevao’
3472 ma kagaka `uma o le kākou- gofoaala . . .
‘and all the people of our- subvillages’ . . .
3473 pokopoko `ua fiu `ua alu legā aso `o kakali
‘gathered together (we are) fed up, the day has gone still
waiting’
58 The avoidance of intentional discourse

3474 leai se faipule `o sau ma se mea . . .


‘no MP (in sight) who comes with something . . .’
3475 `ae se`i gofogofo
‘but just sitting (all the time)’
3476 fu`e mai fo`i le suāvai a kaulele`a . . .
‘the food from the oven made by the untitled men is
ready . . .’
3477 ia` savalivalia`i,
‘well (he) walks around’
3478 fealu- fealualua`i Loa pei e lē- e le popole i le mea lea. (. . .)
‘Loa wand- wanders around as if he doesn’t-doesn’t worry
about that thing . . .’
3479 `ae faigakā Fagaloa ma o lākou `āuala mai (. . .)
‘but the road for those who come from Fagaloa is hard . . .’
3480 `ua kala`i ma- maliu mai ma afio mai. (. . .)
‘(matai) have arrived and- (orators) and (chiefs) . . .’
3481 ogo `o kala`iga a Loa. (. . .)
‘because of Loa’s summoning . . .’
3482 e lē se mea mamā lā i lo`u magaku. (. . .)
‘in my mind this is not a light thing . . .’
3483 e lē faigofie fo`i (. . .)
‘it is not easy either . . .’
3484 auā `ā lē faia, (. . .)
‘because if it is not dealt with,’
3485 `o la`a fa`apea fo`i se isi- se isi aso
‘there will be someone-else-some other day’
3486 `a fa`asala aku fo`i, (. . .)
‘when getting fined,’
3487 ia` oga ka`u mai lea `o Loa ma le mea
‘then Loa and the thing’
3488 `ua ga faia i kokogu o le gu`u.
‘(that he) did in our village will be mentioned.’
After this clarification, the orator Fa`aonu`u speaks again, this time asking Iuli
to forgive Loa. Then, addressing Loa directly, Fa`aonu`u tells him that he
(Loa) got himself in a strange situation (/`ese fo`i lā kūlaga gā `ua `e iai/), by
not hiding the fact that he and the MP are related.
At this point, the other senior orator, Moe`ono, takes the floor. In his
remarks, Moe`ono makes it clear right away that he does not want to get
involved in Iuli’s complaint about Loa. He brings up a procedural problem,
namely, that Iuli had not properly brought up the issue at the beginning of the
meeting21 and that he rushed things by proposing a fine without letting the
other representatives of the different subvillages give their opinion about the
issue and discuss whether it is appropriate to impose a fine. Moe`ono says,
‘this is not the way in which we do things in our village, Iuli’ (/e le fa`apegā le
faiga o le kākou gu`u Iuli/).
Unintended consequences 59

Iuli then takes the floor again and backs down by declaring that there will
not be a fine if the rest of the village does not want to impose it. Moe`ono takes
the floor again. This time he first takes the opportunity to scold the orator
Fa`aonu`u for expressing his opinion regarding one of the cases that was
discussed earlier and which had to do with a chief, Savea Sione, who is from
Fa`aonu`u’s own subvillage. This is framed as a lesson in political etiquette
(with Moe`ono telling the lower-ranking Fa`anu`u how to behave in a fono),
but it can also be read as a complaint or even a warning given that during the
earlier part of the meeting Fa`aonu`u had expressed an opinion that went
against Moe`ono’s own request that Savea withdraw his suit against the MP
Fa`amatua`inu, that is, the same person who never showed up with the food for
the matai in Falefā. Then Moe`ono returns to Loa’s case, in direct disagree-
ment with Iuli, who is roughly his equal, puts aside the argument made by Iuli
that Loa is responsible because he and the MP are related, and focuses on what
he thinks Loa should have done, namely, first ask the MP or someone else
from his village to come to Falefā (referred to as ‘the seaward village’) and
then to stop continuing to say that the MP was going to come.
(6) (Fono April 7, 1979; speaker: Senior Orator Moe`ono)

3585 Moe`ono; pe `āiga pe lē `āiga,


‘whether or not they are related,’
3586 kasi ā le mea `ua sesē Loa, (. . .)
‘one is the thing that Loa did wrong . . .’
3587 sā kakau ā ga fai iai maliu pea ā `oe i kai. (. . .)
‘he should have told him “come to the seaward
(village)” . . .’
[. . .]

Moe`ono is also the first to implicitly blame the MP himself by saying that he
is just ‘a little kid who went to school with foreigners’ (/`o le kamaikiki sā
a`oga i ō pālagi/) who therefore does not know anything about the tradition
(/ga ke lē iloa gi agagu`u/), which implies that he does not know how to behave
(and be generous). Moe`ono concludes by proposing to wait until they all go to
meet with the MP and the other matai in Lufilufi. That will be the time to bring
up the issue.
How to interpret these exchanges and the arguments presented so far? In
my earlier discussions of the case (e.g., Duranti 1988), I pointed out that,
contrary to what one would expect in a western judiciary context, Iuli’s
accusation and the responses by the other matai are almost exclusively
focused on the consequences of Loa’s announcement of a visit by the MP
that did not materialize and his apparent lack of concern for the situation
that his reported speech had created. None of the matai who speak, includ-
ing Loa himself, say that Loa should not be accountable for reporting what
60 The avoidance of intentional discourse

the MP had told him. If the MP’s reported words are treated as a “promise,”
as apparently made explicit in Moe`ono’s speech to the MP nine days later –
see example (12) below – it is striking that the discussion never comes
down to saying that the MP violated his promise and for making him
responsible for the embarrassment the lack of fulfillment created. Despite
the centrality of the speaker’s intention in the conditions of satisfaction for
the speech act of promising as described by Searle (see §2.1.1), the MP’s
intentions are not mentioned or questioned during the April 7, 1979
meeting. No one wonders whether the MP meant what he said, but it is
Loa and not the MP who is said to have caused the inconvenience of
important people and contributed to their public loss of face. Even
Moe`ono, who, as we saw in (6) above, dismisses Iuli’s contention that
the family relationship between Loa and the MP is an aggravating circum-
stance, does not challenge Iuli’s accusation by bringing up the fact that Loa
might not have known that the MP was not going to come and therefore that
Loa was not responsible. Even though here, as in the two earlier cases
discussed on that day, Moe`ono is critical or even disdainful of the MP, he
agrees with Iuli that Loa did something wrong (sesē): Loa did not stop
delivering the MP’s message (/`a e le ku`ua e Loa ia le fā`aliga a le Faipule/)
once he realized that the MP was not coming.

3.3.1 Loa’s defense


And what about Loa? Does he defend himself? And with what arguments? As
shown in excerpt (3), during Iuli’s first accusatory speech Loa occasionally
reacts with the conventional agreement mālie! ‘nice!’ or ‘nicely (said)!’ to
some of the points made by Iuli, even when they constitute serious allegations
about his (Loa’s) behavior. Following the fono protocol, Loa lets other matai
(the chief Teva, the orator Fa`aonu`u, and the senior orator Moe`ono) speak
before taking the floor and, in fact, he speaks only after the issue has been put
to rest by Moe`ono’s request to wait until they all go to Lufilufi to meet with
the MP and the other matai from his village. When he finally takes the floor,
Loa acknowledges the previous speakers, a conventional opening of any fono
speech, and then offers to clarify the “other issue” which is understood to be
Iuli’s case against him. Loa starts out with the statement that he has not
betrayed the village nor did he make fun of it (lines 3626–3628) and only a
fool would (dare) deceive the council (this line is left out). Loa’s statement that
it was the MP who said ‘I’m coming to summon our village’ (lines
3638–3639) could be understood as an indirect indictment of the MP but in
fact, as suggested by the subsequent statement in line 3640, seems to be a
defense against the accusation of having lied about the MP’s announcement of
his visit with gifts.22
Unintended consequences 61

(7) (Fono April 7, 1979)

3626 Loa; `ia `ae pei o:- `ou ke lē olegia lo kākou gu`u. . . .
‘well but I do not deceive our village. . . .’
3627 Iuli, (. . .) e lē- e lē iā ke a`u fo`i segā kūlaga,
‘Iuli . . . it’s not- that sort of action is not me,’
3628 `ou ke kaufa`avalea po `o le ā se isi kūlaga, . . .
‘(that) I try to fool (people) or something else . . .’
[. . .]
3636 `ua po`o iai gi kuagia o le faipule
‘the MP might have had other commitments’
3637 `ua mafua ai oga lē kaugu`u mai. . . .
‘that explained why he did not arrive . . .’
3638 `a `o ia ā ga sa- ga saugoa mai
‘but it was he who said (to me)’
3639 `ou ke sau e fofoga lo kākou gu`u. . . .
‘I come to summon our village’
3640 e lē fa`apea lā `ou ke faku fua se kalapelo. . . .
‘it is not the case that I made up a lie for no reason.’

Loa also provides a brief narrative account of himself going to the (public)
phone to make a call to find out the reason (māfua) why no one was coming
and adds that he had the inclination (agāga) to bring something to eat for the
matai waiting for the MP but arrived too late: the matai had already had the
food cooked in the oven (suāvai) for the day (line 3657). Loa concludes with a
general statement about his respect for the village – which here means for the
other matai – and a note of caution as to whether the issue makes sense. These
remarks should be interpreted in the context of an accusation that focuses on
the embarrassment that Loa’s actions or lack thereof caused among the matai
in the village. Reaffirming his respect for the assembly and its members is a
way for Loa to reassure his audience that he cares about the consequences of
his actions. Loa’s self-defense shows that Loa is trying to address both his
responsibility for what happened and the truth of his announcement that the
MP was going to come.
When we examine his narrative, we see that there are few factual details. For
example, he says that he came to make the phone call – the phone booth was
relatively far from his house and close to the house where the fono is taking
place – to find out why the MP was not coming. But in the telling he switches
from the singular ‘I came’ (/go`u sau/) to the plural ‘the two of us came’ (/mā
ōmai/) without mentioning who the other person was. Then he never says
whether he actually made the phone call and reached anyone (I do know that at
that time the phone was often out of order so it is possible that he was not able
to complete the call). The nature of verbal interaction in the fono, which favors
long turns-at-talk as opposed to question–answer sequences, does not favor the
62 The avoidance of intentional discourse

display of the details that might seem necessary to get a fuller account of what
actually happened (I will return to this point in Chapter 5).

3.3.2 The relevance of positional roles


One way of making sense of Loa’s responsibility in the events of this case is to
remember Loa’s status, rank, and positional role with respect to the village of
Falefā and in relation to the MP, who holds the title Fa`amatuā`inu in the
village of Lufilufi, the capital of the Atua district (see Figure 3.2).
In Falefā, Loa’s orator title comes from the subvillage of Sanonu, which is
ranked lower than the other three subvillages (see Figure 3.3) but has some
special functions in village affairs, among them the right to determine the exact
fine that someone must pay after having been condemned by the members of
the fono and the right to decide how to distribute any food that is brought to a
formal gathering of matai. In excerpt (8) below, we can see that this role is
mentioned by the orator `Upu (from Falevao) who takes the floor after Loa’s
self-defense. In lines 3678–3679, `Upu explains that the title of “Lautogia”
(here pronounced /Laukogia/) also comes with a duty. If something goes
wrong, that is, if there is no food, the Lautogia must go and find the food to
feed the matai of the village.
(8) (Fono April 7, 1979)

3675 `Upu; ia `o le si fo`i vāega


‘so, as for the other part (of the discussion)’
3676 e kakau lava oga magakugaku iai iā- `oe le Laukogia, . . .
‘one should have reflected upon, you the Lautogia . . .’
3677 se `o lou kou pikogu`u e- e pulea kaumafa a la kākou gu`u.
‘it is your subvillage (that) controls the food of our village’
3678 `a fa`alekogu le aso, . . .
‘if anything goes wrong (on) the day . . .’
3679 `o `oukou ā fa`akau ma faga le gu`u.
‘you all are the ones who buy and feed the village.’
3680 Loa; ((laughs softly)) huhu.
3681 `Upu; lea sā iai le māfaufau fa`apegā i legā aso.
‘that was the thought on that day.’

Loa also has a responsibility toward the MP Fa`amatuā`inu, who holds the
highest government office in the district. As a lower-ranking orator and a
member of the council of matai in Falefā, Loa is expected to support and
help Fa`amatuā`inu. Loa’s kin relation with the MP can only add to such an
expectation. In fact, the kind of situation in which Loa finds himself,
namely, being blamed for something that should be seen as someone else’s
fault, is not uncommon. Lower-ranking orators might be blamed for
Absolute liability 63

something that might have been initiated, condoned, or even mandated by a


matai of higher rank. From this perspective, Loa is guilty regardless of his
intentions.

3.4 Absolute liability, cultural preference, and empirical evidence


I will now first consider and then dismiss the interpretation of this case as
providing evidence for the claim that Samoans – or some of them – subscribe
to the legal doctrine that jurists (and legal anthropologists) call “absolute (or
strict) liability.” This is a way of assigning liability – answerability to law, if
not necessarily actual responsibility – based solely on the effects or conse-
quences of someone’s actions (or omissions). The actor’s intentions are not
relevant, and even the idea of “accident” does not block the assignment of
responsibility. This is how Goldman (1993) interpreted Ochs’ (1982, 1988)
brief mention of legal arenas in the midst of her discussion of the Samoan
tendency to ask non-specific types of clarification questions (e.g., “What?”
“Where?”) instead of guessing what a child meant to say.
Discourse in the Samoan legal arena is oriented “almost exclusively” (Ochs 1988: 141)
towards the social consequences of utterances and actions; distinctions prevalent in
Western middle-class societies between inadvertent, accidental, intentional, and
planned acts are “not terribly important” (ibid.). In other conversational contexts, this
reluctance to talk about “mind” is, according to Ochs, indexed by clarification proced-
ures where, in the face of unintelligible utterances, speakers will elicit a repeat perform-
ance rather than guessing at what was said. Moreover, other reflexes of this cultural
disposition are the noted absence of test questions, riddles, and verbal games based on
guessing routines. This, then, is a society . . . with AL [Absolute Liability] . . . (Goldman
1993: 284–285)
The first problem with this interpretation is that neither Ochs nor I (Goldman
also mentions my 1988 article about intentions) ever made a claim of
Absolute Liability, and Ochs repeatedly used the notion of preference
throughout the book, a notion that is quite distinct from anything “absolute.”
Despite such misinterpretation, however, Goldman makes two types of con-
tributions to the investigation of intentional discourse: (a) he called for an
empirical study – at the level of lexicon, grammar, and genres – of how
people actually display their orientation to intentions, and (b) he showed that
different contexts show different orientations by participants toward inten-
tions and conventions, a point subsequently supported by Justin Richland
(2006).
One way to further pursue the empirical investigation I initiated in the 1980s
and continued in this chapter is to add the analysis of the Samoan expressions
that might encode a concern for causal relations between events (speech acts
included).
64 The avoidance of intentional discourse

3.4.1 Causes, reasons, and sources


When we look at Samoan adult speech in the context of the fono we find a
general avoidance of speculation about individual intentions or motives for
past actions, which results in a tendency to produce only vague and ambiguous
allusions to past events. At the same time, Samoans do have expressions for
what we would call ‘reasons’ or ‘causes’ of events and behaviors. In the fono
I recorded, there are a few examples of words like the verb māfua translated by
Milner (1966: 120) as ‘originate from, be caused by’ and the corresponding
nominalizations māfuaga and māfua`aga that can be translated with the
English nouns ‘cause,’ ‘reason,’ and ‘source.’ In (10) below, we see the senior
orator and chairman Moe`ono explain the ‘reason’ or ‘cause’ for the meeting (I
will later return to this exchange to discuss how Moe`ono frames the rest of his
explanation):
(10) (Fono January 25, 1979; first topic of the day: attendance in the fono

86 Moe`ono; `ou koe f- koe fola mālamalama aku fo`i le māfua`aga


I again d- again display understand Dx-out also Art origin
‘I am again- again clarifying the reason/cause’
87 `ua mafua ai oga kākou fa`apegei. (1.0)
‘(that) has caused us all to be like this [i.e. gathered
together]’
or
‘(that) has been the source of our being like this [i.e.,
gathered together]’
88 auā `o fogo `ua kuaga`i. (1.0)
because Pred meeting Perf pass
‘because past meetings’
89 ia` e le `o-
well it’s not-
90 e le`o `ākoa pokopoko lo kākou gū. (2.6)
‘it is (was) not the whole village that was gathered . . .’
[. . .]

The verb māfua in line 87 and its derivation māfua`aga contain the word fua,
which means ‘fruit, flower, bloom,’ suggesting a semantic blurring between
‘reason, cause’ on the one hand and ‘origin, source’ on the other.23 This is a
temporal interpretation of causation that follows the same logic captured by the
Latin phrase post hoc ergo propter hoc, meaning that an event following
another is seen as caused by it. There are also in Samoan spatial metaphors
for representing reasons and causes as shown by the use of such terms as ala
‘path’ and `auala ‘road’ (see Milner 1966: 31). An example of the use of
`auala ‘road’ meaning something close to our ‘reason’ is given in the next
excerpt, in (11). Here `auala might express what in English could be conveyed
Absolute liability 65

with the colloquial phrase ‘how we got there.’ The context is an earlier part of
the same April 7 meeting discussed so far, and the orator Loa is being scolded
by Senior Orator Moe`ono for not making sure that the news of the convoca-
tion of the meeting be delivered (pagi) in time to the subvillages of Sauago and
Sauatele to come to a fono falefitu (see Figure 3.3 above).
(11) (Fono April 7, 1979; during the opening speech the senior orator Moe`ono
inquires about the absence of the matai representing the two nearby
subvillages of Saletele and Sauago in Fagaloa Bay)

783 Moe; Loa, `aua le ta`alo i le pagi.


Loa don’t Art play Prep Art message
‘Loa, don’t fool around with the delivery (of the message).’
784 `o le pagi `o le afioga ma so mā kāofi
Pred Art message Pred Art decision and ArtþIn we-Du
opinion
‘the message is the decision [by the chiefs] and the decision
of the two of us (exclusive) (i.e., Moe`ono and Iuli)’
785 (e-) e kau momoli.
‘to be delivered.’
786 (4.0)
787 Loa; ga malaga iai Faikamakau
‘Faitamatau traveled there (i.e., to the two subvillages).’
788 ia` `a `ua leiloa lā po `o le ā le auala
well but Perf not-know there Qu Pred Art which Art road
‘well, but it is not known what the reason is’
789 e lē fa`apea e iai se isi e ka`alo (0.5) i le afioga.
‘it’s not that anyone is fooling around . . . with the decision
(by the chiefs)’
790 (1.0) `ua fa`akaugu`uiga ā le afioga.
‘. . . the decision was followed through.’

examples (10) and (11) show us that Samoans are indeed concerned with the
origin or cause of certain states-of-affairs especially when they have negative
consequences. In (11) we see the orator Loa interpreting Moe`ono’s scolding
as an indirect request for him (Loa) to provide a justification for the failed
communication between subvillages. At the same time Loa’s response shows
reluctance to either offer or find the ‘reason’ (literally ‘the road’ in line 788) for
the current state-of-affairs. This is not unusual. In the fono most speakers tend
to stay as vague as possible when uttering statements that could be interpreted
as accusations. Even though, as shown above, in the fono there may be concern
for reasons, causes, and explanations, the discussions of such reasons or causes
are not very elaborate. They tend to be formulated in terms of generic or
stereotypical behavior, like, for example, being under the influence of alcohol
and thus not in control of one’s behavior. For example, in the April 16,
1979 meeting held in the village of Lufilufi, Moe`ono confronts the MP
66 The avoidance of intentional discourse

directly about his folafolaga, which was translated as ‘promises’ by a bilingual


native speaker in 1979 but could also be rendered with ‘announcements’ or
‘the things you announced’ (see Chapter 4 about these translations). In so
doing, Moe`ono takes for granted that the reason for the MP’s irresponsible
behavior is due to the fact that he had been drinking beer, which is understood
as meaning that he must have been drunk. Here is the relevant passage.
(12) (Fono April 16, 1979; Lufilufi, capital of the district of Anoama`a East and
residence of the family of the MP Fa`amatuā`inu)

Moe`ono; `o fea `au folafolaga?


‘Where are your promises?’
(or ‘(the things that) you announced’)
oi ke `ole`ole- e ke `ole`ole `Āiga?
‘oh! did you betray- did you betray (the dignity of) the Chiefs?’
fea `au folafolaga
‘where are your promises?’
(or ‘(the things that) you promised/announced’)
ga `e susū aku
‘(saying) that you will come’
fa`akali le gu`u iā `oe!
‘the village waits for you!’
fa`akali `Āiga ma mā`ua
‘the chiefs and the two of us24 (were) waiting’
`o fea `o iai
‘where were (you)?’
soia!
‘don’t!’
`aua le faia- faia iguga pia
‘don’t drink beer.’ (lit. ‘don’t do- do the drinking of beer!’)
Similarly, a few minutes later, in the same speech, Moe`ono characterizes the
scene when the MP must have told Loa to summon the matai in Falefa
(something Moe`ono is imagining) as an occasion that involved drinking beer –
although this time instead of the word pia (from English beer) he calls it `ava,
a metaphorical extension of the term for the ceremonial drink kava.
It is not unusual for the powerful senior orator Moe`ono to be the one who
mentions the cause of the problem. The mention of ‘reasons’ or ‘causes’ is
hierarchically skewed. In the fono I recorded in 1979, requests or offers of
reasons or causes tended to be made by higher-status individuals, with the
senior orator Moe`ono being a prime example. Lower-ranking individuals like
the orator Loa – as shown in excerpt (7) above – tended to avoid giving
reasons for past or present events or providing details of issues that may
involve higher-status individuals. In my book From Grammar to Politics,
I wrote about this tendency in discussing the vague quality of the official
announcement of the agenda of the fono. To make sense of it, I proposed a
Conclusions 67

strategy, “say the least,” which can be understood as a way to avoid the
problem of having to blame, accuse, or rebuke someone whom you are
expected to honor (see Duranti 1994: 116–121). Similarly, in the case dis-
cussed in this chapter, we saw that once he is accused by the senior orator Iuli,
the lower-ranking Loa finds himself in a difficult situation. He needs to defend
himself, but he has to do it in a way that does not directly challenge Iuli’s
authority and judgment.
These observations suggest that, in uncovering Samoan ethnotheories of
self, relationships, and social agency, the issue is not just whether Samoans
have words that hint at or entail individual will or planning, which they do, but
(a) in what contexts something corresponding to an individual intention is
made explicit and relevant and (b) whether the expressions used in such cases
correspond to the speech-act types described in the literature on English and
often implicitly assumed to be universal. To test this line of argumentation, in
Chapter 4 I will explore in some detail whether we could say that there is a
Samoan speech act corresponding to the English promise.

3.5 Conclusions
Overall, the data discussed in this chapter support the view previously argued
by a number of ethnographers that Samoans are more eager to act upon
conventions, consequences, and cultural expectations about what a certain
type of person (e.g., with a particular status or role) would or should do in a
given context than they are to rely upon an explicit reconstruction of what an
individual’s intentions or unexpressed goals might have been. The Samoan
matai in the fono I examined in this chapter avoided speculations about
intentions – or about motivations (Mead 1928) – as shown by the fact that
inquiries regarding why someone did something are rare and quite limited in
scope. More generally, I still feel confident saying that the Samoan speakers in
my recordings display a discursive dispreference for introspection and for
non-generi, that is, individual-specific psychological explanations of past
behaviors.
Given that human action, and speech as one type of such action, is goal
oriented, Samoans, like any other people in the world, must and do interpret
each other’s doings as having certain ends with respect to which those doings
must be evaluated and dealt with. The problem – for us, and, perhaps at times,
for them as well – lies in the extent to which, in interpreting each other’s
behavior, Samoans display a concern for an actor’s alleged subjective reality.
The fact that a society can carry on a great deal of complex social interaction
with little concern for people’s inner thoughts about what should be done and
with a much more obvious concern for the public, displayed, performative
aspect of language is, in my opinion, an important fact which any theoretical
68 The avoidance of intentional discourse

framework concerned with the process of interpretation should take into


account. We need a theory of the force of language – or pragmatics – that
can recognize not only speakers’ knowledge, needs, and wants, but also the
constraints that speakers have in conceiving, expressing, and realizing certain
types of social acts through language. I will return to these issues in Chapters 5
and 6. In the meantime, I need to acknowledge the fact that what I have
discussed so far is limited to so-called “institutional discourse.” In several
respects, the Samoan fono I recorded correspond to parliamentary debates or
court proceedings in the US or Europe, that is, events that have been shown to
have some different rules and conventions than ordinary conversation (Atkin-
son and Drew 1979; Heritage and Clayman 2010). What is needed then is an
examination of some conversational exchanges in Samoa. I will do so in
Chapter 8 with the explicit goal of identifying the manner and extent to which
Samoan speakers engage in reading other intentions or mind-reading outside of
institutional settings like the fono.
4 The invention of promising in the Samoan
translation of the Bible

To breed an animal with the right to make promises – is not this the
paradoxical problem nature has set itself with regard to man? And is it not
man’s true problem?
(Friederich Nietzsche, The Genealogy of Morals, Second Essay)

[Morality] has, at least politically, no more to support itself than the good will
to counter the enormous risks of action by readiness to forgive and to be
forgiven, to make promises and to keep them.
(Hannah Arendt, The Human Condition)

4.0 Is promising a human universal?


In the typology of speech acts proposed by Austin (1962) first and then further
developed by Searle (1969, 1976), promising was a key example of a commis-
sive – that is, a type of act whereby we commit to doing something (like the
acts expressed by the English verbs commit, threaten, vow, pledge, offer, bid,
warrant). Such commissives, as discussed in Chapter 2, have among their
“felicity conditions” (for Austin) or “conditions of satisfaction” (for Searle) the
speaker’s intention to do a number of things, namely, (a) carry out the act,
(b) accept responsibility for such an act, (c) be under the obligation of doing it,
and (d) convince the hearer that (a), (b), and (c) will be the case by using
conventional language that will allow the hearer to fully understand what the
speaker is up to.
The act of promising played an important role in Michelle Rosaldo’s (1982)
argument that speech act theorists were Anglo- and Eurocentric. Her claim that
among the Ilongots there was no notion of promise, or at least not in the sense
of an act whose conditions of satisfaction are in the privacy of people’s mind
or in their “psychological orientation” (1982: 212), was part of a more general
trend in the anthropology of the late 1970s and early 1980s. Whether they
subscribed to a view of social behavior as something to be explained by means
of larger socio-economic forces or something to be interpreted as public
displays and interpretations of cultural values (or conceptions of “person”),
most anthropologists did not believe that psychological or internal states could

69
70 Promising in the Samoan translation of the Bible

explain the social meaning of human actions. Thus, meaning of actions was
more likely to be seen in terms of contexts and observable or perceived
consequences of certain behaviors. The theoretical issue then for Rosaldo
was not so much the universality of the notion of promise (a claim that had
never been made by either Austin or Searle), but with the epistemological or
even ontological implications of the conceptual apparatus through which
speech acts were being described by analytic philosophers and, ultimately,
made sense of by speakers within particular speech communities.
I argue . . . that the act of “promising” is alien to the Ilongot repertory of kinds of
speech. More immediately relevant, however, is the question as to why, and with what
consequences, the act of promising has been used as a paradigm in theories presently
available. To think of promising is, I would claim, to focus on the sincerity and integrity
of the one who speaks. Unlike such things as greetings that we often speak because, it
seems, “one must,” a promise would appear to come, authentically, from inside out. It is
a public testimony to commitments we sincerely undertake, born of a genuine human
need to “contract” social bonds, an altruism that makes us want to publicize our plans.
Thus the promise leads us to think of meaning as a thing derived from inner life.
A world of promises appears as one where privacy, not community, is what gives rise
to talk. (Rosaldo 1982: 211)

As discussed in Chapter 2 (§2.3), when Searle finally responded to Rosaldo’s


critique in his 2006 article in the journal Anthropological Theory, he ignored
her argument against the view of promises as coming from inner life and
requiring sincerity. Instead he focused on the challenge to the universality of
promise, which he dismissed as an issue that was irrelevant to his analysis
(Searle 2006: 27). But Rosaldo’s objection to Searle’s conditions of
satisfaction for promising was not just based on an empirical counterexample
(you may have “promises” but my people do not). It was a more general and
fundamental critique of promising as a social act that (a) needs to take the
context (including people’s relationship) into consideration and (b) may not be
conceptualized as the expression of some inner state of mind. Such a theoret-
ical stance is already present in the passage above and is made even clearer in
the passage where Rosaldo criticizes the analysis of the use of promise in
English (see §2.3).
Rosaldo’s tragic death in 1981 while in the Philippines deprived us of her
brilliant intellect and of a follow-up in her own words. In this chapter, I take on
the challenge of continuing her line of argument by examining in some detail
whether there is a Samoan expression that would correspond to the English
promise. The availability of written texts, including the Samoan Bible, that are
between 100 and 150 years old provides us with a unique opportunity to apply a
method of study that has been applied by other linguistic anthropologists
examining cases of linguistic acculturation or language-mediated cultural
change (see §4.1.1 below). In joining the growing literature in this domain
Is promising a human universal? 71

of study, I will try to make my method as explicit as possible, thereby offering


another example of how to investigate linguistic material in documenting
cultural change. I will propose that promising was not a native Samoan speech
act at the time of European contact and was introduced in the Samoan repertoire
of speech acts by European speakers in the mid-1800s, probably in the context
of the translation of the Bible into Samoan. The evidence for this hypothesis is at
the moment circumstantial and yet consistent with what we know about the
translations of other terms found in the Bible and with other pieces of the
evidence in the historical record and in current linguistic practices.

4.0.1 Promise and promising in English


One of the common assumptions in the understanding of the term promise in
its Standard English use is that it is an act that most typically is made by one
individual to another. This interpretation is captured in the following definition
provided by The Shorter Oxford English Dictionary (SOED):
1. A declaration made to another person with respect to the future, stating that one will
do, or refrain from, some specified act, or that one will give some specific thing.
(SOED, p. 1597)

Searle’s (1969) conditions of satisfaction (corresponding to Austin’s “felicity


conditions”) of the speech act of promising (see Chapter 3) reflect this basic
understanding of promise when he states that for a promise to be properly
performed (a) the hearer would prefer the speaker to do what is predicated in
the promise (condition 4), (b) normally such a course of events would not
happen (condition 5), (c) the speaker has the intention to do the action
(condition 6), and (d) will have an obligation to do so (condition 7), and,
finally, (e) the utterance used by the speaker is a conventional expression
through which the speaker’s intention to carry out the promise can be recog-
nized by the hearer – this is what makes the meaning of the speech act “non-
natural,” in Grice’s (1957) sense of the term. In all of these conditions, the two
parties involved are understood to be two individuals, namely, a speaker and a
hearer. Of course, they do not have to be individuals, given that promises
between a person and a group or between two groups are possible, but the
unmarked or prototypical promise in English is conceived as an act that
involves (and requires) two individuals.
Regardless of whether Searle’s definition of the conditions of satisfaction
for the act of promising are valid or appropriate for the English practice of
promising (see Downie 1985; Vitek 1993; Jacques 2001/2; Habib 2008), the
specific meaning of a personal commitment sincerely made by one person to
another was not a native Samoan concept or activity when the missionaries
arrived in Samoa in the 1830s. The relevance for this claim in the context of
72 Promising in the Samoan translation of the Bible

this book is that Searle’s description of the speech act of promising, as


anticipated by Rosaldo, foregrounds (or does not sufficiently contextualize a
belief in) inner and private psychological processes that are supposed to
motivate and sustain an obligation into the future regardless of unforeseen
contextual circumstances. I believe that it is the private dimension of the
English promise – as shown by the fact that we can even say, for example,
I promised myself that . . . (see below) – that is absent (or not logically
necessary) in other types of speech acts that commit the speaker to do
something in the future. Promising is thus quite different from other kinds of
commitments, like, for example, making an oath, where the commitment is
largely established by the typically public quality of such an act. As we shall
see, the most common solution to translate the English promise in Samoan has
been to use a verb that describes a visible and typically publicly performed act.

4.1 The invention of promising


As in other parts of Polynesia and elsewhere in the world, the task of translat-
ing the Old and New Testaments into Samoan forced the missionaries and their
native consultants to introduce a number of lexical innovations not only for the
vast number of proper names that populate the genealogies and stories in the
Bible, but also for some of the most basic concepts of the Christian faith. In
addition to the Christian notions of spirituality and afterlife, and their associ-
ated moral values, there were other, more ordinary concepts and activities that
needed translations. One of them, I am arguing, was promising, an act that, as
I will show below, is found in a considerable number of passages in the Bible,
especially in the New Testament, but, in my reconstruction, was not an
indigenous concept or speech act in Samoa at the time of contact with
Europeans. My evidence is based on a detailed textual analysis of the Samoan
Bible and of other Samoan texts collected in the last decade or two of the
nineteenth century. I will show that the Samoan terms chosen for describing or
performing promises in the Bible have other more basic meanings, which they
have retained until today. Such meanings display a family resemblance that
can be made sense of by positing a temporal sequence of semantic and
pragmatic extensions that have been observed in other language families.
A major candidate for expressing promise in Samoan is a lexical item whose
(past and current) most common meaning is a type of physical act roughly
corresponding to the English verbs lay, put down, or unfold. These are all
actions that are available for others to directly experience through the senses
(e.g., vision, hearing). Similar semantic changes have been proposed by
cognitive linguists for Indo-European languages where verbs of sense-
perception or movements have been extended to describe mental activities.
For example, just as the contemporary understanding of intention comes from
The invention of promising 73

the Latin intentio originally understood as an embodied movement or ‘tension’


(see §2.7), the English grasp also came to be used to mean understand
(Sweetser 1990: 28).
It is also possible that Europeans misinterpreted certain Samoan terms as
corresponding to the English promise even though they were describing acts of
a slightly different meaning. This is what I will argue in my analysis of written
records of pre-Christian traditions and legends collected in the nineteenth
century by two Germans, Oskar Stuebel and Augustin Krämer (see below).
The arguments presented in this chapter should not be interpreted as mean-
ing that the Samoans I met or those whose speech I recorded did not under-
stand the act of promising. A great number of Samoans are fluent English
speakers and have lived abroad. A considerable number of them have also
been educated (in English) in colleges or universities in Samoa, New Zealand,
Australia, or the US. This means that, as in most places in the world today,
western meanings, including those that we associate with modernity, coexist
with a recurrent, if sometimes conflictual, attention to ‘tradition’ (aganu`u) or
to ‘the Samoan way’ (`o le fa`aSāmoa). More specifically, as I will show,
during my fieldwork in the late 1970s and during the 1980s, native Samoan
research assistants and consultants occasionally used on their own – that is,
without my prompting – the English word promise to translate certain Samoan
expressions I had recorded in spontaneous interactions. In evaluating those
translations, the issue for me is whether they fully capture the local and
context-specific meaning of those expressions. My answer will be that the
use of the English promise for translating some of the examples I collected
from spontaneous interaction might be a practical solution to rendering what is
being said, but at the same time it might also obscure a fundamental quality of
traditional Samoan ethnopragmatics. This quality, which coexists with
changing attitudes and practices, is the recognition of social action as typic-
ally – or even “essentially” – public. This means that the evaluation of
responsibility is done in terms of the effects of a person’s acts rather than in
terms of the original intentions, as defined in the conditions of satisfaction
proposed by Searle (see Chapter 2). Furthermore, this specific point is linked to
a more general point about Samoan ethno-theory of the self. In contrast to
Searle’s contention that one can speak of social action only when there is an
(invisible) intention behind the (verbal or non-verbal) act (see §2.2), I agree
with previous ethnographic accounts that for Samoans there is real “action”
when a person’s acts are done in public and as such can be evaluated by others
who might be affected by the consequences of those acts. This means that
commitments are binding not because of what a person thought, felt, or
believed, but because they afford public projection of whether something will
actually be done, regardless of what a person, in Searle’s language, “intended.”
It is in this respect that Loa’s case, discussed in Chapter 3, must be understood.
74 Promising in the Samoan translation of the Bible

As shown in example (12) in Chapter 3, the MP Fa`amatuā`inu was blamed


for not having carried through his commitment to bring presents to the matai
of the village of Falefā. But Loa’s reporting of what the MP had said he
was going to do was also to be blamed because of its effects, i.e., the
gathering of the matai waiting for gifts that never came. Loa’s reporting of
the announcement became a public act and as such was judged to be an
“action” by the senior orator Iuli and a few others regardless of Loa’s sincerity
or intentions.
Once we view commitments in this way we are not surprised that, as
I will discuss later in this chapter, Samoans had the concept of “oath,”
which was understood as a public act, involving several people as wit-
nesses, like, for example, when they are trying to find out the truth about
someone’s statement (see §4.7). This is a different act from a promise
understood as a commitment from one person to another over an extended
period of time without consequences for its violation.
As implicit in Austin’s and Searle’s definition of promise as directed toward
the future, temporality plays an important part in commitments – this was also
one of the points made by Adolf Reinach in his treatment of promises (Reinach
1913, 1983). This suggests that the conceptualization of time and timing might
also be relevant in cross-cultural comparisons of promises. For how long one
person must honor the obligation and for how long the other may continue to
hold on to his claim, for example, might vary from one community to another.
These differences might be inferred from local attitudes toward situations and
acts different from promising, including the phenomena of retaliation and
forgiveness.
The history of early encounters between Polynesians and European sailors
is full of episodes of conflicts and tragic deaths on both sides, including the
one that resulted in the death of Captain Cook in Hawaii (Beaglehole 1967;
Sahlins 1995) and of Captain de Langle and eleven other Frenchmen in
Samoa. The written accounts of such tragic encounters suggest that, whether
or not Samoans and other Polynesians reacted to unreasonable attitudes,
demands, or retaliation by European sailors and traders, they were also quick
to forget or forgive, or at least they were seen as such by the Europeans. This
inference can be made, for example, from the journal left by Count
Lapérouse (sometimes spelled La Pérouse), who had been sent to the Pacific
with two ships by Louis XVI to continue Captain Cook’s exploration. After
two days of trading with Samoans at sea, on December 11, 1787, La
Pérouse’s second in command, Fleuriot de Langle, decided to go ashore to
get supplies of fresh water. Following some exchange that might have been
interpreted by the Samoans as disrespectful, de Langle and eleven of his men
were killed by the Samoans, who, in turn, suffered even heavier casualties
(Tcherkézoff 2008: chapter 4).
The invention of promising 75

To La Pérouse’s astonishment, the next day five or six Samoan canoes came out to trade
with the French.1 Full of supplies La Pérouse brushed them off, firing a cannon near the
canoes to splash but not actually harm the occupants . . . La Pérouse then left.
The next vessel to stop at Tutuila appears to have been the Pandora (Gilson 1970: 67).
The Pandora’s tender repulsed a Samoan night attack on June 22, 1791, causing “terrible
havoc” and the death of several Samoans . . . Following this conflict . . . trade relations
took on a more positive tone. George Bass, the next European to trade at the island,
described the Samoans as “friendly and receptive” . . . (Borofsky and Howard 1989: 263)
Minimally speaking, these accounts suggest that Europeans and Samoans had
different expectations about what constituted a reasonable amount of time to
mourn, retaliate, resume interaction, or display respect for one another’s feelings
about the consequences of a recent tragic event. From a cross-cultural perspec-
tive, Europeans’ surprise at the timing of the Samoans’ acts of forgiveness (or
apparent forgetfulness) reminds us that, as pointed out by Bourdieu (1977) in his
discussion of gift-exchanges, the temporality of all kinds of act sequences,
including commitments and their fulfillments, should not be taken for granted
and is likely to be subject to cross-cultural variation. We might thus hypothesize
that for certain societies, including Samoa at the time of European contact, the
temporal span of promises, which goes from a few minutes to years or decades, is
so long as to seem unlikely or unreasonable. This hypothesis has an economic
corollary associated with the control that an individual has to satisfy the promise.
If we believe David Hume’s thesis in A Treatise of Human Nature that “the
obligation of promises is an invention for the interest of society” (Hume 1985:
576) and that it is meant to solve the problem of “the transference of property and
possession by consent” beyond the immediate and face-to-face “barter of com-
modities” (Hume 1985: 572), we might want to consider the type of society
where promises make sense. This was indeed one of the points made by Rosaldo
(1982) about the Ilongots. In Samoa, to a large extent still today, the extended
family (`āiga) has ultimate control over property, from pieces of land and its
products to precious artifacts like fine mats (`ie toga) exchanged for labor. Even
though at any given time a matai, as the holder of a family title, has some
authority to make decisions for the family as a whole, such authority is limited
in scope, including its temporal unfolding. From this point of view, the very idea
of promises as ways to exert some control over future action – an idea that is also
found in Hannah Arendt’s (1998: 243–247) discussion of promises – is contin-
gent on the type of socio-economic order of the community.

4.1.1 Searching for promising in Samoan texts and dictionaries


In my 1984 paper on Samoan intentionality – but not in the revised version of
that paper provided in the previous chapter – I had followed in Rosaldo’s steps
by briefly noting that in Samoan there is no special term for promise and that the
76 Promising in the Samoan translation of the Bible

Samoan translations that had been proposed of the English term promise suggest
that Samoan might have a different idea about what it means to promise. In this
chapter I return to my original claim in order to refine it through a more detailed
analysis of the expressions that have been offered as translations of the act of
promising. In addition to the data I collected in Samoa, I will rely more than
anywhere else in this book on written texts, which include two Samoan–English
dictionaries, the first complete Samoan translation of the Old and New
Testaments, which appeared in print in the 1880s,2 and other texts collected
and originally translated in German, also at the end of the nineteenth century, by
Oskar Stuebel (1846–1921), the German consul-general in Western Samoa
(Scarr 2013: 136, 159), and Augustin Krämer (1865–1941), a German Navy
medical officer who developed in Samoa a passion for ethnology and over a
period of several years was able to collect a considerable amount of precious
information about Samoan society and culture, including legends, myths,
genealogies, and descriptions of social events, customs, and artifacts (Krämer
1902, 1903).3 On the basis of my investigation of these texts, I will refine my
earlier and cursory discussion of the act of promising and propose that, although
it seems to exist as a possible speech act, it was probably not a native Samoan
concept when the missionaries arrived in the 1830s. Its appearance within the
Samoan repertoire of speech acts is likely to have occurred in the context of the
translation of the Bible into Samoan, a process of translation-as-acculturation
documented for other societies by a number of anthropologists, including
Bambi B. Schieffelin (1996, 2007) for the Bosavi (Kaluli) of Papua New Guinea
in the twentieth century, Webb Keane (2007) for Sumba, Indonesia, in the
nineteenth and twentieth centuries, and William Hanks (2010) for the Yucatec
Maya in the second half of the sixteenth century. As we shall see, this hypothesis
is based on the following evidence:

1. The word folafola, which dictionaries give as a Samoan translation of


promise (as a verb or as a noun, and in a number of possible derivations
and compounds), is extensively used with this meaning in the Samoan
Bible but very rarely if at all in the non-Christian texts of the same period
collected by Stuebel and Krämer.
2. Even in the two cases in which folafola is translated (by Stuebel, see below)
with the German verb versprechen ‘promise’ one could (and I will in fact)
argue in favor of an alternative gloss.
3. When we examine the full range of meanings of folafola (and its com-
pounds), it becomes apparent that its most basic meaning is that of a
particular type of movement with which objects are laid or spread out and
only secondarily that of a verbal act (‘spreading the news, announcing’),
which, in turn, can be further understood as ‘declare,’ and as such acquire
the meaning of a personal commitment by the speaker to future action.
Samoan translation of English promise 77

4. Such a metaphorical extension of folafola from a motion verb (‘laying out,


spreading’) to a personal commitment that requires the internalization of the
intention to act in the future manifests a family resemblance with other
metaphorical extensions introduced in the Samoan Bible, whereby external,
public, and perceptible actions, dispositions, or qualities of persons became
reinterpreted as entities, processes, or referents that are supposed to exist
inside of a person and as such are not available to human perception and
public scrutiny.
5. One speech act of commitment recorded in eighteenth-century texts is the
oath. As shown later in this chapter (§4.7), this was something done with
respect to the past, e.g., to swear to the truth of a statement made. Today the
same term (tautoga) is used to refer to commitments oriented toward the
future such as “the oath of office” (tautoga i le tofi) or “the oath of
allegiance” (tautoga o le loto faamaoni) that public officers in the Parlia-
ment or in the judiciary must take (articles 28, 34, 61, 67, and 78 of the
Samoan Constitution, 1960 and 2008). Differently from promises, oaths are
by definition public acts, as shown by the description of the “Oath of
Allegiance” for Members of Parliament that must take place “before the
Assembly” (i luma o le Fono Aoao) (Article 61).
6. Even today there is some ambiguity and uncertainty about the use of
folafola for expressing the notion or the act of promising, thereby suggest-
ing a continuous cultural dispreference for the use of commissives that do
not imply public exposure.
7. The noun māvaega, which is sometimes translated as ‘promise’ in English,
at close analysis can be shown to have had and to continue to have a
different meaning that oscillates between a person’s wish or will before
dying (particularly appropriate for matai) and a decree, stipulation, or
agreement between parties.

4.2 Samoan translation of English promise in dictionaries


To examine how Samoans might express the act of promising, we will first
look for promise in Samoan–English dictionaries, two of which, in our case,
have the advantage of being quite thorough and yet separated by more than
half a century from one another. By consulting Rev. George Pratt’s Grammar
and Dictionary of the Samoan Language (1893, third edition) and G. B.
Milner’s (1966) Samoan Dictionary we will have the advantage of two sources
that were compiled at different times and by scholars with a different profes-
sional training. Pratt’s dictionary is the work of a nineteenth-century mission-
ary who, as we shall see, played a major role in the translation of the Bible into
Samoan. Milner’s dictionary was produced by a professionally trained lexi-
cographer, informed by mid-twentieth-century linguistic theories and methods.
78 Promising in the Samoan translation of the Bible

As we shall see, in some cases the two dictionaries provide very similar
translations. In other cases, they offer different insights and cover a different
set of vocabulary items.
Both Pratt and Milner translate promise with the Samoan word folafola, the
only minor difference between the two dictionaries being the spelling: Pratt
spells the word folofola and Milner spells it fōlafola, using a macron on the
first vowel to mark o as a long vowel (phonemically /oo/, phonetically [oː]).
The different transcription of the same word is not surprising given the
different attention given to vowel length by the two authors. To avoid confus-
ing readers, I will hereafter use only the spelling folafola (without the macron
to mark the first long vowel) in my discussion of the term, unless I am quoting
from Milner’s dictionary or other sources that mark the long vowel.4
When we go to the Samoan–English part of Pratt’s dictionary, and we look
up folafola, we find the following English glosses: “FOLAFOLA, v. 1. To
spread out. 2. To unfold, to preach. 3. To promise” (Pratt 1893: 162). In
addition, Pratt gives a similar set of glosses to the non-reduplicated form fola:
“1. To spread out, as mats to sleep on. 2. To unfold, as the hand. 3. To promise;
redup. FOLAFOLA; pl. FOFOLA; pass. FOLASIA” (Pratt 1893: 161–162).
We also find both fola and folafola as parts of compounds. One is con-
structed with the causative prefix fa`a-, namely, (a) fa`afola ‘to open up, to
spread out, as a cloth, or the fingers,’ and (b) fa`amāfolafola ‘to spread out, to
unfold’ and also ‘to explain’ (Pratt 1893: 122). Another compound is māfola, a
verb meaning “1. To be spread out. 2. To be extensive, to be wide. 3. To be
plain, to be perspicuous, of a speech.” Pratt also lists the reduplicated form
māfolafola with no gloss (but Milner translates it as ‘be flat’).
In the Samoan–English part of Milner’s dictionary we find similar glosses
for fōlafola and its nominalization, fōlafolaga.
fōlafola. v. 1. Announce (publicly). Sā __ e le faife`au le fa`amanatuga mo lea Aso Sā:
The pastor __d a communion-service for that Sunday. 2. Acknowledge (a gift) by
public announcement (Samoan custom). `Ua __ i fafo meaalofa a le nu`u: The presents
from the village were __ d outside (the house). 3. Promise. Na `ou __ `i ai le lima
pauni: I __d him 5 pounds. fōlafolaga n. 1. Announcement. __ a le faife`au: The
pastor’s __. 2. Acknowledgment. __ o meaalofa: __s of presents. 3. Promise. `O lau __
`iate a`u: Your __ to me. (Milner 1966: 68)
As shown below, Milner, unlike Pratt, does not translate fola, with ‘promise.’
fola v. 1. Strew. __ le fale `i `ili`ili: __ the floor of the house with pebbles. 2. Spread. n.
Floor. `Ua ufitia le __ `i le `ili`ili: The floor is covered with pebbles. (Milner 1966: 68)

Nevertheless, Milner’s translations of fola are consistent with what appears as


the basic and, as we shall see, most common meaning of both fola and fōlafola,
namely, spreading, putting, or laying out. The meanings ‘announce’ and
‘acknowledge (publicly)’ make sense as semantic extensions, from spreading
Samoan translation of English promise 79

out, laying out objects, it is not too far to spreading information. Two examples
are provided in (1) and (2) below from the transcripts of conversational
interactions I recorded in Samoa.
(1) (Dinner # 3; 1988)
T; o`u alaku lā o`u moe
‘I went there (and) slept’
leai ma gisi-
‘There was no one- (else)’
folafola ou fala o`u moe.
spread myþPl mat I sleep
‘(I) spread out my mats (and) I slept’
Sk; agapō?
‘Last night?’

(2) (Women of the Congregation; 1988)


Vg; `oi `oi! mm. lā lava e kalagoa lā F[NAME] ma N[NAME]
‘Oh-oh! (yes) while F and N are talking’
`ae lau `uma ifo ā e P[NAME] iā ke a`u ia kala a N[NAME]
‘(and) P told me (cried out) everything (about) N’s stories’
fai le folafolaga i mea `ua fai.
do Art announce-Nom Prep thing Perf do
‘(she) made an announcement about the things that were done.’
okaoka gi mea `ua fai!
‘Oh gee, the things that were done!’
In (2) the term folafolaga (nominalization of the verb folafola by means of the
suffix –ga) is used as part of a series of affective displays of surprise (`oi! `oi!!)
or disbelief (okaoka!) by the teller (Vg) for the things told by N, which
P repeats, as well as for the very fact that P shows bad judgment by repeating
such things in public. This use of folafola (with suffix –ga) corresponds to the
first gloss provided by Milner (see above): ‘announce (publicly).’ The same
meaning is found in the expression folafola `ava used for the kava announcer,
as shown in (3), where speaker S inquires about the compensation received by
the kava announcer in an event attended by the people he is talking to.
(3) (Dinner #3; 1988)

S; e ka`i fia lā gi kou mea ga maua?


‘how many things did each of you get?’
P; e limasefulu kālā le lāuga
‘fifty dollars (for) the speech (that is, for each speech)’
S; `ae fia le folafola`ava?
‘and how much (for) the kava announcer?’
The close family resemblance between the meanings ‘spreading out’ and
‘distributing,’ on the one hand, and between ‘spreading out’ and ‘announcing’
80 Promising in the Samoan translation of the Bible

(in the sense of ‘spreading news or information’) on the other is further


supported by the fact that the ‘kava announcer’ (folafola `ava) mentioned in
example (3) can also be referred to as tufa `ava ‘(the one who) distributes (tufa)
the kava (`ava),’ in which tufa literally means ‘distribute, share.’
In the context of the meetings of the village council (fono) discussed in
previous chapters, folafola is also used in the sense of ‘laying out’ an argu-
ment, ‘presenting’ evidence. An example is provided in (4), where the senior
orator Moe`ono is in the middle of a long speech where he is recapping the
events that have brought about the crisis the assembly is called upon to resolve.
(4) (Fono January 25, 1979; p. 30 of ms.)
Moe`ono; `o le mea lea e fōlafola māe`a
Top Art thing this TA lay thorough
‘this is to (be) laid down thoroughly’ (i.e., ‘to be explained’)
Similarly, folafola is also used by Moe`ono in the same speech in combination
with the word mālamalama (‘clear’ and ‘understand’) to mean ‘explain’ – as in
‘making clear,’ ‘laying out’ something for others to ‘understand.’
All of these examples show folafola to be a polysemic lexical item. Its
various meanings are semantically related metaphorical extensions that are
easily grasped even by non-native speakers. The missing piece at this point is
the further semantic extension to mean ‘promise,’ which is absent in my data,
with two possible exceptions, both of which I will contest. The first is
reproduced in the first line of example (12) in Chapter 3. The senior orator
Moe`ono is shaming the MP for not coming through with his folafolaga (in the
plural, as shown by the possessive `au), namely, the never-delivered gifts that
are at the origin of Loa’s case discussed in Chapter 3. In the original transcript,
folafolaga was translated as ‘promises’ by one of my Samoan consultants. But
the translation ‘announcement’ or ‘announced things’ is equally acceptable.
The second case is the word māvaega, which I will discuss later in this chapter
(in §4.6), where I will argue that the more appropriate English translation of
this word is ‘agreement.’
The use of folafola as ‘promise’ is, however, predominant in an important
written text, the Samoan translation of the Old and New Testaments – O le Tusi
Paia o le Feagaiga Tuai ma le Feagaiga Fou lea ua Faasamoaina (hereafter
shortened to Tusi Paia). Before examining this text in some detail, I will
briefly review the history of its translation.

4.3 Challenges in the translation of the Bible into Samoan


In the early 1830s members of the London Missionary Society (LMS), first
among them John Williams, arrived in Samoa (Williams 1832, 1837; Prout
1843; Ellis 1844). They brought with them native teachers from Tahiti and the
Challenges in the translation of the Bible into Samoan 81

Cook Islands (known at the time as Hervey Islands), where Christian missions
had already been established, who could act as culture- and language-brokers.
As elsewhere in Polynesia, one of the first tasks in establishing a mission in
Samoa was the production of Christian texts and the simultaneous introduction
of literacy. By the time six more LMS missionaries arrived in 1836, in addition
to a spelling book and a book of hymns, the entire Gospel according to
Matthew had already been translated into Samoan (Murray 1888: 40). As
elsewhere in the Pacific, in Samoa also the translators were following rules
established by the British and Foreign Bible Society. Here is a passage in
which Rev. Archibald W. Murray narrates some aspects of the process of the
translation and the revision.

We, with our native pundits, used to sit about nine hours daily, except on Saturdays,
when we took half a holiday to recruit a little and prepare for the Sabbath services.
Whether at home at our own stations or elsewhere there was always a demand for
preaching when the Sabbath came round. Our sittings extended over two, three, and
sometimes four, or even five weeks, according to the length or difficulty of the book we
had in hand, but we used to feel that three weeks were about long enough for a
continuous sitting. We were pretty well furnished with critical helps, and versions were
of service to us. The Septuagint, the Vulgate, and our South Sea versions which had
preceded our own – the Hawaiian, the Tahitian, the Rarotongan, and the Tongan –
always were on our table. Boothroyd’s English version also, and of course the author-
ised version, and English commentaries, were all laid under contribution. (Murray
1888: 43)
Murray’s account is supported by other sources, including correspondence
between Rev. George Pratt and Rev. Henry Nisbet that I was able to review.5
Nisbet’s journal (1836–1876) – available in microfilm (Mitchell Library,
Sydney) – also contains a day-by-day description of his study of Latin, Greek,
and Classical Hebrew at the Turvey Congregational Academy.
From these accounts we learn that each new translation was based on the
“authorised version” (in English) and a number of original texts or older
translations – Murray (1888) mentions the Textus Receptus, the Vulgate, and
the Septuagint – as well as previous translations in other Polynesian languages,
all of which are closely related and thus exhibit lexical and morpho-syntactic
similarities. These similarities among Polynesian languages made it possible
for the British missionaries in Samoa (and elsewhere) to employ native
speakers (the “native pundits” in the above quote) from other island communi-
ties (e.g., from Tahiti) who had already been converted, had acquired literacy,
and had done translation work in a closely related language.
According to Murray (1888: 45), the first translation of the New Testament
into Samoan was completed in 1850. The translation of the Old Testament was
started then and completed within five years. A first revision was done by two
other missionaries, Henry Nisbet and George Pratt. A second revision was
82 Promising in the Samoan translation of the Bible

done between 1867 and 1870 by Pratt, Murray, Turner, and Whitmee (Murray
1888: 48). By the 1880s Samoans had access to bound copies of one single
volume containing both the Old and New Testaments. In my investigations
I have drawn from an 1884 edition reprinted by The Bible Society in the South
Pacific (n.d.) and an 1887 edition that is now available as an electronic file at
http://books.google.it/books/about/O_le_tusi_paia_o_le_Feagaiga_Tuai_ma_le.
html?id=aDs7AAAAcAAJ&redir_esc=y. The examples cited in the rest of this
chapter are identical in the two editions.
Pratt (1817–1894) is an important actor in this challenging, lengthy, and
complex activity. He was a biblical scholar who spent forty years in Samoa
(1839–1879) and was described as “a very accomplished speaker of the
vernacular” (Newell 1911b). Pratt’s Samoan–English dictionary was first
published in 1862 and then revised by him in two subsequent editions, in
1876 and 1893,6 which include a grammar of Samoan. After Pratt’s death, in
1892, J. E. Newell, who is the author of a grammar of English written in
Samoan (Newell 1891) and had access to Pratt’s original manuscripts, pre-
pared a fourth and last edition.7 In the preface to his 1911 revision of Pratt’s
dictionary (dated “Samoa, March 1910”), Newell writes that Pratt was also “an
ardent student of Hebrew and Arabic.” That Pratt had a remarkable linguistic
knowledge is confirmed by a number of sources, including Richard Lovett’s
(1899) history of the London Missionary Society, where he quotes a letter
from “an old colleague and friend, S[amuel] J[ames] Whitmee,” who wrote:
The translation, and then the revision of the Samoan Bible, was the great work of
[Pratt’s] life. To this he devoted almost daily attention for many years, with the result
that the Samoans have a Bible which, as a classic, is, and will be to them, very much
what the Authorized Version has been in England. His Hebrew Old Testament and his
Greek New Testament were among his most cherished companions, whether he was at
home or traveling. He had also a very perfect knowledge of the Samoan language, and
spoke it like one of the natives of a generation now passed away, before the language
had suffered from modern corruptions. He was also familiar with the classic traditions
of the people, and could illustrate and give points to his speech by such telling
references and allusions, that it was always a treat to the natives to hear Palati [Samoan
for the name Pratt] speak. He had no uninterested hearers. He accordingly had little
patience with missionaries who were contented with an imperfect knowledge of the
language of the people to whom they preached, or who were given to careless speech.
(Lovett 1899: 388–389)
As I suggest in §4.4.1 below, Pratt’s (and other missionaries’) knowledge of
Greek and Classical Hebrew will become important for making sense of some
of the lexical choices made in the Samoan translations of what is rendered with
promise in the King James Bible.
One of the challenges for Pratt and his co-workers – as for others before and
after them8 – was to find Samoan lexical counterparts for the multitude of
referents, concepts, and practices found in the Bible. In some cases, the
Challenges in the translation of the Bible into Samoan 83

Figure 4.1 Title page of 1887 edition of the Tusi Paia, the Samoan translation
of the Bible, a collaborative effort by George Pratt, Henry Nisbet, and others
84 Promising in the Samoan translation of the Bible

solutions seemed ready-at-hand, as we learn from Rev. George Turner (1861),


who proudly lists numerous Samoan terms, customs, and metaphors that were
interpreted by the missionaries as appropriate counterparts to or even precur-
sors of those found in the Bible. This is a “continuity” strategy, also adopted
elsewhere in the nineteenth century, which coexists and gets entangled with
the goal and desire to break the ties with the heathen past (see Keane 2007 on
the double message of continuity and change in the adoption of Sumbanese
traditional oratory by Christian missionaries and their local advocates). But the
Samoan Bible also provides many examples of expressions that were inno-
vations. The most obvious cases are loanwords or borrowings (see, among
others, Weinreich 1953; Brown 1999; Thomason 2001; Garrett 2004; Maki-
hara and Schieffelin 2007; Haspelmath and Tadmor 2009), that is, foreign
words (e.g., from English, Greek, or Latin) that were adapted to Samoan
phonology and word structure. This technique introduced such words as filo
‘thread,’ `auro ‘gold’ (from Latin), luko ‘wolf,’ `ekālēsia ‘church, religious
congregation,’ lokou ‘word, logos’ (from Greek),9 temoni ‘demon’ (probably
from English rather than from Ancient Greek daimon), and hundreds of proper
names of individuals, groups, towns, and countries, which were also all
adapted to Samoan phonotactics, which avoids consonant clusters and requires
each syllable to have at least a vowel. The many names imported include
Tavita ‘David,’ Herota ‘Herod,’ Aperaamo ‘Abraham,’ Ruta ‘Ruth,’ Isaraelu
‘Israel,’ Kalilaia ‘Galilee,’ and so forth.
Other new referents and concepts were introduced by extending the mean-
ing of existing Samoan terms. Thus, the traditional role of failāuga ‘speech-
maker,’ a term used to refer to the holder of tulāfale (as opposed to ali`i) title,
was used in the Old Testament to translate ‘preacher’ (e.g., Ecclesiastes 1:1,
1:2, 1:12, 7:27, 12:8, 12:9, 12:10). There was also innovation in the ways in
which native terms were combined to create new lexical items for a wide range
of artifacts, social roles, and social institutions, like, for example, fale sā
‘church’ (lit. ‘sacred/forbidden house’), fafine talitāne ‘harlot, prostitute’ (lit.
‘woman (who)-receives-man’), nu`u `ese ‘Gentiles’ (lit. ‘different/distinct
villages’), and fale puipui ‘prison’ (lit. ‘house-fence’).

4.3.1 Translations for the Christian notions of “soul”and “spirit”


Translations were needed for terms like soul and its quasi-synonyms spirit and
heart, which imply well-established – and often unquestioned – European
contrasts like the one between “body” and “soul” or “body” and “spirit” found
in the passages reproduced in (5) and (6) below from the New Testament (all
English examples are from the King James Bible found online at http://www.
kingjamesBibleonline.org):
Challenges in the translation of the Bible into Samoan 85

(5) (Matthew 10:28)


And be not afraid of them that kill the body, but are not able to kill the
soul: but rather fear him who is able to destroy both soul and body in hell.

(6) (2 Corinthians 7:1)


Having therefore these promises, dearly beloved, let us cleanse ourselves
from all filthiness of the flesh and spirit, perfecting holiness in the fear
of God.

In these cases, the missionaries and their Samoan collaborators, challenged


by concepts and contrasts that were not easily transferable to Samoan lan-
guage, practices, and worldview, resorted to two main strategies. The first
was to give a more literal translation of an originally metaphorical expres-
sion. The second was to alter and thus extend the meaning of existing
Samoan lexemes. In comparing the King James Bible with the Samoan Tusi
Paia, we find that both strategies were adopted for the translation of the
English soul.
When soul seemed to refer to actual human beings, it was translated with
Samoan common nouns referring to persons or with personal pronouns,
thereby eliminating the metonymical relationship between “soul” as the non-
material inner part of a person and his external, corporeal manifestation. For
example, in Genesis 17:14: “that soul shall be cut off from his people”; the
expression that soul is lea tagata ‘that person’; in Genesis 19:20, the expres-
sion my soul in “. . . and my soul shall live” is translated with the Samoan first
person pronoun a`u, de facto turning the clause into “and I shall live” (ona ola
ai lea o a`u). Similarly, in Genesis 34:8 “The soul of my son Shechem longeth
for your daughter . . .” in the Samoan text becomes “my son Shechem, he has
been very much wanting/needing your daughter . . .” (O lo`u atalii o Sekema, o
loo ua manao tele lava o ia i lo outou teine), in which “the soul of” becomes
“he” (o ia).
When soul (or spirit) refers to the spiritual and inner side of the human self,
it was usually translated either with loto and more rarely with agaaga (usually
spelled agaga in the Bible even though it is a reduplication of the term aga
‘conduct, manner of acting’ and as such should be agaaga or agāga – as
recognized by Pratt 1893: 2, 67). In fact, the word loto, which was discussed in
§2.7.1 as one of the candidates for translating intention, turns out to be a
widely used term in the Tusi Paia to cover a wide range of psychological and
emotional constructs, including “soul” and “heart,” as shown in (7) and (8)
below (see also Cain 1979: 61–69).
(7) (Genesis 42:21)
na tatou iloa le puapuaga o lona loto
Pst we-all know Art affliction of his soul
‘we saw the anguish of his soul’
86 Promising in the Samoan translation of the Bible

(8) (Proverbs 23:7)


manatu o ia i lona loto
think Top 3rdSgPro Prep his heart
‘he thinketh in his heart’

The word agaga was regularly used for the expression Holy Spirit or Holy
Ghost (Agaga Pa`ia) and sometimes for soul, including cases where soul is
contrasted with body as shown above in the passage in (5), and flesh in (6).
Their respective translations are provided below in (9) and (10).

(9) (Tusi Paia, 1887, Mataio 10:28)


Aua foi tou te matatau i e fasioti le tino,
Don’t also you TA fear Prep those-who kill Art body
‘And be not afraid of them that kill the body,’
a e latou te le mafaia ona fasioti le agaga;
but TA they TA Neg possible Comp kill Art soul
‘but are not able to kill the soul:’
a e lelei ona outou matatau i le ua mafai
But TA good for you-all fear Prep the-one-who Perf possible
‘but rather fear him who is able’
ona fano ia te ia le agaga atoa ma le tino i Kēna.
Comp perish Prep him Art soul whole and Art body in hell
‘to destroy both soul and body in hell.’

(10) (Luke 23:46)


Lo’u Tamā e, ou te tuuina atu lo’u nei agaga i ou aao: [. . .]
My father Voc I TA place Dx my here spirit in your hands
‘Father, into thy hands I commend my spirit: [. . .]’

The extent to which one believes that this meaning of agaga, like the meaning
‘promise’ of folafola, is a semantic innovation depends on the interpretation of
the native use of agaaga and its base aga, which has been the subject of
considerable debate among students of Samoan culture and society, especially
in the context of the relation between aga and another term, amio ‘behavior’
(Shore 1982, 1983; Love 1983; Freeman 1984, 1985; Mageo 1992). One issue
is whether the semantic difference between aga and amio could or should be
captured in terms of dichotomies such as culture vs. nature, social vs. personal,
and positively vs. negatively valued action, a position held by Shore (1982)
and criticized by others (e.g., Love 1983; Freeman 1984, 1985; Mageo 1989),
or it should be understood as a contrast between an inner, psychological
being (aga) and an outer self or social ways of being (amio), as suggested
by Love (1983). Putting aside the problematic interpretation of the aga/amio
dichotomy as corresponding to culture/nature, I think there is empirical
evidence that aga refers to forms of behavior that are seen as characteristic
Challenges in the translation of the Bible into Samoan 87

or typical of a certain entity, whether human or otherwise, as suggested by


Mageo (1989), who also criticizes Freeman for interpreting aga as something
innate (Mageo 1989: 183). Aga is thus used to refer to some expected, and only
in this sense “essential,” property of human conduct, which is, however,
always detectable by the senses (e.g., sight, hearing, touch) and thus publicly
assessable in cultural terms, that is, as expressed in particular ways of acting in
the world. Mageo suggests our common notion of ‘face’ as a possible transla-
tion of some of the uses of aga in Samoan (Mageo 1989: 183–185). Pratt
himself (1893: 67) under the entry aga as a noun gives the following (untrans-
lated) example: `o aga a le nu`u, which we may translate as the ‘ways of
behaving’ (aga) of a community or group (nu`u). This translation is consistent
with Pratt’s translation of the compound aganu`u as ‘conduct according to the
customs of one’s own country’ (Pratt 1893: 68). Further evidence of this
original meaning of aga is found in other Polynesian languages where aga
denotes custom, habits, ways of acting (in Samoan aga can also be a verb
meaning ‘to do, to act, to go,’ see Pratt 1893: 67). The reduplicated Samoan
form agaaga (or agaga) can be connected to a reconstructed Proto-Polynesian
lexeme *angaanga, which has reflexes for visible and concrete entities like the
body or the head in Kapingamarangi, Maori, and Tuamotu (http://pollex.org.
nz/search/?field=entry&query=aga). Possibly related to these uses is the mean-
ing of the Samoan verb aga meaning ‘to face, confront’ (Milner 1966: 7) and
its compounds.
How do we go from aga and agaaga as referring to detectable social acts
and conduct to the “soul” and the “spirit”? There are at least two possible
paths. One is Mageo’s suggestion that the missionaries adopted the pre-
Christian Samoan notion of agaaga as the disembodied “ghost” of dead people
and transformed it into the notion of the soul surviving after death (see also
Stuebel 1896; Krämer 1902; Cain 1971, 1979: 21–23, 45–49, 70–143). Some
support for this hypothesis is found in the Methodist missionary George
Brown’s brief description of pre-Christian Samoan religious beliefs. Brown
specifically mentions agaga (which he spells anganga) meaning ‘soul’ or
‘spirit’ and interprets as a possible reduplication either of the verb aga ‘to
go, to do or to act’ or of the noun aga meaning conduct or manner of acting
(Brown 1910: 218). In his account, “[w]hen the body dies the spirit flies away,
and it is said ua lele le anganga, that is, the spirit has flown away.” Brown’s
reconstruction, as well as the stories collected by Stuebel and Krämer that
mention the agaga, seem to support at times a disembodied but still visible or
materialized entity.
The soul was always considered to be of the same form as the body, and was much
dreaded by those who professed to see it shortly after the death of the individual.
(Brown [1910]1972: 219)
88 Promising in the Samoan translation of the Bible

Other times, as shown in the Samoan narratives collected by Krämer, the


agaga could be wrapped in a leaf10 or a basket, thereby implying a perceiv-
able, physical quality:
Then Mala‘ulufotu looked up to the ridge pole of the house and saw that the house was
full of little baskets, just hanging around. In the midst of them hung a new basket. So
Mala ‘ulufotu asked Fulu ‘ulaalematoto’s boy: What kind of a new basket is that
hanging on the ridge pole of the house? And the boy answered: That is the soul [agaga]
of Tuifiti’s daughter that my mother brought. Mala‘ulufotu reached up and came down
with Sina’s soul [agaga]. And so Sina was revived. (Krämer 1994: 153–154; original
Samoan text with German translation in Krämer 1902: 124–125)

Another suggestion about the transformation of aga (and agaga) into the
Christian ‘soul’ was made by Bradd Shore in his response to Love’s criticism
of Shore’s cultural interpretation of aga and amio in his 1982 book Sala`ilua:
A Samoan Mystery. Shore (1983: 154) proposed that aga as the center of the
social self (of Samoan culture) becomes in the Bible “the Western, private,
psychological self” of Christianity.
Regardless of whether we agree or not with the details of these claims, these
few examples of semantic innovations remind us that languages are malleable,
and new meanings can be added to existing words and expressions, sometimes
replacing or obfuscating the original meaning, other times coexisting with it,
not always in an overall coherent fashion (see Keane 2007: 132–146). This is
what might have happened to folafola.

4.4 Translations of promise in the Bible


I searched for the word promise in the King James Bible either as a noun or as
a verb. I found a total of 112 cases11 and I matched the sentence in which they
occurred with the corresponding passages in the 1887 edition of the Samoan
Bible (which does not include a translation of the Apocrypha). The term
folafola (in various morpho-syntactic realizations) was used in 69 cases (see
§4.4.1 for their distribution). These occurrences of folafola do not exhaust all
the cases of this term in the Samoan Bible given that it is also used with some
of the other meanings mentioned above in the discussion of the dictionary
entries, e.g., ‘spread’ and ‘preach.’
Example (11) shows folafola used as a verb – with the aspectual prefix `ua
(spelled ua in the Tusi Paia) marking perfect – and example (12) shows
folafola as part of a nominal construction (a head noun plus a relative clause)
to create a Samoan translation of the English promise used as a noun. The
Samoan phrase le upu na folafolaina means ‘the word (upu) that was folafola,’
with the suffix –ina giving a passive flavor to the verb form (even though
Samoan grammar does not have a passive voice). (All examples from the Bible
hereafter are given in three lines: the first is from the King James Bible, the
Translations of promise in the Bible 89

second, in Samoan, from the Tusi Paia, the third is my own word-by-word
English gloss of the Samoan text).
(11) (Hebrews 10:23)
. . . for he is faithful that promised;
. . . auā e faamaoni o ia ua folafola mai
because Pres faithful Pred he Perf folafola Dx

(12) (2 Peter 3:4)


. . . Where is the promise of his coming?
. . . O ifea o i ai le upu na folafolaina ai lona afio mai? . . .
Pred where Pred there Art word Pst folafola Pro his move Dx
Most of the cases where folafola is not used as a translation of the verb or noun
promise in the King James Bible are rendered with verbs of saying – the most
common being the respect vocabulary term fetalai when the speaker is God, as
shown in (13).12
(13) (Deuteronomy 27:3)
. . . when you pass over to enter the land which the LORD your God gives
you, a land flowing with milk and honey,
as the LORD, the God of your fathers, has promised you.
Pei ona fetalai mai ai e Ieova o le Atua o ou tama ia te oe.
As Comp speak Dx pro Erg Jehovah Pred Art God of your father to you
‘as Jehova, the Lord of your fathers, said to you [sing.]’ (my back translation)
In three cases we find the noun afioga, a respect vocabulary term that in
addition to referring to a person who holds a (high) chief title (used for both
reference and address, as in lau afioga ‘your honor’), it also describes a chief’s
speech (Milner 1961: 308; 1966: 6). The same word is also used in the fono
I recorded to mean a chief’s (or chiefs’) ‘decision’ (perhaps in the sense of ‘the
position at which a chief has arrived’). A more general term to encompass
these various uses would be ‘delivery’ or ‘delivered opinion.’ Two of the three
examples with afioga are from Psalms. Here is one of them:
(14) (Psalm 19: 154)
give me life according to thy promise!
ia e faaolaola mai ia te au e tusa ma lau afioga.
Imp you revive Dx to me according with your decision/delivery

The “common” (i.e., non-honorific) verb fai ‘say’ (but also ‘do, make, have’)
and the noun `upu ‘word’ are used for mortals and, as shown in (15), where
God is speaking to Solomon, for God’s quoted direct speech, showing that
God’s speech is made to follow the pragmatic rule according to which speakers
should not use an honorific form in talking about themselves. Verbs of saying
are typically combined with deictic particles like mai or atu indicating the
directionality of the act, i.e., who is saying what to whom (Platt 1982).
90 Promising in the Samoan translation of the Bible

(15) (from 1 Kings 9:5)


Then I will establish your royal throne over Israel for ever,
As I promised David your father,
e faapei ona ou fai atu ai i lou tamā o Tavita
Pres as Comp I say Dx Pro your father Pred David
A more literal back translation would be: ‘like I said to your father David.’

These two translations of promise – one with folafola and the other with verbs
of saying (or nouns, like `upu ‘word,’ referring to acts of speaking) – are also
found in the previously mentioned manual written in Samoan by Rev. J. E.
Newell (and published in London by the London Missionary Society in 1891)
for teaching English grammar to Samoans. Newell translates promise with
folafola as well as with two verbs of saying, ta`u (‘to tell, to mention’ in Pratt
1893: 281; Newell 1911a: 295) and fai mai (see above) (Newell 1911a: 126).
I found no examples of ta`u used for promise in the Samoan Bible, but see the
compound ta`utino discussed below.
I also searched for all the uses of folafola in the Tusi Paia and found that a
relatively frequent meaning of this word in the New Testament is ‘preach,’
which can be made sense of as a metaphorical extension of ‘spreading the
Word’ or the ‘Good News’ (tala lelei), as in ma ua folafolaina foi le tala lelei i
e matitiva ‘and the poor have the gospel preached to them’ – Matthew 11:5). In
the Old Testament, preaching is tāla`i (spelled tala’i), a word that Pratt (1893:
296) translates as ‘to proclaim’ and Milner (1966: 234) as ‘to propagate’ as
well as ‘boast’ and ‘brag.’ The same verb can be used in a nominalization for
‘the one who preaches’ as an alternative to the term failāuga (‘speechmaker’)
mentioned above.

4.4.1 Differences between translations of promise in the Old


and New Testaments
At a closer look we realize that the above-mentioned uses of folafola corres-
ponding to translations of promise (as a verb or as a noun) in the King James
Bible is unequally distributed in the Old and New Testaments. In the Old
Testament folafola only in a few instances appears in passages where the
English text has promise, but in the New Testament folafola matches almost
every English occurrence of promise. In the Old Testament, when the King
James Bible has promise, the Samoan Tusi Paia tends to have Samoan verbs
meaning ‘speak, say’ or nouns meaning ‘speech’ or ‘word.’ The most common
word used in these cases is the honorific verb fetalai ‘speak,’ a respect
vocabulary item (`upu fa`aaloalo) appropriate when the referent of the subject
of the verb is an orator (tulāfale). Other, less frequent translations are afioga
‘speech, opinion (of a chief),’ and ordinary (non-honorific) words like `upu
(spelled upu, without the glottal stop) ‘word’ and fai atu (two cases).
Translations of promise in the Bible 91

I propose that to make sense of these differences between the Samoan Old
and New Testaments we need to take into consideration the original Hebrew
and Greek texts.13 Before doing so, however, we need to establish that the
missionaries were in fact examining the original texts.
According to Murray (1888: 46), Rev. George Pratt and Rev. Henry
Nisbett were the two people who were put in charge of gathering all the
materials previously produced by the translation team of missionaries and
native speakers and editing a “final revision” (of what will turn out to be
the first complete edition) of the Samoan Bible.14 I believe that there is
sufficient evidence to support the thesis that Pratt and Nisbet not only
knew Classical Hebrew and Ancient Greek, but were in fact systematic-
ally consulting various versions and early translations of the original
texts.
Murray himself, for example, mentions that Pratt had helped Rev. W. G.
Lawes in translating the Bible into Niuean (another Polynesian language)
(Murray 1888: 56) and then quotes Lawes (1888: 58), who lists a number of
“originals” that he had consulted, including Van der Hooght’s Hebrew Bible,
various critical versions in Greek and English, as well as the Textus Receptus
and the Septuagint, which were both popular at the time. I also learned from
reading Nisbet’s journals that in preparation for his missionary work, together
with the other members of the London Missionary Society, he had been sent to
be trained in Latin, Ancient Greek, and Hebrew, and was expected to practice
in those languages by translating from the Bible.15 Direct evidence of the
attention paid to the original texts by the missionaries from the London
Missionary Society is found in Pratt’s correspondence with Nisbet from the
late 1860s (now at the Memorial Library in Sydney, see endnote 5) where he
mentions the Polyglott (a text that contains the Bible texts in various languages
including Hebrew and Greek) and gives lists of words and phrases for which
he is considering changes for the revised Samoan translation.16 In those lists
we find mentions of Greek terms such as πάντοτε (pantote) ‘always, all the
time’ translated with the Samoan so`o (spelled soo by Pratt and in the Tusi
Paia).
Given this background, a possible explanation for the fact that folafola
corresponding to promise in the King James Bible is rarely found in the
Samoan version of the Old Testament is that Pratt, Nisbet, and the other
missionaries in charge of the translation were trying to capture the meaning
of the original text, where there was no Hebrew word closely corresponding
to promise. The choice of the Samoan verbs of saying mentioned earlier
makes sense given that the most common root used in the Hebrew text for
the cases of promise in the King James is dbr ‘to speak.’ In the Old
Testament, that is, God does not “promise,” he “speaks.” This is consistent
with the more recent English translation of the Five Books of Moses done
92 Promising in the Samoan translation of the Bible

by Robert Alter (2004), where the LORD does not promise, as shown in
Alter’s translation of Deuteronomy 28:3, previously shown in example (13),
here reproduced as (13’) with Alter’s translation followed by the relevant
Samoan translation:
(13’) (Deuteronomy 27:3)
. . . as the LORD God of your fathers has spoken to you. (Alter 2004:
1008–1009)
Pei ona fetalai mai ai e Ieova o le Atua o ou tama ia te oe.
As Comp speak Dx pro Erg Jehovah Pred Art God of your father to you
‘as Jehova, the Lord of your fathers, said to you [sing.]’ (my back translation)
One particularly telling example and additional evidence for the attention to
the Hebrew text is the Samoan translation of Ezekiel 13:22, which in the King
James Bible includes the phrase “that he should not return from his wicked
way, by promising him life” (emphasis mine). In this case the Samoan text
does not have anything corresponding to “by promising.” It turns out that the
original Hebrew text does not either.
Unlike Classical Hebrew, Ancient Greek, the lingua franca in which the
New Testament had been written, had the noun ἐπαγγελία ‘promise’ and the
verb ἐπαγγέλλομαι, of which promise (as a noun or as a verb) in the King
James Bible constitutes a literal translation. As we saw earlier, the Samoan
translation of the New Testament (O le Feagaiga Fou) has sixty-three cases of
Samoan folafola matching all but two of the total sixty-five cases in which the
Greek text has ἐπαγγελία or the corresponding verb form. The Samoan text
does not have folafola in six cases. Of these, three do not have promise in
Greek.

4.5 Promises in non-Christian Samoan texts


A search of Krämer’s two volumes, which are rich repositories of stories and
accounts of cultural practices narrated in the Samoan language of the late
1800s, resulted in no example in which the word folafola or its compounds
have the meaning of ‘promise’ (verprechen in German) and no example of
any other Samoan term used with such meaning. When folafola is used it is
shown in a few cases with the basic meaning ‘spread out’ (German ausbrei-
ten) – including the semantically related adjectival mafolafola meaning ‘wide,
broad’ (German breit) – and in two cases (a few sentences apart within the
same narrative) folafola describes an action that Krämer adequately inter-
preted with the verb ausrufen ‘call out, exclaim’ (but his English translator
misleadingly renders with ‘called for’). The context of the narrative indicates
that the latter is the use of folafola for ‘announcing’ – in this case for goods
that have been exchanged – along the lines indicated above in the discussion
of example (3).
Promises in non-Christian Samoan texts 93

(16) (Krämer 1902: 111; the English translation is mine)


[. . .] Ona alu ane lea ‘o Pulele’i’ite,
Da ging Pulele`ite`ite hin und
‘then Pulele`ite`ite went (and)’
‘ua folafola le aofa’i o talo ma ‘ulu o le umu. FIRST EXAMPLE
rief die Zahl der Tarostücke und Brotfrüchte aus der Küche auf.
‘announced the total (number) of taros and breadfruit from the oven.’
Ona folafola fo’i lea ma ‘ato niu. SECOND EXAMPLE
Dann rief er auch die der Kokosnusskörbe auf.
‘then he announced the baskets of coconuts.’
Ona fa’apea lea ‘o le ‘upu a Pulele’i’ite:
Darauf sprach Pulele’i’ite folgende Worte:
‘here is the word(s) of Pulele`i`ite:’
Fa’alogologo mai la tatou malaga,
Hört, meine Reisebegleiter,
‘listen, our travel party,’
se’i ‘ou ta’uatua ia niu.
ich will euch etwas von den Nüssen sagen.
‘let me tell you about the coconuts.’

A search among the 110 Samoan texts collected by Oskar Stuebel (1896)
yielded two cases of folafola that are translated by Stuebel with the German
verb versprechen ‘promise.’ This at first seems to confirm that in the late
nineteenth century folafola was understood as corresponding to ‘promising’
also in non-Christian texts about Samoan customs, legends, and history.
However, a careful reading of the narrative context of these occurrences of
the term folafola suggests an alternative interpretation, whereby folafola could
be adequately translated with one of the other meanings discussed above, that
is, ‘put out,’ ‘announce,’ or maybe ‘offer,’ rather than ‘promise.’ I will here
discuss one of these two examples.
The text in question (entitled by Stuebel In welchen Fällen Land veräussert
werden konnte) is given as an example of cases (Fällen) in which a piece of
land (Land) might be transferred (veräussert) from one family to another (an
unusual situation given the typically inalienable relation between land and
titles in the Samoan descent system). The narrative is about a matai, who,
having found himself without enough food to properly welcome a visit (from
another village) by his sister or his sister’s son, goes to ask and obtains help
from another matai who gives him one of his animals (in the original German
translation, this is contextually understood to be a pig). After the traveling
party has left, the matai who had benefited from the generosity of the other
matai, after discussing it with his own extended family, goes to visit the one
who had helped him and offers him an outrigger (vaa) or, alternatively, a piece
(fasi) of land (fanua). The other matai rejects the boat and accepts instead the
offer of the piece of land for his own children’s benefit.
94 Promising in the Samoan translation of the Bible

As shown in (17) below, in the narrative collected by Stuebel we find two


instances of folafola. The first folafola, with the suffix –ina17 and followed
by the deictic particle atu ‘away from the speaker’ – the indexical origo
(Hanks 1992) – is translated by Stuebel as ‘offer’ (German: ich dir . . .
anbiete ‘I offer you’). The second folafola, which is followed by the deictic
particle mai indicating an action toward the speaker (or origo), is translated
instead as ‘that you have promised me’ (German: welches du mir verspro-
chen hast). In (17) I give the relevant passage extracted from the (longer)
original text with the original German translation and my own English
translation of the latter.

(17) (from Stuebel 1896: 128, 214–215)


e lua mea nei ua ou sau ma a`u
Ich bringe zwei Sachen,
‘I come with two things’
ou te folafolaina atu e tali a`i lau pule FIRST FOLAFOLA
welche ich dir als Gegengabe für dein Geschenk anbiete
‘(that) I offer you in return for your gift’
o Ie vaa a e Ie finagalo18 i ai
nämlich ein Kanoe, willst du es aber nicht
‘a boat, but (if) you do not want it,’
ou te avatu Ie fasi fanua e taui a`i lau pule
so werde ich dir als Gegenbage für dein Geschenk ein Stück Land geben.
‘I will give you a piece of land in return for your gift.’
Ua tali atu Ie alii po o Ie matai
Der Häuptling oder das Familienhaupt antwortet
‘the chief or family-head replied’
maliu mai ua faafetai
Sei willkommen, ich danke dir,
‘(you) are welcome, I thank you,’
a e tuu atu pea Ie vaa e aoga i ou ala
aber behalte das Kanoe, das für deine Söhne nützlich ist
‘but keep the boat (that) is useful to your sons,’
a e tuu mai o Ie fasi fanua ua e folafola mai nei SECOND FOLAFOLA
und gieb mir das Stück Land, welches du mir versprochen hast,
‘and leave me the piece of land that you have promised/offered,’
e aga`i atu i ai la`u fanau o alu pule lea.
Damit meine Kinder es in Besitz nehmen.
‘So that my children take possession of it.’

Even though the use of either German versprechen (or English promise)
for the second folafola found in this text makes sense, alternative transla-
tions are also possible and reasonable, including ‘offer,’ which was used
for the first instance of folafola in the story, as well as ‘put out, lay
out, announce, declare.’ In fact, if folafola was meant to correspond to
The case of the word māvaega 95

something akin to ‘promise’ (or the German versprechen), why not trans-
late it as such in the first instance when the speech act is performed?
Instead, it seems that folafola is turned into the verb for ‘promise’ (ver-
sprochen hast ‘you have promised’) in the metapragmatic account offered
at the end of the passage, which is a recasting of the action performed by
the chief who has come to compensate for the gift he had received in a
moment of need.

4.6 The case of the word māvaega


As mentioned above, a search in a number of transcripts of conversations and
fono meetings recorded in Samoa between 1978 and 1988 produced only two
cases in which a Samoan term was translated in English with the word
promise. One describes the famous announcement made by the MP, which
Loa reported (and got in trouble for). The second is found in the original
handwritten transcripts of the January 21, 1979 fono I audio-recorded and
annotated while still in Samoa during the same year. It is the English transla-
tion of the word māvaega made by one of the bilingual matai in Falefā who
was helping me interpret the fono proceedings. The word was repeated twice
by the same speaker. As shown in example (18), when I cited one of the
utterances with māvaega in my From Grammar to Politics book – to illustrate
the use of the ergative marker e for marking the (semantic) Agent noun phrase
(here the village of Lufilufi) (Duranti 1994: 128) – I did not follow the
translation given me by the native consultant and translated māvaega as
‘agreement.’
(18) (Fono, January 21, 1979; the senior orator (matua) Moe`ono is providing the
background that is meant to explain why he decided to run against the
incumbent MP, a matai from the nearby village of Lufilufi)
Moe`ono; oga `ua soli e Lufilufi le kākou māvaega.
‘because Lufilufi violated our agreement.’

I had a number of reasons not to follow the “native” translation of māvaega.


Most importantly, in (18) the noun māvaega co-occurs with the possessive
/kākou/ (tātou), the inclusive ‘our’ (or ‘we’), and since the village of Lufilufi is
the Subject-Agent of the sentence (marked by the ergative marker e), māvaega
refers to something that both the people of Lufilufi and the people of Falefā had
done. If Lufilufi had violated the promise, the possessive should have been
their and not the inclusive our.19 The inclusive our makes sense only if we
think of it as an “agreement” or “mutual commitment” and not as a promise by
one party (the people of Lufilufi) to another (the people of Moe`ono’s village,
Falefā).
96 Promising in the Samoan translation of the Bible

Māvaega is the nominalization – produced with the same –ga suffix found
in the noun folafolaga in example (2) – of the verb māvae ‘to be left, inherited’
and also ‘(be) apart, separate.’ As a noun, māvae also means a ‘crack.’ Māvae
itself is, in turn, a compound made of the verb vae or vavae meaning ‘divide,
separate’ and ‘cut’ (Milner 1966: 308). The connection among vae, māvae, and
māvaega is revealed by the meaning of māvaega as ‘farewell, departing
words’ (Milner 1966: 142), that is, something that happens at a moment of
separation. This is the reason for Milner’s translation of māvaega as an orally
delivered ‘will’ in the sense of a ‘parting promise.’ It is used to refer to a
chief’s legacy or, as explained in Krämer ([1902] 1994: 663), “an official
decree when [matai] feel their end is near, i.e. they officially state who is to be
their successor in name and title and what privileges the other children will
inherit.” This also shows that māvaega at least in its original sense could not
really be a promise because the person who is performing the act does so
before departing from this world and as such his decree or wish will have to be
realized by others (see Krämer [1903] 1995: 110). The conditions of
satisfaction for the act of promising as described by Searle do not apply
because the speaker will not be able to fulfill the commitment.

4.7 Promises, oaths, and public announcements


The linguistic analysis presented so far should not be interpreted as implying
that Samoans in the past did not have ways of expressing commitments or did
not care about whether an action was carried out as stated, announced, or
reported. There is plenty of evidence that as individuals and in groups, like
other people in the world, Samoans were in the past and are today accountable
to each other for their actions, words included. During my stays in Samoa
I witnessed a recurrent concern with whether something really happened in the
past or will happen in the future. In fact, truth is one of the main concerns of
the formal proceedings like the fono (see Chapters 3 and 5). Truth is also a
concern outside of the fono, in more mundane exchanges. If someone says
something that is suspicious, friends are quick to confront the alleged liar with
expressions like `aua le pepelo! ‘don’t lie!’ (found in my transcript “The
Inspection,” December 1978) or with accusations like pepelo ia! ‘(he/she/
you) really lied!’ Being caught lying is an embarrassment for children as well
as adults even though people might find it hilarious at times to discover that
someone tried to get away with lying.
And of course, storytelling in general and gossip (faitala) in Samoa, like
elsewhere, are concerned with lessons learned from all kinds of social miscon-
duct, including lying or not fulfilling a social expectation. The word pepelo can
be used regardless of the intentions of the person who is being accused of
lying. For example, if I had been asked whether I was going to drive to the
Promises, oaths, and public announcements 97

capital and said that I was not but I had been later seen driving back from the
capital, I would be accused of pepelo without being given a chance to explain
my actions on the basis of changed circumstances – this speech act was often
accompanied by a smile, implying a certain pleasure for having been able to
catch me lying.
A Samoan concern with finding the truth was recorded in the nineteenth
century by the European missionaries and other visitors. In his Nineteen Years
in Polynesia: Missionary Life, Travels, and Researches in the Islands of the
Pacific (1861), the missionary George Turner recounts an episode that illus-
trates this concern by describing a stratagem for finding out who did a given
(bad) deed (in this case a theft).
On looking out one afternoon, I saw all the grown-up people of the village coming and
sitting down before the door. They all looked very demure, and I wondered what was
up. Presently one of the old men commenced speechifying. “We have been talking
about your horse which has got a lame foot, and which is supposed to have been stoned
by some one. We wish to know who has done it, but all deny, and we cannot find out. It
is our custom when anything is concealed, for all to assemble and take an oath. That is
our plan. Will you please to hand out a Bible, and let us all swear here, that we may
know who is the guilty party?” It was their custom thus to assemble, and each laying his
hand on the sacred stone, or shell, or cup, which might be considered the representative
of the god, to implore vengeance and speedy death, if he touched the stone and told a
lie. Of course I thanked them for their respect for my nag Tom, but told them that such
imprecations were wrong, and that the simple yea or nay in such a case was quite
sufficient. They were satisfied, and by and by it appeared that the horse was lame not for
a stone, but from rheumatism. (Turner 1861: 24)
We do not know the possible outcome of the proposal made by the people
mentioned in this episode, but further evidence of old Samoan stratagems to
uncover the culprit of a theft is found in one of the texts collected by Stuebel
(1896: 130, 217): O le tauto faasamoa ‘the Samoan oath.’ The custom is also
described in Krämer (1995: II, 106), who might have lifted it from Stuebel and
Turner. Milner (1966: 257–258) also mentions the same Samoan practice for
fact finding in translating the verb tautō and its nominalization tautōga ‘oath’
(as a noun). The use of this lexical item and its compounds fa`atautō (with the
causative prefix fa`a-) and fa’atautōina (with both the causative prefix fa`a-
and the so-called “transitive” suffix –ina) are shown to be used for the taking
of an oath by an official (this is also the verb I mentioned earlier for the
marriage vows). Here and in all of the examples that I found in the older
secular texts, commitments, regardless of the English translations we might
agree upon, have one common characteristic: they are made in public, with
multiple parties as witnesses. The word promise in English, of course, can
have this public usage, like when politicians make promises to the voters (Hill
2000). But the English promise also has a possible use that is more private, as
demonstrated by expressions like I promised myself, which describes an act
98 Promising in the Samoan translation of the Bible

that happened as part of an interior dialogue. Has the use of folafola extended
to such speech acts? Given that I could not find examples of such a use of
folafola in either the data I collected or elsewhere, I asked two young Samoan
friends (via email) whether one could say in Samoan “I promised myself
I would stop smoking” and similar utterances. They answered that it is possible
and each gave me a different translation. It is relevant for our discussion that
one person used folafola20 but the other did not accept it as an appropriate
translation because she interpreted folafola as ‘declare,’ as opposed to ‘prom-
ise.’ She proposed instead to use the word tautō,21 which I mentioned above as
meaning ‘oath.’ Even though one must be careful when generalizing on the
responses of two speakers, the fact that two bilingual speakers (who, I should
add, grew up in the same household) could have different intuitions about how
to render ‘I promised to myself’ suggests that folafola is far from solidly
established as the Samoan translation of English promise even though such a
meaning is found in the Bible, which was translated in the nineteenth century,
and in Milner’s dictionary (1966). It is thus possible that the innovation
constituted by the use of folafola for ‘promise’ may still be in progress.

4.8 Conclusions
In this chapter I have taken on the challenge of corroborating my earlier claim
that there is no special word corresponding to the English promise in Samoan
by engaging in a detailed analysis of two terms, folafola and māvaega, that
have been proposed as a translation of the English promise in dictionaries, in
the Samoan translation of the Bible, and by the Samoan speakers who helped
me translate my recordings of Samoan speech in spontaneous interaction.
I have shown that the basic, prototypical meaning of folafola is that of a
physical action roughly corresponding to the English ‘(to) put,’ ‘lay,’ ‘spread
out,’ and that by metaphorical extension it can be used to mean ‘(to) spread,
lay out information,’ ‘spread the word,’ ‘announce,’ ‘declare,’ and even
‘explain.’ On the basis of my review of texts produced or collected in the
second half of the nineteenth century, namely, the Samoan Bible or Tusi Paia
and Stuebel’s and Krämer’s volumes dedicated to Samoan legends, stories,
beliefs, and various cultural practices, I have hypothesized that the notion of
promise was at first introduced into Samoan in the process of translating the
Christian Scriptures and especially the New Testament, where folafola is
frequently used to translate expressions meaning ‘promise’ in the original
Greek text (and rendered with the English promise in the King James Bible).
The data gathered and analyzed in this chapter suggest a potential scenario
of historical change for the lexical item folafola, which went from meaning
‘spread out, distribute’ to ‘announce’ (a native activity) and later to ‘preach,’
an unknown activity in pre-Christian Polynesia, and ultimately to ‘promise,’
Conclusions 99

probably in the process of looking for a native Samoan term that could render
the many occurences of ‘promise’ in the Bible, especially in the New Testa-
ment. This proposed reconstructed semantic change is summarized below in
terms of four stages.
Stage 1: folafola means ‘lay out, spread out,’ especially flat and flexible
things such as mats or goods.
Stage 2: ‘spread out’ ! ‘distribute,’ a native metaphorical extension out of
several traditional practices, including the display on the floor of
the house or in the ceremonial ground (malae), what one group is
giving to another, and the distribution of the kava drink (`ava)
during ceremonial or formal occasions.
Stage 3: ‘spread out (goods)’ ! ‘spread information, announce,’ e.g., during
ceremonial exchanges where someone will let those outside the
house know what has been given or during a kava ceremony where
the person who is in charge of ‘distributing’ the kava is also the one
who ‘announces’ who is getting the next serving or ‘cup’ (ipu).
Stage 4a: ‘spread information, announce’ ! ‘preach,’ in the Christian sense
of announcing the arrival of the “Good News” (Tala Lelei) and
spreading the Christian faith.
Stage 4b: (potentially simultaneous with 4a): ‘announce, declare’ ! ‘prom-
ise,’ especially to refer to God’s promises to his prophets and to the
people of Israel.
The metaphorical extension of folafola to describe a personally felt commit-
ment to a future action most typically between two parties seems to be
consistent with other semantic innovations introduced by the missionaries,
including the notions of ‘soul’ understood as the inner, spiritual, immaterial
aspect of human life and experience.
Despite the widespread use of folafola to render promise in the New
Testament, it is possible that this word may still not have the full illocutionary
force of the speech act of promising in contemporary Samoa. For example, the
current vow at a marriage ceremony is the expression ou te folafola ma
ta`utino atu, a complex phrase that includes the verb ta`utino, a compound
made out of ta`u ‘speak, say’ and tino ‘body,’ translated as ‘declare plainly’ by
Pratt (1893: 289) or ‘speak out, speak frankly’ by Milner (1966: 250). The
whole phrase would then literally mean something closer to ‘I announce and
declare’ or ‘I announce and frankly say.’ This suggests that, for such an
important occasion, folafola is not considered sufficient for conveying a
spouse life-long commitment.
Overall, the data analyzed in this chapter support the view presented else-
where in this book that the linguistic expression of a private intention to
commit to doing something in the future might not fit with the traditional
100 Promising in the Samoan translation of the Bible

Samoan propensity to focus on the visible effects and consequences of actions


and feelings rather than on an individual’s inner thoughts (see Chapter 3). As
we saw in §4.3.1, even the notion of agaga, which was used by the mission-
aries to translate soul, spirit or ghost as in Holy Ghost (Agaga Pa`ia), was
understood as something of a physical and, thus, visible nature.
Of course, languages do change over time, and it is difficult to know when
we can be sure that a given change, especially a semantic or pragmatic one,
happened only or mostly because of contact due to a group of outsiders coming
into a community and introducing a new metaphysics, e.g., a new religion. But
as shown by Hanks’ (2010) remarkable analysis of the process of reducción
among the Yucatec Maya, linguistic anthropologists have powerful tools to
illuminate how cultural innovation takes place, especially when we are fortu-
nate to have access to the historical record of written documents that can be
compared with current linguistic practices. In addition to contributing to our
discussion about intentional discourse in Samoa, this chapter offers an exem-
plification of a method of argumentation and data gathering that others inter-
ested in historical change could adopt, improve upon, and extend in new
directions.
5 Intentionality and truth, revisited

5.0 Introduction1
In the previous chapters I have reviewed the anthropological critique of the use
of intentions in speech act theory and presented some data that further question
the universality of a theory of human action mostly, if not exclusively, founded
on intentions. My move to reduce or at least contain the role of an individual’s
intentions in social theory should not be understood as a return to behaviorism
or a flat rejection of cognitive theories that rely on the reconstruction of
rational inferential processes. Rather, my intellectual engagement with the
literature, on the one hand, and with analysis of spontaneous verbal exchanges,
on the other, has been motivated by the need to show that in some situations
participants do not seem to invoke intentions as much as a certain type of folk
psychology might lead us to expect. Furthermore, a historical and cross-
linguistic perspective also shows that the disembodied notion of intention
found in the philosophical literature is not a universal. These data raise the
issue of whether a rethinking of intentionality in anthropology might affect the
ways in which we write about culture. Once we start to examine this issue, we
realize that it is difficult to talk about intentionality without also rethinking the
notion of truth that is adopted in ethnographic accounts.

5.0.1 Anthropological thinking and writing


The very process of doing ethnography often requires asking others questions
regarding their own or others’ intentions, beliefs, or motivations. This means
that, in the act of getting to know our subjects, we cannot but be involved in
assessing the truth of their assertions, the sincerity of their expressed feelings,
and the relationship between what they say they are up to and what we see as
their unfolding actions. Intentions may or may not be a topic of discussion, but
the relation between intending and doing fulfills claims of moral responsibility,
one of the domains of human experience where ethnographers are more likely
to feel different from the people they live with and study. It is through the
discussion of these intersecting experiential dimensions that we come to define

101
102 Intentionality and truth, revisited

the uniqueness of others’ ways of being-in-the-world as well as their humanity.


Unfortunately, when we go and look for traces of these interpretive processes
in anthropological writings, we find little information or discussion. Ethnogra-
phies rarely give us the amount of detail (or the meta-comments) one would
need to assess exactly how truth was determined, agency represented, inten-
tions taken into considerations, and responsibilities assessed. At best, we are
given summaries of some of these processes, but the inferential processes
remain hidden and the data on which generalizations are based are not avail-
able for scrutiny. For text-oriented anthropologists like Dennis Tedlock
(1983), it is the dialogic source of ethnographers’ generalizations that is erased
in the process of writing about culture. Knowledge comes from conversations
that have been either heard or initiated by the fieldworker. Any tension,
disagreement, contestation, or approximation that emerged in the interaction
among subjects as speakers of culture is eventually removed in the interest of
producing ethnographic descriptions or glosses that can fit into an established
field of anthropological knowledge and a genre of writing about culture and
the Other. For cognitive anthropologists like Dan Sperber, it is the lack of what
he calls “descriptive comments” that makes it difficult to assess the truth of
what is being described. Ethnographers end up being “entangled in . . . pseudo-
concepts and pseudo-questions” that make it difficult or impossible to reach
consensus on a particular cultural phenomenon, let alone the goals of the
discipline of anthropology as a whole (Sperber 1985: 30).
These issues are deeply philosophical. And yet, confronted with philosoph-
ical texts that they consider unempirical and culturally deprived characteriza-
tions of human thought and human action, most ethnographers have avoided
entering debates in the philosophy of mind or philosophy of language fields.
For example, the publication of Austin’s (1962) theory of speech as social
action initially inspired a few ethnographers to engage with Austin’s frame-
work in a constructive way (e.g., Finnegan 1969), but this early interest
quickly faded away as ethnographers convinced themselves that analytic
philosophers were too “western” in their ways of thinking about human action,
utterances included, and that little was going to be gained by borrowing their
models of human thinking and acting.
It was in the context of this intellectual climate that, in the late 1980s,
I decided to enter the discussion and examine the use of truth in anthropo-
logical discourse. In representing my 1993 article on truth and intentionality
(originally written in 1990) as a chapter of this book, I need to mention that
formal training in linguistics had primed me to think of “philosophy of
language” almost exclusively as “analytic philosophy of language.” This might
help clarify my framing of the problem of the study of truth and intentionality
against the backdrop of a dualistic model that opposes mind and world, as well
as Subject and Object and Self and Other. In fact, this dualism – together with
Introduction 103

the idea of language as the “representation” of an already constituted reality –


was refuted by at least two philosophical traditions, American pragmatism
(e.g., James 1910; Rorty 1982) and post-Kantian “Continental Philosophy”
(see McCumber 2011). Hegel temporalized truth, starting with his quick but
emblematic discussion of the changing meaning of expressions like “this,”
“now,” and “here” (or their German equivalents) in his Phenomenology of
Spirit, and others followed in his steps, including Heidegger, who criticized the
“agreement” characterization of truth and reinterpreted truth (Greek aletheia)
as “uncovering”:

To say that an assertion “is true” signifies that it uncovers the entity as it is in
itself. Such an assertion asserts, points out, “lets” the entity “be seen” . . . (Heidegger
1962: 261)

In the 1980s this type of “continental” critique had been absorbed and reformu-
lated by French poststructuralists and deconstructivists and as such had also
entered the anthropological discourse, partly through the work of Bourdieu,
Gibson, and a few others. Among various proposals for new methods, like, for
example, “dialogical anthropology” (Tedlock 1983), the questioning of the
essence of truth – and of “essence” more generally – soon became a rethinking
of the mission of the anthropologist as a searcher of truth. This meant that in
the 1980s and early 1990s the very practice of ethnography, the building block
of the entire discipline, was being questioned, reanalyzed, or recontextualized.
This soul-searching produced a considerable body of self-reflexive writings,
including a series of articles, monographs, and edited collections dedicated to
participant observation, fieldnotes, and the practices through which anthropo-
logical descriptions are produced and made authoritative (e.g., Rabinow 1977;
Sperber 1985; Clifford and Marcus 1986; Clifford 1988; Geertz 1988; De Vita
1990; Sanjek 1990; Fox 1991). During that period, we witnessed an epistemo-
logical revolution, in which the observers became the observed and the modes
and goals of the enterprise – a supposedly objective or at least detached
comparison of different ways of being “human” – came under attack (e.g.,
Fabian 1983; R. Rosaldo 1989; Abu-Lughod 1991; Sheper-Hughes 1993).
Whether or not one agreed with the meta-statements and ontological questions
brought about or inspired by this new wave of critical anthropology, it became
problematic for many fieldworkers to continue to carry out their research in the
traditional way. For some, but not all, the solution was to get away from
studying the Other and turn to study their own society.
It was in the middle of such a crisis of anthropological thinking that linguists
in anthropology departments abandoned their own intellectual ghettos and
moved beyond the important but limiting roles as teachers of phonemic
analysis, collectors of lengthy native texts, or guardians of the still poorly
understood (and never clearly formulated) “Sapir-Whorf hypothesis” (Hill and
104 Intentionality and truth, revisited

Mannheim 1992; Lucy 1992b). In the course of this paradigm shift (see
Duranti 2003a), the linguists’ subjects graduated from “informants” to talking
social agents whose conversation-actions could be recorded in the same way in
which their postures could be (but usually had not been) photographed or
filmed. Although these techniques and the analytical procedures that accom-
panied them did not make participants’ verbal contributions easily interpret-
able or our ethnographic accounts more objective in any naïve sense of the
term (Duranti 2006b), they made them – at least in the spirit in which linguistic
anthropologists were working – more open to further inquiry. By routinely
recording native discourse, we discovered that participants were “naturally”
inclined, to borrow Michael Moerman’s (1988) apt metaphor, to talk their
culture not only in interviews, but also in their everyday talk. In the midst of all
kinds of social situations, people could be heard (and sometimes seen) making
sense of their own behavior to themselves and to each other through accounts
of past, present, and future actions. In one contribution after another, linguistic
anthropologists showed that a theory of culture (or of “the cultural” as some
started to say) can be expressed not only in the symbolic oppositions found in
local taxonomies and ritual performances – and sometimes captured in state-
ments given in open-ended interviews – but also in the words and turns
exchanged among people while teasing, arguing, instructing, gossiping,
joking, or telling and co-telling narratives of personal experience (e.g., Bren-
neis 1984; Ochs and Schieffelin 1984; Briggs 1986; Haviland 1977, 1986,
1989; Besnier 1989, 2009; Ochs, Smith, and Taylor 1989; Schieffelin 1990;
Watson-Gegeo and White 1990). It was in this intellectual climate that I was
exploring the relevance of dominant conceptualizations of intentionality and
truth for the production of anthropological knowledge.

5.1 Truth as correspondence between the mind and the world


Veritas est adaequatio rei et intellectus
According to Aristotle, the essential nature of truth is a correspondence
between human judgment (in the mind) and things (in the world).
As pointed out by Franz Brentano (1966), Aristotle’s definition of truth is
found not only in the medieval doctrine that “veritas est adaequatio rei et
intellectus” (which can be roughly translated as “truth is the correspondence
between an object (thing) and a state of mind”), but also in the Logic of Port
Royal (1660) and in Kant’s Critique of Pure Reason, where we read: “The
nominal definition of truth, that it is the agreement of knowledge with its
object, is assumed as granted . . .” (Kant 1965: 97) and “[i]f truth consists in the
agreement of knowledge with its object, that object must thereby be distin-
guished from other objects; for knowledge is false if it does not agree with the
Truth as correspondence between mind and world 105

(some state of) Mind (some aspect of the) World

Figure 5.1 Truth as correspondence between mind and world

Proposition State-of-affairs
Figure 5.2 Interpretation as correspondence between a proposition and a state
of affairs

object to which it is related, even though it contains something which may be


valid of other objects” (1965: 97). Figure 5.1 schematically represents this
view of truth.
If we substitute the notion of judgment or mind (intellectus) with that of
language or proposition or simply extend the Latin definition to include
enunciationis ‘statement, proposition’ (veritas est adequatio rei et intellectus
sive enunciationis), we find that the classic view of truth is at the basis of the
notion of interpretation in modern logic, as schematically reproduced in
Figure 5.2.
In the logical tradition originally inspired by Wittgenstein’s Tractatus
([1922]1961) and further articulated by members of the Vienna Circle (e.g.,
Carnap 1947) and the school of thought known as logical positivism, to know
the meaning of a proposition means to know the conditions under which that
proposition would be true (Wittgenstein 1961: 4.024). The same view, in a
slightly different language, is found in contemporary “possible worlds”
semantics. Thus, for instance, for Stalnaker (1984: 2), “A proposition is a
function from possible worlds into truth-values.” In this tradition, to under-
stand a proposition means to be able to match the states-of-affairs represented
by an utterance with their respective truth-values in specific “worlds” or
domains of discourse. Some propositions – representing what Leibniz called
“necessary truth,” Kant “analytic truth,” and Carnap “logical truth” – are
always true (and their negation always false). Examples of this type of
proposition include hammers are tools, squares have four sides, all bodies
take up space. These are all cases (of true propositions) in which the concept
106 Intentionality and truth, revisited

expressed by the predicate (are tools, have four sides, and take up space) are
contained or rather entailed by the (denotational) meaning of the subject.
These are also propositions that are not based on experience. The majority of
the propositions expressed through our everyday talk, however, are not of
this kind. They tend to be based on experience. They are contingent truths
like, for example, the mayor of Los Angeles is a man, the French president is
a socialist, most Americans buy foreign-made cars, etc. By reading (or
hearing) such statements we do not immediately know whether they are true.
They are not necessarily true. Since there is nothing intrinsic to the concept
of mayor that entails the gender identity of the person who holds that office,
before we can ascertain the truth of such statements we need to obtain some
additional information that is not contained in the meaning of the subject
noun phrase.
Whether we are dealing with analytic (a priori) truth or with contingent
truth, this view of meaning assumes two sets (and kinds) of entities: (1) states-
of-affairs (in particular worlds) and (2) their representations. When the two
match or there is a (positive) correspondence between them, we can say that
the proposition expresses a true proposition. If we use propositions, we can say
that there are only two possible relations between a proposition and a given
state of affairs (see Figure 5.3).
When we look at actual social interactions, such a binary view of truth may
be useful as an approximation or starting hypothesis, and in some cases as a
bare summary of some final resolution or judgment, but it cannot adequately
account for how people make sense of what is happening or what is being said.
Truth is typically a social matter because of the ways in which people work
around it and through it. This is so not simply because we need conventions
and public criteria for the assessment of truth, but because in social life truth
itself becomes an instrument, a mediating concept living in particular practices,
through which important social work gets done. This at least is the kind of
“truth” that ethnographers are more likely to encounter in their mingling with
others, in their professional accounts and recounts of what happened last night
or a hundred years ago, when our subjects’ great-grandfathers (or someone of
their generation) ruled the country, or the village. The truth discussed in public
arenas, for instance, is debatable not only because there are parties representing
and pushing for different versions of it, but also because it is through the
confrontation of different versions of truth that a collectively acceptable reality
is searched for, one where divisions and differences will not be resolved for
ever but can momentarily rest, until the next social drama (in Turner’s [1974]
sense) breaks out. As Lamont Lindstrom (1992) argued in his work about
debates in Tanna (Vanuatu), anthropological inquiry shows that truth is not
just a goal in such public confrontations; it is a criterion through which power
is both defined and claimed.
The Reconstructionist model of interpretation 107

5.2 Intentionality as a relation between the mind and the world


As discussed in Chapter 2, according to the classic definition introduced by the
Scholastics and further elaborated by Franz Brentano and Edmund Husserl,
intentionality is the property of human consciousness to be “directed toward”
or “be about” something (hence the English term “aboutness” used as a
synonym of “intentionality” by some authors writing in English). We can
attend to the same referent (e.g. a person, a thing, an idea) or the same state-
of-affairs (e.g., event, situation) as an entity that we love, hate, need, want,
admire, feel curious about, despise, miss, judge, recognize, remember, feel
longing for, and so on. Intentionality is the human faculty that makes possible
these many ways of relating to a given entity in the imaginary or real world.
However, unlike truth, which implies only two values, true and false (see
Figure 5.3), intentionality implies a potentially unlimited set of relations
between the mind and the object or situation. Thus, as shown in Figure 5.4,
the same thinking/feeling/perceiving Subject can at different times entertain a
variety of intentions toward the same entity.

5.3 The reconstructionist model of interpretation


Despite their differences, the dualism of these views of truth and intentionality
has led to interpretive theories that are about reconstructing the Subject–Object
or Language–Referent relation also in terms of a dualistic universe. What is
lacking in any model based on a mind–reality dualism is a recognition of the
ability that humans continuously display to utilize and reassess a variety of
natural and cultural resources around them not only to produce descriptions or
induce actions, but also to sustain a complexity of perspectives that are the
bread and butter of everyday cognition and interaction. One contrast here is

True

Proposition State-of-affairs

False
Figure 5.3 Two possible relations between a proposition and the state-of-
affairs it describes
108 Intentionality and truth, revisited

Subject A Object G

Figure 5.4 The same Subject (A) can be directed toward the same Object (G)
through different intentional acts, e.g., admiration, fear

between a basic folk psychology that sees human behavior and hence social
action as produced by mental beliefs and attitudes (Greenwood 1991) and a
more experiential, interactional, and ultimately ambiguous notion of humanity
(in the sense of “being-human”), a notion that is not reducible to individual
states of mind and/or propositions about them.
In the rest of this chapter I will address some of these issues, with particular
emphasis on those that arise when we widen the context of inquiry and start
thinking in terms of the larger human practices in which acts of intentionality
and judgments of truth-value are embedded. My review of these problems is
by no means exhaustive of the issues implied in past and current theories of
truth and intentionality. It should thus be taken as an attempt to clarify or in
some cases simply bring together for the purpose of reflection a number of
objections and proposals that have recently been made within a number of
fields concerned with what might be thought of as “interpretation-in-practice,”
with linguistic anthropology being one of such fields but not necessarily the
only or most important one. As a way of organizing what is a rather intricate
set of issues, I focus on four dimensions of the interpretive process that are
often ignored or underestimated in the reconstructivist and dualistic model of
interpretation:
1. the medium (or code) through which symbolic acts (representations
included) are performed;
2. the audience or co-participants in the interaction which produces
symbolic acts;
3. the cultural context, local theories included, which gives meaning to the
symbolic acts by locating them within larger or related activities and
conceptions;
4. the actions constituted through symbolic mediation.
Problems with goals, plans, and intentions-to-do 109

I will start by discussing some of the limitations of the reconstructivist model


of interpretation, and I will then continue with a brief examination of some of
the communicative and interactional dimensions usually ignored in the truth-
functional model.

5.4 Some problems with goals, plans, and intentions-to-do


There is no question that human activities are characterized as goal oriented
and as such requiring the actors’ ability to focus upon real or imaginary objects
as they plan or try to act upon such objects. This is the original meaning of
intentionality mentioned above. Knowledge implies objects of knowledge; to
believe means to believe something, to perceive also means to perceive
something, to want implies the existence of an object of want, etc. Despite
disagreement among different philosophical traditions as to the ontological
status of the Subject (the one who perceives, knows, wants, etc.) and of the
Object (the perceived, known, wanted, etc.), it is generally accepted that
human beings are generators, creators, or experiencers of intentional acts and
that such intentional acts constitute an important element of human action
(praxis) in the world. Historically, the recognition of intentional states or
human intentionality has been important in establishing the relation between
human actors and the world as quite different from the relation between things
and events in the natural world.
Our ability to think, imagine, believe, want, need, like, dislike, pretend,
analyze, etc. is instrumental for our capacity to propose new (complex) sets
of relationships among ourselves, the environment, and other humans.
Depending on the tradition of study, however, this process has been discussed
in rather different and sometimes limited terms. Regardless of the discipline –
logical semantics, linguistics, psychology, or artificial intelligence – we find in
the mathematical and social sciences a continuous attempt at isolating inten-
tional acts or intentional states from the larger contexts and activities within
which acts of intentionality are realized. This applies not only to the philoso-
phers of the Enlightenment, but also to praxis-oriented thinkers. Thus not even
Marx could escape celebrating human actors’ ability to foresee and execute a
goal on their own:
[W]hat distinguishes the worst architect from the best of bees is this, that the architect
raises his structure in imagination before he erects it in reality. At the end of every
labour-process, we get a result that already existed in the imagination of the labourer at
its commencement. (Marx 1912: 198)

It is hard to find here the (also) Marxist notion of praxis as both subjective and
historical and the notion of labor as a collective, joint activity, which will play
a role in Leont’ev’s (1981) theory of the phylogeny of human consciousness.
110 Intentionality and truth, revisited

Marx’s rebellion against crude materialism strangely aligns him here with
those who explain human praxis, interpretive processes included, in terms of
individual conscious acts (cf. Dumont 1977 for an anthropological critique of
Marx’s view of the individual). For cognitive psychology, the next step is
often inevitable: the reduction of the sociocultural context in terms of those
features that can be represented in the subject’s mind and through individual
cognitive categories. In this tradition, which characterizes most of so-called
“cognitive models” of human action, the predominant starting point is the
western notion of person as unit of analysis. Grounded in the western rational-
ist tradition, interpretation is understood as a process which focuses on an
individual’s mind as the meaning-making organism and on an individual’s acts
as the reflections or consequences of his states of mind (see Bilmes 1986 and
Heritage 1990/91 for a criticism of this tradition from an ethnomethodological
and conversation analytical viewpoint). It is in this context that we must
evaluate the fundamental role of individual intentions in contemporary philo-
sophical and psychological theories of communication. For speech act theory,
for instance, to understand the meaning of someone’s words means to
understand his or her intentions as expressed through conventional signs
(e.g., a common language) (Grice 1957; Searle 1969, 1983). As mentioned
in Chapter 2, for Dennett (1987) the assumption of intentionality in a system is
the only way of making predictions and explaining that system’s behavior in
rational terms. Such a view is appealing for its apparent simplicity, but it runs
into serious problems as soon as we leave the logic of western folk psychology
and take on a cross-cultural perspective. The reconstructionist view of inter-
pretation is but a corollary of the western universalist perspective criticized by
Dumont (1964, 1970, 1977). In the dominant western ideology – deeply rooted
in Descartes’ dualism – the individual actor is the main agent of social insti-
tutions and the point of reference for a universal ethics. We can guess what is
in another’s mind because it is assumed that our beliefs, wants, and feelings, as
well as the notions (and practices) of good and evil are (or should be) the
same as those of others. We are then left with the empirical question of
whether such concepts can be used to understand non-universalist social
systems, e.g., India’s caste system, which prescribes to each individual his
or her own specific place, duties, and rights based only on membership in a
particular group. Hence, the importance of studies that examine such non-
egalitarian and non-universalistic social systems as they experience pressure
from inside (e.g., from groups advocating women’s rights) or from outside
(through pressure brought about by a global flow of information). As anthro-
pologists we need to pay special attention to the socio-historical limitations
and constraints that shape the meanings of actors’ actions and hence, in a
sense, fashion their subjective intentionality. One way of understanding the
role of such constraints is to consider how individual intentionality is affected
Code-related limits on speaker’s intentions 111

by the limits inherent in any system of representation, natural languages


included. The tendency in the past has been to think in terms of cognitive
constraints (e.g., Whorf 1956; Lucy 1992a, 1992b). We have now moved to a
more inclusive and complex notion of what “language” is. Thus, the possible
role of a given genre or register, or simply the turn-taking system appropriate
to a given event, may guide speakers’ forms of participation and thus their
acts just as much as or more than the lexical or morphological choices
discussed by those interested in older versions of linguistic relativity (Duranti
2011). An example of the influence of genre is found in Chapter 7 of this
book, where readers will be able to appreciate the extent to which the
conventions of email messages either afford new content to manifest itself
or are simply a rephrasing (or translation) of what can be already found in the
other chapters of the book.

5.5 Code-related limits on speaker’s intentions


A basic question that must be faced by any theory of intentionality is the role
of the individual vis-à-vis his or her community (or tradition) in intentional
acts. Although it is perhaps easy to imagine or accept simple intentional acts –
e.g., the intending of specific physical (or even imaginary) objects, e.g.
anticipating the offer of food by our host, looking for a person in a crowd,
trying to remember a person’s name – it is difficult to assess the extent to
which an individual can have complex intentional states that are independent
of their semiotic representation, whether verbal or otherwise. This is in a sense
the old question of the “language of thought”: can we engage in complex
reasoning without language (or without a language)?
Structural linguistics has often emphasized how any semiotic code, natural
languages included, imposes a system of choices – a finding that was, for
example, at the basis of Ferdinand de Saussure’s distinction between syntag-
matic and paradigmatic relations (Saussure 2005). Linguistic anthropologists
in the first half of the twentieth century extended this view to claim that the
very use of a code implies the acceptance of the classificatory and conceptual
system that gives meaning to each of its elements (Sapir 1933; Whorf 1956).
This implicit “contract” between users and their code would take away control
from the speakers and give it to tradition and culture, through “habituation”
that makes something we say “highly probable” or “virtually unavoidable”
(Hanks 1996: 238), either triggered by or feeding into a particular social order.
Despite the obvious existence of alternative choices and acts of resistance to
hegemonic codes, as demonstrated by sociolinguists’ discovery of covert rules
or preferences (Trudgill 1972), the basic thesis here is that symbols, like tools –
and hence both symbolic and material culture – can empower their users but at
the same time place constraints on both forms of behaviors and forms of
112 Intentionality and truth, revisited

interpretation. This seemed obvious to the student of language and culture


Edward Sapir who wrote:

It is strange how frequently one has the illusion of free knowledge, in the light of which
one may manipulate conduct at will, only to discover in the test that one is
being impelled by strict loyalty to forms of behavior that one can feel with the utmost
nicety and can state only in the vaguest and most approximate fashion. (Sapir [1927]
1963: 549)

Similar considerations were made about the speaker’s subjective intentions


within historical materialistic views of the communicative process. It was
suggested that intentions can be recognized (by the addressee(s) first and by
the analyst later) only to the extent to which they conform to the expected or
expectable intentions in the ongoing or reconstructed context. Thus, Vološinov
wrote:

The individual motives and intentions of a speaker can take meaningful effects only
within limits imposed by current grammatical possibilities on the one hand, and within
the limits of the conditions of socioverbal intercourse that predominate in his group on
the other . . . No matter what the intentions of the speaker means to carry out, no matter
what errors he may commit, no matter how he analyzes forms or mixes them or
combines them, he will not create a new pattern in language and he will not create a
new tendency in socioverbal intercourse. His subjective intentions will bear a creative
character only to the extent that there is something in them that coincides with
tendencies in the socioverbal intercourse of speakers that are in process of formation,
of generation; and these tendencies are dependent upon socioeconomic factors.
(Vološinov 1973: 143)

As a reformulation of linguistic determinism, these statements are suspicious


(Chomsky 1973; Cole and Scribner 1974; Hill 1988). However, if we redefine
“code” to be a dynamic set of discourse patterns, which includes but goes
beyond what is usually meant by grammar or lexicon, we might start agreeing
with Vološinov that the code may indeed be an important factor in shaping
intentions or possible interpretations. Here “code” must be seen as ways of
speaking (Hymes 1974), that is, a set of semiotic resources (e.g., genres, styles,
registers, verbal routines) to rely upon in constituting interactions or socio-
cognitive activities – as opposed to a binary set of form–meaning
correspondences. It is in this vein that language has been seen as an important
socializing factor (Ochs 1988; Ochs and Schieffelin 1984; Duranti, Ochs, and
Schieffelin 2012). A mother’s or an older sibling’s explicit interpretation of a
child’s attempt to communicate is de facto (also) an instruction on what is a
“possible message” in that context and on how that message fits within an
actual world of social relations, kinship, etc. Furthermore, worldwide recurrent
use of verbal routines with young children such as ‘say x’ or ‘tell so-and-so
that x,’ etc. indicates specific areas where culturally valued messages and
Code-related limits on speaker’s intentions 113

relationships are communicated and shaped through language use (Duranti,


Ochs and Schieffelin 2012; Schieffelin 1990; Schieffelin and Ochs 1986b).
Chomsky’s insistence on the “creative uses” of language by the child is
compelling when compared with Skinner’s behavioristic model of verbal
behavior (Chomsky 1959) but becomes limiting for any research program
interested in what actually happens to children while they are acquiring a
language. As Hymes (1972b) made clear through his notion of communicative
competence, a child cannot produce any kind of meaning he or she wishes to
communicate even though it might be grammatical (or, in Chomsky’s terms,
well formed). There are linguistic routines that inform the activities in which
people are involved and the meanings that are possible, acceptable, and
interpretable. Borrowing a powerful term from Gibson’s (1986) ecological
approach to perception, it might be useful to think of semiotic codes as systems
of affordances. For Gibson, any medium (e.g., air, water) interacts with an
organism in such a way to favor certain kinds of behaviors and disfavor others.
Thus, for instance, air affords breathing for humans but not for fish, the soil
affords walking by a heavy animal, and water affords swimming. The func-
tions that the environment (or medium) allows for a given organism are called
affordances. We could then say that certain linguistic conventions or frames
afford particular speech acts and require particular felicity conditions that, in
turn, make some interpretations possible and acceptable. Such interpretations
also open up the possibility of further social action.
Another way in which the code in its broader sense overrides subjective
intentions or motivations is on those occasions in which a highly ritualized
speech may reduce the role or relevance of individual intentions. This is the
case for instance in ritualized speech, including divination (Du Bois 1987) and
political oratory (Du Bois 1986), where the frequent use of stock phrases,
common metaphors, and proverbs places the emphasis on a de-personalized
stance of the speaker, who, as it were, renounces his own will and hides his
intentionality by using the words of the ancestors or those of “tradition” (Bloch
1975; Duranti 1992a). Even in those cases in which the speaker seems to be
speaking in the first person, there is a systematic ambiguity as to the social
identity of the speaker – it is at this point that distinctions must be made
between speaker, sender, addressor, and even “voice” (Bakhtin 1981; Duranti
1986; Firth 1975; Haviland 1991; Hymes 1972a; Vološinov 1973).
In guiding speakers to assume particular stances and avoid expressing
potentially dangerous opinions or emotions, certain speech genres may force
people to say what they do not mean or, in some cases, liberate them from the
responsibility of certain positions or deliberations. Whether or not a speaker
has certain intentions (e.g., to do something in the future, to speak the truth, to
reveal certain feelings) may be made irrelevant by the genre that is spoken and
the role that such genre entails.
114 Intentionality and truth, revisited

5.5.1 An example of how a genre constrains what a speaker can say


For example, in the first speech in a fono (see Chapter 3), which is modeled on
the ceremonial genre called lāuga (Duranti 1981a, 1981b, 1994), there is a
section dedicated to the “topics of the fono.” As shown in (1), the announcement
of the agenda is the sixth part of the speech and thus it happens after a consider-
able amount has been said about entities and events that are not up for discus-
sion: the power of the Christian God and the need for His guidance and help
(part 2), the important events in the history of the country (part 3), the authority
of the chiefly titles (part 4) and the acknowledgment of all the major titles and
descent groups in the village (part 5). Part 6, the agenda of the meeting, is
different because it deals with specific recent events, that is, contingencies.
(1) Sequential organization of the parts (vāega) of a fono speech
1. `ava (‘kava,’ mention of the previous offer or drinking of kava)
2. fa`afetai i le Atua (‘thanksgiving to the Lord’)
3. taeao (‘mornings,’ a metaphor for important events in the history of the
country)
4. pa`ia (‘sacred, holy,’ a recognition of the chiefly titles in pre-
Christian terms)
5. fa`atulouga o le fono (‘acknowledgment of the council’ or recognition of
the major families and titles represented in the fono or even beyond)
6. matā`upu o le fono (‘topics of the fono’ or ‘agenda’)
7. fa`amatafi lagi (‘clearing of the sky,’ or good wishes to participants)

There is then a sharp contrast between the contingent quality of specific


“topics” and “events” to be discussed next and the non-contingent quality of
everything else that is being talked about in the same speech (see Duranti
1994). As shown in excerpt (2), such a contrast is minimized by providing only
a very brief and vague account of the agenda. The first speaker of the day, Loa
(the same person who was a major protagonist in the events discussed in
Chapter 3), only says that the agenda of the meeting is about two people
because of something that happened on the day of the national election.
(2) (March 17, 1979 meeting of the village council or fono; passage extracted
from the first speech of the day, performed in the ceremonial style called
lāuga)

1 Loa; ia `o:- . . . pei `o makā`upu o le aofia ma le fogo . . .


‘so . . . as for the topics of the assembly and the fono . . .’
2 `o lo`o iai, . . . ka`akia i- `i pā`aga,
‘(that) there are . . . (they) are about the partners,’
3 ia` `o:- . . . `o Vaikoko lava ma: le afioga iā `Alai`a-sā, . . .
‘so, it’s . . . Vaitoto himself and the honorable Alai`a-Sā, . . .’
4 oga `o le aso (o) le kākou pāloka, . . .
‘because of our election, . . .’
Code-related limits on speaker’s intentions 115

5 ia` . . . ka`ilo gā po `o:- iai ma gisi makā`upu, . . .


‘so, . . . I don’t know if- there are other topics, . . .’
6 ia` `a `o le ā- . . . lumāmea lava i le aofia ma le fogo,
‘so, but they will be brought in front of the assembly and the
fono’
7 gisi fo`i makā`upu `o lo`o:- . . . iai kokogu o le kākou aofia, . . .
‘any other topics (that) exist inside of our assembly, . . .’

In a highly stratified society like the Samoa of the late 1970s when I recorded
the above speech, it would have been risky for an orator to even suggest that
someone of higher rank might have misbehaved and deserve punishment. In
this case, Loa, by holding a matai title that belongs to a family from the
subvillage of Sanonu, is expected to deliver the first speech,2 which includes
the announcement of the agenda of the fono. However, in the passage above
one of the two people he mentions is someone who holds the title of Alai`a-Sā,
one of the four highest chiefs (ali`i) in the village. It could be argued, as I have
in the past (Duranti 1994), that narrative minimization in this case protects Loa
from future retaliation for saying the wrong thing or being disrespectful. As we
saw in Chapter 3, such retaliation was not out of the range of possibilities for
Loa himself.
If we accept the premise that Loa had reasons to be careful in what he said
and in how he would say it, we can easily interpret what he did in (2) as the
product of his own intentions, which became realized through the use of a
particular set of linguistic resources he had access to by being a Samoan
speaker and a skillful speechmaker. One of these resources is the utterance
/oga `o le aso (o) le kākou pāloka/, ‘because’ (or ‘on account of’) ‘our
(inclusive) election’ in line 4. This is, syntactically speaking, an elliptical
sentence. It lacks a predicate that would further specify what happened on
election day, or that something did happen then. It follows a syntactic pattern
found elsewhere in the fono speeches. Pragmatically speaking, one could say
that Loa can get away with the vagueness of this verbless utterance simply
because everyone present at the meeting that day probably knows already what
the two “partners” are being accused of – in addition to the likely circulation of
information through informal channels, there had been a fono a week earlier
when their behavior had been discussed but without the two defendants being
present. In this account, Loa is a speaker who can choose to say what he knows
to be appropriate to the context and not say what he knows to be inappropriate
and dangerous. Loa is the agent, and the language he chooses is the medium,
that is, the instrument for realizing the goal of mentioning the agenda in the
culturally proper way and the safest possible way he can think of.
But there is also a way of interpreting what is reproduced in (2) as an
attempt to redistribute the agency behind the utterances and their overall
illocutionary force (see Besnier 1989 for a similar claim within the context
116 Intentionality and truth, revisited

of conversational interaction in Tuvalu). Loa is here the carrier of a tradition he


did not invent and he does not individually control. It could thus be argued that
in performing his speech in the formal genre appropriate for starting a fono,
Loa lets himself be guided by such a formal genre. This hypothesis is consist-
ent with the study of genres as “orienting frameworks, interpreting procedures,
and sets of expectations” (Hanks 1987: 670) or as “orderly and ordering
principle[s] in the organization of language, society, and culture” (Briggs
and Bauman 1992: 144–145). Along similar lines, I see a genre like the lāuga
as verbal and embodied practice of sense-making, which does have an agency
of its own. This agency of course should not be understood as detached from
speakers but as emerging from their actions. The verbal performance of a
genre, with all its intra- and inter-textual relations, activates its force. In
speaking in the style appropriate for the expression of certain emotions, we
evoke and embody those emotions, letting them come into the context of
interaction among the participants of the speech event.
As a sequentially structured series of different “parts” (vāega), the lāuga
sets up the temporal boundaries of the mentioning of the agenda just as the
score of a popular song predetermines when certain chords should be played
even though a pianist might be free to play the chord on different parts of the
keyboard and a guitarist may choose among different positions along the guitar
neck and among the possible arpeggios or strummings. As a repertoire of
lexical choices, tropes, and style of expression, the lāuga is a blueprint for
performance, which sets up the tone of what will be said and predisposes the
speaker to enter a particular intentional mode where certain speech acts and
speech choices are preferred (and easier to execute) and other ones highly
dispreferred and in some ways hard to come up with. By the time Loa arrives at
the sixth part of his speech – see (1) above – he has been channeling the
authority of traditional Samoan hierarchy and values for several minutes. The
slow speed and subdued tone of his speech have helped establish a connection
between an unquestionable historical-mythical past and an uncertain and
unknown present that is potentially threatening to the ideal harmony or ‘mutual
affection’ (fealofani) among the family heads attending the meeting. The
aesthetic attributes of the lāuga genre with its metaphors, proverbs, special
vocabulary, and frequent parallelisms and repetitions are thus more than a
toolkit for sounding rhetorically skilled and knowledgeable about tradition.
They have a role in organizing one’s thoughts and one’s utterances. As typical
of ritual speech (Du Bois 1986), the lāuga leaves no room for doubting the
truth of what someone has said.
To get to accounts of (recent) events whose truth can be debated, the lāuga
genre must give way to a different, less predictable, and more flexible register,
a hybrid way of speaking sometimes referred to as talanoa (plural talatalanoa)
‘to chat, converse, talk’ (Duranti 1981a, 1984). The fono organization allows
Code-related limits on speaker’s intentions 117

for this shift because after the opening lāuga, which can be one or as many as
the subvillages represented in the meeting, the conventions of the lāuga are
partly abandoned and instead of mentioning lists of matai titles or unquestion-
able truth, details of actual events might be mentioned, high-ranking individ-
uals might be criticized or scolded, and speakers’ statements might be directly
questioned. Even during this part of the fono, however, the convention of long
turns at talk combined with the authority of the individual speakers who
represent extended families and sometimes important lineages makes it diffi-
cult to confront someone regarding a particular claim or to bring out possible
contradictions. In order for this to happen, the oratorical genre needs to be
broken. This is what happens in the following excerpt from a dramatic moment
in a fono where a young orator (holding the matai title “Vaitoto”) and a high
chief (holding the title “Alai`a-Sā”)3 have been accused of publicly insulting
the other members of the council on the day of the national election. In (3)
below we see the powerful senior orator (matua) Moe`ono Kolio interrupt the
younger and lower-ranking defendant Vaitoto in an attempt to induce the latter
to either admit responsibility or accuse someone else. The speechmaking
format is broken and the exchange becomes confrontational. Eventually, the
orator Tapusoa will ease the tension by returning to a speech style that is closer
to the traditional lāuga.
(3) (March 17, 1979: the junior orator Vaitoto has denied any wrongdoing while
saying that he will accept the decision of the council to fine him; he is almost
done with his speech, when Moe`ono interrupts him)

Vaitoto; `a `o gā `ua `ou kala mamā aku


‘but there I have spoken purely to you’
Iuli ma Moe`ogo ma le mamalu . . .
‘Iuli and Moe`ono and the honorable’
`Āiga ma aloali`i . . . ma lo kākou gu`u // ia` (?)
‘chiefs and son-chiefs and our village . . .’
Moe`ono; ((interrupting)) `o loga uiga lā o lau fekalaiga Vaikoko,
‘the meaning of your honorable speech Vaitoto,’
e lē `o lau kōfā `a `o Alai`a lā e oga `upu?
‘it’s not Your Honor’s words but Alai`a’s (that were said)?’
Vaitoto; `ou ke le iloa se // (`upu)
‘I don’t know any (words)’
Moe`ono; `o `oulua sā i kokogu o le ka`avale
‘you two were inside of the car’
uā `o `upu e lafo aku i kokogu o le ka`avale
‘because the words came out from the inside of the car’
oga `ua lua- lua kokogu o le ka`avale.
‘because you two were inside of the car.’
Vaitoto; e leai se `upu a // Alai`a (se [unclear])
‘there are no words that Alai`a said (?)’
118 Intentionality and truth, revisited

fai mai fo`i le afioga iā Alai`a


‘His Highness Alai`a also told me’
e leai se `upu a Alai`a ga faia. . . .
‘not a word was said by Alai`a.’
ke le iloa gi- gi mea e mafai iai (lā ia/Alai`a) . . .
‘(I) don’t know of any thing that (Alai`a) could do.’
`ou ke mālamalama lelei iā ke a`u.
‘I understand myself very well.’
Moe`ono; ga lua feosofi loa i kokogu o le ka`avale lafo aku ma `upu!
‘you two jumped inside of the car then cast some words!’
Tapusoa; fa`afekai aku lava i le pa`ia ma le mamalu o l-
‘thank you very much honorable and dignified (members) of th-’
`ua aofia fo`i- . . . pa`ia i `Āiga ma aloali`i
‘the honorable chiefs and son-chiefs have gathered.’
[. . .]

Orator Tapusoa’s intervention succeeds at ending the confrontational talk


between the senior orator Moe`ono and the young orator Vaitoto by starting
his intervention with two phrases that acknowledge the sacredness (pa`ia) or
dignity (mamalu) of the participants. In so doing, Tapusoa changes the
interpretive frame of the activity. His use of a more traditional oratorical
genre reactivates a turn-taking system that grants a speaker the floor for an
extended period and thus ends the quick back-and-forth that the senior orator
Moe`ono had started with his interruption in the middle of Vaitoto’s self-
defense.
This brief excerpt from a fono I recorded in 1979 reminds us of the close
relation between content and form of verbal communication. But precisely
because there is no independent propositional content that can be conveyed
regardless of the style of speaking, the struggle about truth is also a struggle of
style. The fact that Tapusoa can succeed at stopping the quasi-conversational
exchange introduced by the powerful Moe`ono by reintroducing the longer
turns underscores the force of tradition in the social arena of the fono. It also
shows that the discourse produced in such a context is not under anyone’s full
control no matter how powerful that person might be. Once Tapusoa has
started, Moe`ono does not challenge him. To do so would be an extreme
violation of the fono etiquette.

5.5.2 Audience co-participation in shaping intentions


A number of theoretical and empirically based contributions to the hermeneut-
ics of texts have reiterated the basic point that any act of speaking and its
interpretation are partly constructed by the audience’s response (Duranti and
Brenneis 1986; Gadamer 1976; Goodwin 1981; see also Chapter 10). This is
Code-related limits on speaker’s intentions 119

related to the retrospective, typically post hoc nature of any act of interpret-
ation (see Wittgenstein 1958). Without others to carry on our messages, to
complete them, expand them, revise them, etc., we would not be able to
communicate; ultimately, we would not be able to know what we are talking
about. Current and past debates within the humanities and social sciences on
the need to have usually repressed voices heard falls within this perspective.
The related proliferation of fields like feminist anthropology, cultural studies,
ethnic studies, and gender and sexuality studies are also attempts at getting the
path clear for certain groups to be able to “hear themselves” through their own
ears instead of having to be translated by a mainstream, dominating audience,
however defined, that imposed the hegemonic discourse of the past.
That the audience counts is of course a commonplace notion to those who
have been studying the details of everyday encounters. As demonstrated by
conversation analysts, even the simplest routine, e.g. starting a telephone
conversation, is based on a rapidly unfolding joint activity that has normative
expectations, preferences, and structural constraints but cannot be completely
predicted and must thus rely on both parties to cooperate toward a formulation
that “works” (Schegloff 1986). Any of our words, phrases, and utterances, not
to speak of larger units, contains several possible “next moves.” Addressees
are co-authors not only when they join in and help us figure out what we
“really” mean (e.g., when the addressee is a teacher, a therapist, or any kind of
socially recognized “expert”), but also when they are not present. This is the
“Internalized Other” of G. H. Mead’s constructivist psychology (cf. Mead
1934; Mehan 1981), which can, in turn, surface as the “voices” expressed
through direct, indirect, or quasi-direct speech (Vološinov 1973; Bakhtin
1981). The work of the linguist and discourse analyst in these cases is to look
at the techniques used by speakers to index multiple, alternative audiences.
Such devices as code-switching as well as prosodic features of utterances
(Blom and Gumperz 1972; Gumperz 1982, 1992; Woolard 2004) can convey
important information as to who is the primary audience of a given message
(Brenneis 1978). But what these different devices do should not be explained
simply in terms of intended meaning. Some of them are rather like nets which
are thrown out in the interactional sea to catch the fish that happens to be slow
or unfortunate. What has been described as typical of African American verbal
dueling, where an utterance is not an insult until the recipient interprets it as
such (Kochman 1983; Labov 1972; Mitchell-Kernan 1972), may very well be
the ideological recognition, through the institution of a particular verbal genre,
“signifying” – or what used to be called “sounding,” or “playing the dozens” –
of a practice that may explain a much wider range of interactions than we
might be willing to accept (Morgan 1989, 1996). What we need is a model of
communication that takes the emergent, interactive nature of the process into
consideration while at the same time capturing the local, culture- or even
120 Intentionality and truth, revisited

context-specific techniques used to diffuse possible motives or open them to


negotiation. In its negative definition, verbal communication could be seen as a
series of constraints on the uncertainty of possible interpretations of the
Subject’s intentions. When the audience is cooperative, we do not realize the
work that is being done to maintain the direction of talk within predictable
boundaries. When the audience is less cooperative (e.g. legal and political
arenas), uncertainty and ambiguity might have to be foregrounded, as is the
case with verbal indirection (Brenneis 1986).

5.5.3 Culture-specific constraints on introspection and responsibility


The view of interpretation as a guessing game whereby one tries to find out
what the speaker really thinks or wants is strongly grounded in our western
notion of person and the economic metaphors that govern our theories of
communication (cf. Geertz 1983; Shore 1982; Holquist 1983; Rosaldo 1982;
Rossi-Landi 1983). The belief that one can know what someone else is
thinking or have access to his or her mental processes is not shared across
cultures (cf. Duranti 1988; Rosaldo 1982; Rosen 1985; Schieffelin 1986). As
mentioned earlier, the ability to read other people’s minds as the generalized
method for correct interpretation is connected to the western assumption of a
universal ethics. Empirically speaking, interpretive strategies differ cross-
culturally (or cross-contextually) and are closely tied to institutions. The
meanings embedded in a family conversation need not be considered in a
business meeting among colleagues. Cross-culturally, there seems to be a
difference between cultures that tend to focus on the subjective state of mind
of speakers (viz. the illocutionary force of their utterances) and those that are
rather more concerned with the consequences of talk (perlocutionary force) (cf.
Duranti 1988). Originally Austin thought of the difference between the two
types of acts as a question of conventionality, with the former but not the latter
meeting the conventionality criterion. Thus, if I say that I cannot take you to
the movies and you decide that I don’t like you, one should not speak of the
“don’t like you” message as part of what I “meant,” i.e. it is not part of the
illocutionary force of my words. One of the problems with this view is that it
does not take into consideration those cases in which certain consequences, not
directly encoded in the message, are equally recognized by a culture – here
understood as a network of interacting subjects – as potentially predictable and
hence, in a sense, conventional. It also assumes that the relevant conventions in
any given interaction are an a priori specifiable set. But this is not the case,
given (a) the distributed nature of knowledge, whether of ritual or mundane
behavior, and (b) the negotiated nature of human interaction, whereby rules
and conventions are continuously reassessed and redrafted. Finally, it does not
offer any solution for what Bogen (1987) has called the “problem of
Code-related limits on speaker’s intentions 121

heterogeneity,” that is, the problem of utterances that have different but equally
acceptable interpretations according to the audience and conventions we take
into consideration.
These issues also tie intentionality to responsibility. Much of legal and
political debates center in fact around issues of responsibility and as such
constitute a group’s attempt at defining the factors relevant to making a given
interpretation legitimate (e.g., Gluckman 1972). This is a typical arena for
power struggles and for exerting social control. Much of the work that is being
done by those in power to control information and outcome is focused on the
criteria for bringing out evidence rather than on the “facts.” If a group succeeds
at making the criteria acceptable, the details of a crime or even the identity of
the offender may be seen as a relatively unimportant matter. This possibility is
both recognized and actualized in Samoan political arenas, where, as discussed
in Chapter 3, an orator may get into trouble for the actions of a chief (or a high-
ranking government official), if the orator did not act to prevent him from the
wrongdoing or did not try to remedy the chief’s misconduct or failure to fulfill
expectations. It might not be accidental, then, that the Samoan word for
meaning (uiga) is the same as that for deed, action. The local practice of
interpretation is one in which the interpreters look at the actions and ignore the
actor’s intentions (Shore 1982). In Samoan, one cannot say “I didn’t mean it”
(Duranti 1988: 17) and during our fieldwork in the late 1970s some of the
younger transcribers used to mock the use of “sorry” in English as an excuse
for a mistake. Eve Danziger (2006) also reported that the Mopan Maya
consider “I didn’t mean it” inappropriate as a defense of wrongdoing.

5.5.4 Multiplicity of goals


Any utterance has potential or actual multiple goals. This multiplicity corres-
ponds to a general underspecification of number and types of speech acts for
any given utterance. This is particularly the case when we take time into
consideration. In Gibson’s terms, language affords reinterpretation, recontex-
tualization. Utterances are not linked once and for all to one specific situation.
The matching of certain locutions with their meanings is but one contextually
defined choice or selection among a range of possibilities that cannot be
completely specified in advance. This process of selection is itself embedded
in systems of social relations and obligations. To take something as a compli-
ment has consequences – and different ones for different people in different
contexts. To engage in assigning intentionality means to engage in the econ-
omy of power relations in a given community. The fact that the analysts seem
immune to the effects or consequences of their interpretations has to do with
the common physical, geographical, and institutional distance between
researchers and their subjects. When the subjects are able to read and react
122 Intentionality and truth, revisited

to the anthropologists’ descriptions, the metaphor of “neutral” or “objective”


accounts becomes harder to maintain. In particular, it is difficult for anthro-
pologists to predict certain uses or abuses of their work. In some cases, the
sharing of information among well-meaning colleagues or between researchers
and the funding institutions may prove damaging to the “subjects.” These
ethical and political issues of fieldwork are part and parcel of the general
property of multiplicity of goals and interpretations inherent in any semiotic
activity.
The distinction drawn within speech act theory between illocutionary force
(as produced by the intentions of the speaker) and perlocutionary force (as
made evident by the consequences of the act, e.g., its impact on the addressee)
is only a first step in trying to deal with some these phenomena. Once we
introduce time and context in the analysis, we realize that what speakers might
consider as an unintended consequence and, thus, a “perlocutionary effect” in
Austin’s terminology, might receive interpretations that either challenge or
ignore speakers’ intentions. For example, on September 12, 2006, Pope Bene-
dict XVI gave a speech in Germany in which he quoted from a 1391 dialogue
written by the Byzantine emperor Manuel II Paleologus where the teachings of
the Prophet Mohammad were described as “evil and inhuman.” Muslim
leaders, and some historians of Islam immediately reacted with outrage and
asked for an apology. In the days that followed the Pope’s statement, acts of
violence such as the burning of some churches in the Palestinian territories
were also reported and interpreted as retaliations for the Pope’s statement. Five
days later, on September 17, 2006, the Pope issued a statement from his
residence in Castel Gandolfo, Italy, in which he regretted the reactions to his
remarks about Islam, which, he said, were a quotation from a medieval text
that did not reflect his own personal thought. Even though a number of
authorities in the Muslim world downplayed the significance of the Pope’s
words or attributed them to a misunderstanding of the Qur’an, others pointed
out that the Pope’s follow-up statement, which was quoted in the English press
as “I am sorry,” was not a true “apology” or was not adequate (The New York
Times, September 18, 2006).4
What this example shows is that it is difficult for a public figure like the
Pope or a political leader to control the meaning of his or her words and an
apology will be required regardless of a speaker’s intentions, even when those
intentions could be textually defended (e.g., by claiming that one was quoting
someone else’s opinion). Similarly, even a statement like the one that the Pope
delivered five days later is open to a number of interpretations, with journalists,
politicians, and religious leaders showing very different views as to whether it
was or it was not a true (or sincere) apology.
This example also reminds us that not only is it difficult in many situations
to say what a particular act is (e.g., offense, provocation, condemnation,
Problems with the truth-functional model 123

apology, declaration about a personal feeling), but it is also possible that the
ambiguity or vagueness of an act’s pragmatic force be routinely exploited by
speakers and their audience, both consciously and unconsciously. The test of
explicit performative verbs of the type I promise you . . . or I blame you . . . or
I accuse you . . . , I apologize for saying . . . is usually not available in
spontaneous interaction, and the conditions for deciding once and for all the
illocutionary force of a given speech act are not a discrete number or independ-
ent of the parties’ interests.
The addressee’s knowledge and expectations as well as the nature of the
encounter are crucial features in determining the kind of act that is being
performed. Take for instance the difference between an accusation and a
complaint. At first, the former seems to imply not only a violation of a social
code, but also the expectation of some legally or otherwise induced conse-
quences (for instance, a payment), whereas the latter does not seem to
have the same implications. Concentrating on the speaker’s attitude or the
responsibilities of the party about whom the accusation or complaint is being
made provides, however, a very limited view of the factors that play a part in
the distinction between accusation and complaint. There are definitely many
other elements potentially at work. The social identity of the speaker and his
or her rights are also very important. The question as to whether a given
utterance is a complaint or an accusation cannot be resolved once and for all
at the time of speaking, given that part of the reasons for such an utterance
might precisely be the testing of the very conditions that would make it into
an accusation (as opposed to a complaint). This is what routinely happens in
political contexts. It is thus what with Giddens (1979) we can call the
“duality” of social structure in general and of speech in particular that
undermines philosophical speculations about out-of-context speech act
types. There are “rules” (e.g., of syntax, semantics, pragmatics) for the
interpretation of speech acts, which make communication and social action
possible and unavoidable (e.g., one feels the social pressure to answer a
question or respond to an invitation), but the use of speech also makes
possible the reproduction of the system as well as the constitution of new
contexts or interpretations.

5.6 Problems with the truth-functional model


The very foundations of logical semantics in its modern sense are built on the
distinction between the apodictic properties of relations between expressions,
given some hypothetical truth-values, and the actual states-of-affairs repre-
sented by those expressions. Thus, it was very important for the early Husserl
to define “pure logic” as the “Ur”-Science that does not have to deal with
the “experiential content” of the life-world – in this, Husserl follows Kant. The
124 Intentionality and truth, revisited

notion of meaning as the conditions under which a given expression can be


said to be true does not change the original emphasis of modern logic on
abstract, universal properties of logical relations (e.g., negation, implication)
and the fundamental lack of interest in the actual states of affairs represented
by the propositions whose truth-values and logical relations are being evalu-
ated. This distinction is represented in Carnap’s (1947) use of the terms
“logically true/false” vs. “factually true/false.” This tradition is inherently
universalistic in its search for context-independent properties of propositions
(and relations among them).
This exclusive interest in the “formal” properties of meaningful propositions
can be traced back to Kant, who, in the Critique of Pure Reason, wrote:
Now a general criterion of truth must be such as would be valid in each and every
instance of knowledge, however their objects may vary. It is obvious however that such
a criterion [being general] cannot take account of the [varying] content of knowledge
(relation to its [specific] object). But since truth concerns just this very content, it is
quite impossible, and indeed absurd, to ask for a general test of the truth of such
content. A sufficient and at the same time general criterion of truth cannot possibly be
given . . . But, on the other hand, as regards knowledge in respect to its mere form
(leaving aside all content), it is evident that logic, in so far as it expounds the universal
and necessary rules of the understanding, must in these rules furnish criteria of truth.
Whatever contradicts these rules is false. (Kant [1787] 1965: 97–98)
It is this almost exclusive concern with the formal properties of the relation
between language (whether formal or natural) and the world (or “objects”) that
has been criticized as insufficient for representing the complexity of meanings
which characterizes daily uses of natural languages and symbolic phenomena
in social life. Certainly, logical theorems and proofs give us a “technology of
knowledge” (in Foucault’s sense) through which truth and hence forms of
interpretation become possible, but we know now that actual interactions will
reveal more subtle domains of signification, for which we need access to more
dynamic notions and models.5
There are devices in languages that allow speakers and hearers to tem-
porarily suspend the ordinary criteria for truth verification. The most obvi-
ous examples are conventional metacommunicative statements that alert
hearers that what is coming next is part of a virtual and alternative reality
with a distinct practical logic. Such framing devices are common in
storytelling.
As soon as we hear a generic framing device, such as “once upon a time,” we unleash a
set of expectations regarding narrative form and content. Animals may talk and people
may possess supernatural powers, and we anticipate the unfolding of a plot structure
[with] an interdiction, a violation, a departure, the completion of tasks, failure followed
by success, and the like. (Briggs and Bauman 1992: 147)
Problems with the truth-functional model 125

But narrative fiction is not the only context in which we suspend our ordinary
standards for truth finding. There are situations where the interaction seems to
be structured in such a way as to make it difficult for participants to fully
challenge the veracity of what is being said or “get to the bottom” of the claims
and counterclaims that have been made. One such context is political oratory.
One common way of characterizing the problems with truth assessment in
political debates is to simply say that “politicians lie” or cannot “be trusted.”
But such an analysis turns out to obscure the very conditions that make it
difficult to get a satisfactory answer regarding the truth of a candidate’s claim.
In the next section I will suggest that one such condition is the exchange
system that is conventionally adopted for debates, including the long turns
during which time a candidate can have the floor without being immediately
challenged for the veracity of his or her claims.

5.6.1 Truth and turn-taking


Between November 1995 and November 1996, I documented the political
campaign of Walter Capps, then a Professor of Religious Studies at the
University of California at Santa Barbara, for a seat in the US House of
Representatives. By following him around with my video camera I had the
opportunity to record a number of interactions Capps had with members of his
campaign team, family, supporters, and opponents. These interactions
included political debates in front of a live audience. Such debates followed
a format that is fairly familiar to candidates and audiences in the US, even
though there is always room for variation across situations, and moderators
may impose or try to impose particular restrictions regarding length of
speeches by candidates or the type of response the audience is allowed to
engage in while the candidates are speaking. One particular feature that is
common across all kinds of debates, including the televised presidential ones,
is the long turn format (or “macro-turn”), which happens to be not that
different from the one I described for the Samoan fono in Duranti (1981a)
and elsewhere (see §3.2). Candidates are given a certain amount of time
during which they can present their views, woo their audience, and criticize
their opponents. The familiar political debate format alters the ordinary
conversational turn-taking in a number of ways. It pre-allocates the floor
establishing who will speak next – something that in conversation is instead
negotiable and emergent from the features of the context of interaction. It also
allows speakers to hold the floor for an extended period during which time
other participants who might have something relevant to say are not expected
to respond to criticism, comment, or ask questions, even when they hear
something that they might consider to be an error or a false and damaging
126 Intentionality and truth, revisited

statement. This has consequences for the establishment of truth in public


debates and may put candidates in a double bind because they have to decide
on the spot whether they can afford to wait for their turn to correct what has
been said or speak right away, violating the agreed-upon rules of turn-taking
and thus risking being reprimanded by the moderator in front of the audience.
Here is an example from my 1995–1996 project on a political campaign for a
seat in the US House of Representatives (see Chapter 6 for more information
on this). At one point toward the end of a public debate in San Luis Obispo,
on August 15, 1996, Democratic candidate Walter Capps starts to emphasize
the fact that there has been a general tendency to reduce the role of the federal
government and give more authority to the states, but that such a position is a
mistake “because the states will not pick up those responsibilities and obliga-
tions in a manner that is collectively moral.” The responses by the other
candidates show that indeed they all see a more limited role of the federal
government. Incumbent Seastrand, for example, says that “there is a real
movement, pushed by many of my freshmen colleagues including myself to
see to it that many of the things that have been usurped by the federal
government uh would come back to the states.”
When Capps has a chance to speak again, he returns to his earlier point, this
time casting himself as the only candidate who is “not a libertarian” and the
only one “who wants to say something good on behalf of federal government.”
This is strange, he argues, given that he has “never been to Washington . . .
never served there.”
(4) (Debate in San Luis Obispo, August 15, 1996)
Capps; [. . .] I think I’m the only candidate up here that’s not a libertarian.
that is I (may be) the only one here who. wants to say something
good on behalf of federal government. now that’s a strange role for
me to be in because. I’ve never run for public office before. I’ve
never been to Washington– never served there. easy to blame
bureaucrats in Washington for all our national problems. I think
that’s way too easy. (while) certainly I want strong, state
government. certainly I want strong local government. If I had my
druthers I would take local government over state government.
because that’s where I think the most important decisions are made.
but we do need to have. strong effective. Efficient. minimal.
national federal government. [. . .]
The video recording I made that day shows that right after Capps said “I’ve
never been to Washington,” incumbent Seastrand makes an overt expression of
surprise and writes a note on a pad in front of her. A few minutes later, when
her turn comes, Seastrand returns to Capps’ statement to undermine its ver-
acity. She says that Capps misrepresented himself as someone who has “never
been to Washington” because, in fact, he has been to Washington.
Problems with the truth-functional model 127

(5) (Debate in San Luis Obispo, August 15, 1996)

Seastrand; and, as far as my friend (uh- p-) Professor Walter Capps –hh
uh– I’m amazed that you’ve stated on several occasions in this
meeting that you’ve never-been-to-Washington. I’m gonna
have to go into my files and look at The Santa Barbara
News-Press because I think that they reported that you even
went into the Oval Office. of the President himself. Your– uh
one of your family members has or works there with (such
Ste–) George Stepanopolous’ office. and so uh I was given the
impression on reading that article that when you go to
Washington, you meet the President. So anyway it’s
interesting but it’s an election time. [. . .]

Seastrand is here pointing out that Capps cannot claim that he has not been to
Washington because, according to the newspaper The Santa Barbara News-
Press, he visited the President in the White House. A careful reading of Capps’
statements reveals that for him “Washington” stands for “government” and he
is not literally claiming that he has never traveled to Washington but that he
has never been an elected official. This is made clear by the discourse context
and by the self-correction “never served there.” (“I’ve never been to Washing-
ton– never served there.”)
I would like to suggest that in a conversational interaction it would not be
difficult for Capps to try to correct Seastrand’s interpretation and for hearers or
viewers to get a chance to make up their mind about whether what is being said is
true or false, but, as shown in excerpt (6) below, the turn-taking rules defined by
the moderator are not designed to allow for interruptions or immediate rebuttal.
(6) (Debate in San Luis Obispo, August 15, 1996)

Moderator; each candidate is going to have three minutes . . . to introduce


themselves . . . and then we’ll start the question. I’ll ask a
question and then they’ll all answer two minutes . . . on the
same question. and then after everyone has answered on the
same question for two minutes . . . then we’ll have one minute
a rebuttal type thing. and if you applaud too long, . . . that’s
gonna cut down on a lot of people’s time. . . . So maybe if
we just uh- . . . applaud now for all of the candidates even
though Andrea [Seastrand, the incumbent] isn’t here yet. let’s
applaud and get it out of our system. Okay? ((he starts
clapping))
Audience; ((claps))
Moderator; thank y’all. uh- I would also like to say if any of you have any
catcalls . . . or any snide remarks . . . to make to any of these
candidates please take them outside now. . . . these people are
guests of our AARP here and we really don’t want uh-
128 Intentionality and truth, revisited

somethin’ like that. . . . and I’d also like to ask the candidates to
kind of hold down on their . . . criticism of each other . . .
cause we really just want the facts here. as they used to say “we
want the facts ma’am and that’s all.” So- . . . uh- . . . we’ll pick
out one of the questions here. Let’s see. [. . .]

These rules force candidates to wait for their turn to rebut what someone else
said. This means that, when they do get another chance to talk, candidates have
to decide whether they are going to give up some of the time they have to
correct a point that listeners might have already forgotten, to examine a point
or detail of what was said that might make them look fussy or even defensive.
In some cases, such a chance to make corrections or provide a counterexample
might never come. This is the case in the debate I introduced above. After
incumbent Seastrand is done with the long turn from which I took excerpt (5),
the moderator takes the microphone and closes the debate.

5.7 Truth and affect


The role of truth-values in matching mind and world is particularly problem-
atic when emotions are at play. How can we possibly know under which
conditions a person is really feeling sad or happy or angry? It is in these
contexts that cross-cultural and cross-contextual differences can be expected.
One dimension that is often invoked in the anthropological literature is the way
in which the expression of emotions indexes not so much or not simply “states
of mind,” but social relationships and the cultural values attached to them
(Lutz and White 1986). Truth is in these cases either irrelevant or just one of
the elements against which an appropriate interpretation must be posited. It is
in these contexts that speakers are likely to feel either constrained or relieved
by the limitations imposed by the code they are using. Again, cross-cultural
and cross-contextual differences abound. The extent to which people are
expected to be “original” in their expression of sorrow, for instance, is highly
variable. The conventionality of giving others the job of conveying our
emotional responses to a given event (e.g. birth, death, marriage, etc.), whether
through ritual wailing (Briggs 1992; Urban 1988), traditional oratory, or the
exchange of material goods, forces us to rethink the very object of the
intentional act. If the goal (or one of the main goals) of the performed acts is
to reaffirm social alliances, for instance, rather than mourn the dead or be
joyful about the newborn or the newlyweds, the sincerity of the performer is
irrelevant, especially when the performer is just the messenger.
These cases can be dismissed by assuming a standard, unmarked personal
commitment to truth to be characteristic of everyday interaction and conversa-
tion. Such a standard would be momentarily lifted on special occasions,
Truth and affect 129

including public speechmaking. This is in fact Bloch’s (1975) position. On the


other hand, in both public and private arenas, emotional outbursts and confron-
tations are often about questions of truth. In political debates as well as in
arguments within more private contexts (e.g., within family), there is a consid-
erable amount of discourse that is dedicated to finding or providing evidence
for a given event or position. It is in these contexts that the distinction between
truth and authority or truth and responsibility becomes more and more difficult
to draw. As pointed out by John Haviland (1987, 1989), it is not by accident
then that the same linguistic devices used for the expression of evidence vis-à-
vis a given assertion are also used to provide moral assessment.
Evidentials encode not only what a speaker knows or how he knows it; but also what an
addressee can be taken to know, or should know, or apparently (perhaps culpably) fails
to know. (Haviland 1987: 343)

This point ties the discussion of truth with a point I made earlier about
intentions: in both cases, we are forced to expand the communication model
assumed in the classic treatment of intentionality and truth. We cannot simply
talk about Speaker, Message, and Referent. The potential or actual role played
by the audience in constructing a sequence of interaction (Goodwin and
Goodwin 1987, 1992) is essential for explaining the multifunctionality of such
linguistic devices as evidentials, dubitative forms, hedges, affect particles, etc.
What speakers seem to be doing in these cases is to exploit truth rather than use
it as the ultimate criterion for interpretation. When we look at verbal inter-
actions that deal with controversial issues (whether in informal gossip or in
formal court interaction), we realize that affect is built into the message
precisely by the manipulation of the relative epistemic status of what is being
said (Besnier 1990, 2009; Haviland 1977; Ochs and Schieffelin 1989).
Speakers may hint at possible (embarrassing, threatening, unexpected) truths.
They may describe a world or a situation that is only half true or characters that
are only half real. The audience is often expected to do the rest of the job by
either accepting or denouncing the potential descriptions of the world more or
less vaguely suggested – but not necessarily asserted – by the speaker. The
goal in everyday interaction is not simply to move the audience, as suggested
by traditional rhetoric, but to move with the audience, that is, to make the
audience at least partly responsible for what is being asserted or implied.
Sometimes the speaker’s political or moral success is judged not so much on
his or her ability to make the audience see the truth, but rather on his or her
skills in making truth irrelevant.
Another dimension closely related to affect is the relative position or attitude
displayed or hinted at by a social actor in performing a given act (whether
verbal or otherwise) – what some authors call “stance” (Biber 1988; Biber and
Finegan 1989; Du Bois 2007; Berger 2009; Jaffe 2009). Reported speech – in
130 Intentionality and truth, revisited

combination with other grammatical devices such as tense/aspect, impersonal


forms, passive voice, etc. – is a commonly used mechanism for embedding
someone’s voice (even one’s own) within the talk of another (cf. Besnier 1993;
Hill 1985; Lucy 1993; Vološinov 1973). The way in which a given speech act
is attributed to someone else is part of the politics of everyday speaking and it
is in this realm that the framing of information is shown to be closely
associated with the assignment of responsibility.

5.8 Truth and metacommunication


When participants tell each other or infer what is “really” going on, they must
be able to separate different levels of communication. In modern symbolic
logic, this ability is talked about in terms of the distinction between an “object
language” and a “meta-language” (Tarski 1956: 167), that is, a language used
to talk about another language. This distinction is crucial for assigning truth-
values. Bateson’s work on metacommunication and interpretive frames (Bate-
son 1972) recontextualized this distinction – in particular Russell and White-
head’s theory of logical types and their solution of the liar paradox – and made
it relevant to communication theory in general. Bateson pointed out that it is
through the ability to communicate at different levels that we can engage in
such complex activities as play or criticism. We not only communicate mes-
sages to one another, we also communicate about how to take or interpret those
messages. In some cases, these different levels can be used with damaging
results. Thus, according to Bateson, the conflicting messages provided by a
mother who says “I love you” to her child and then moves away when the
child gets close to her may create a potentially dangerous “double blind”
situation.
Inspired by Bateson’s work and by Goffman’s (1974, 1981) repeated
attempts to deal with these issues, contemporary linguistic anthropologists
have been studying the uses and functions of metacommunicative devices in
everyday interaction in general and in socialization contexts in particular (cf.
Briggs 1986; Schieffelin and Ochs 1986a). One of the recurrent points in this
body of literature is the distinction between a basic, propositional, or even
“literal” meaning, and a set of “additional” devices used to refine or modify it.
John Gumperz’s (1982, 1992) notion of “contextualization cue” covers a large
class of linguistic resources used by speakers for communicating the appropri-
ate interpretation of a given speech act: intonation, emphasis, tempo, rhythm,
lexical choices, etc. One of the things that this alerts us to is the ability that
speakers have to both give and take back a message or some aspects of it. The
study of contextualization cues also illustrates the hidden mechanisms through
which participants are judged competent in a given social encounter. Gum-
perz’s work on “crosstalk” makes some strong hypotheses on sources of
Truth and action 131

misunderstanding between participants with different linguistic and cultural


backgrounds.
A question that still remains to be answered is “How do we know, in any
given case, what is used to contextualize what?” In other words, how do we
decide what constitutes the “meta-level”? Are there linguistic (and non-
linguistic) mechanisms that are always indexing the framing of talk (e.g.,
prosodic features, markers of reported speech)? Or is it instead the case that
anything can be a contextualization cue for anything else? How can we decide
exactly and exhaustively where in a given text the participants are telling each
other that what they are doing is serious or that it is playful? The various parts
of the message are not always ready-made for the participants (or for the
observer) to be picked up separately. Precisely because we need to rely on
the conventionality of certain choices, we must also realize that the acceptance
of a given act or part of an act as the framing (ground) for another (figure) is
itself an act which may be questioned or rejected (Goodwin and Duranti 1992).
We should be asking ourselves whether, for instance, we should be influenced
by orthographic conventions in privileging certain aspects of language (e.g.,
the production of strings of recognizable sounds) as primary and certain other
aspects, typically non-segmental (e.g., prosody) or non-verbal (e.g., body
postures) as secondary. Eventually, we might have to face the issue of power
implicit in these analytic choices. Framing something as the foreground and
something else as the background may indeed be part of acts of authority.

5.9 Truth and action


Within philosophy, Austin (1962) pointed out that the truth-functional view of
interpretation can be used for a limited set of utterances, namely, those whose
function is to make assertions, that is, describe a given event or state-of-affairs
(this set of utterances or speech acts is variously called “statements,”
“representatives,” “assertives,” “constatives”). But there are utterances such
as commands, promises, requests, declarations, apologies, etc. for which truth
seems the wrong dimension for their interpretation. If someone gives a com-
mand, e.g., I order you to stop!, it cannot be evaluated in terms of truth, as
shown by the fact that one cannot reply That’s false (or That’s true) (Levinson
1983). The same is the case for requests (Give me a quarter) or declarations
(You’re fired!). For this reason, Austin proposed to replace truth-conditions
with “felicity conditions” – which became “conditions of satisfaction” for
Searle – and the notion of “meaning” with “force.” Words do not simply
describe the world, they also constitute it in the phenomenological sense of
making something show itself in a certain way (Zahavi 2003: 72–77). This is
true of the language with which we describe objects of perceptions (as argued
by Whorf) and, in more dramatic ways, of the language that makes certain
132 Intentionality and truth, revisited

actions into what they are, including polite requests and excuses, narrative
accounts of the past, and plans for the future.
Does this mean that truth ceases to be important or relevant for a theory of
interpretation? Not necessarily. Rather, what we learn from looking at actual
language use is that truth is often embedded in or functional with respect to
other dimensions of human understanding and human praxis. After all, argu-
ments – whether public or familial – are often centered on issues of truth
(Duranti 1990; Haviland 1989; Hill and Irvine 1993; Lindstrom 1992). But the
truth of social action has a different flavor from the truth of logical connect-
ives. In social life, truth cannot be taken as a given, either as a static property of
objects or events or as a property of their relations. Rather, it must be thought
of as a process, which requires action to be realized. This is a view that seemed
natural to pragmatists like William James:
Truth happens to an idea. It becomes true, is made true by events. Its veridity is in fact
an event, a process: the process namely of its verifying itself, its verification. (James
[1910] 1963: 89)
This conception is also common in anthropologists’ attempts to capture a
contextual, emergent notion of truth that feels to them different from the
“western” conceptualization with which they were educated. For example,
Lawrence Rosen’s (1985) characterization of the Moroccan concept of truth
in statements about relationships should please old and new pragmatists:
It is, again, like a price mentioned in the market, which is not “true” in any sense until
the parties make it applicable to their circumstance. Similarly, in Moroccan social life, a
statement about relationships has no truth value until something more happens, until
that utterance receives some validation . . . By characterizing all unvalidated utterances
as having no truth value leeway is granted for the negotiation of one’s personal
attachments. (Rosen 1985: 57)
What we are told by Rosen and others who have investigated these issues in
the field is that the process whereby members of a given society produce
acceptable versions of reality is embedded in local theories of what constitutes
an acceptable account and who is entitled to tell the facts and assess their value
and consequences (Lindstrom 1992). When words are used to define truthful
statements, we need to remember that, being deeds, the combination of words
in meaningful ensembles is but an attempt to both portray and shape the world.
It is not by accident that Austin’s view of words as actions was pre-announced
in Bronislaw Malinowski’s discussion of language as action in the second
volume of Coral Gardens and their Magic (1935). In looking for a reality that
matches the linguistic descriptions members of a given culture provide
(whether willingly or inadvertently), ethnographers are also seekers of truth
and as such they might fall prey to the correspondence or agreement theory of
truth. At the same time, this discovery process often forces us to recognize that
Conclusions 133

truth changes and multiple truths are possible if the spatio-temporal and
historical context changes. This is one way of interpreting Geertz’s (1973)
notion-practice of thick description. To that, dimensions of power relations
have been added through the study of Foucault’s (1973, 1980, 1988, 2001)
writings about the technology of self and the practices that legitimize where
truth should be found, e.g., in the institutional recognition of a person as law-
abiding, moral, healthy, or clean.

5.10 Conclusions
In this chapter, I have reviewed a number of dimensions of human communi-
cation that raise doubts about the ways in which intentionality and truth have
been used in explaining action and interpretation. My discussion has been
organized around four fundamental aspects of interpretive processes, namely,
(a) the medium (or code) through which representations are constituted, (b) the
co-participants (e.g., audience) in the performance/interaction, (c) the cultural
context, and (d) the actions constituted through representation. I suggested that
the very notion of “code” needs to be reconsidered so as to allow for the
interactional complexity documented in recent studies of language use and
language socialization. The specific properties of the code as a dynamic set of
discourse patterns and strategies afford (in Gibson’s sense) particular inten-
tional states and matching between encoded messages and possible (socio-
economic) worlds. Different media or different genres of speaking may also be
imbued with different truths or different ways of accessing information about
the world. A question that would be interesting to pursue is the extent to which
different media/codes/genres carry within them, in a fossilized form, as it were,
certain collective or private intentions and certain truths. Previous almost
exclusive interest in the speaker’s point of view has been expanded to include
the study of the role played by the addressee or audience in any kind of
communicative act. One of the devices through which contingent cultural
interpretations are made possible is through the embedding of activities within
larger ones and through the interpenetration (Cicourel 1992; Leach 1972) or
intertextuality (Kristeva 1986) of speech genres and speech activities (viz. the
ability of one symbolic form or message to index another form or message in a
different spatio-temporal domain). The very distinction between foreground
and background information, for instance, is a potential weapon in acts of
social struggle whereby a given version of reality is complemented or
threatened by another. Finally, the multiplicity of goals typical of any semiotic
act when matched with co-occurring codes and embedded in larger activities
argues for a different status of intentionality and truth across socio-cultural
domains (e.g., domestic, political, ritual, etc.). If any act of representation is
a social act, the assignment of intentionality and the assessment of the
134 Intentionality and truth, revisited

truth-value of a given expression are themselves social processes that need to


be understood for their social as well as their cognitive properties. Ethnog-
raphers seem to be in a special position for providing major contributions in
this area. At the same time, as shown in the rest of this book, their observations
must be empirically grounded by means of a careful analysis of discourse
patterns as produced by social actors in the course of their spontaneous
interaction.
6 Speaker intentions and the role of the audience in
a political campaign in the US

6.1 The discovery of what we mean1


In Chapter 3, through a detailed analysis of portions of the discussion of a
Samoan fono, readers were introduced to a discursive universe where some-
one’s intentions are not evoked even when they might seem relevant and
needed to defend someone who is blamed for someone else’s lack of fulfill-
ment of a public commitment. In Chapter 5, I reviewed how the ways in which
we understand the role of intentions in communication and the criteria for
assessing truth claims are affected by the audience, the context, and the
medium of interaction, including the specific speech genre used by speakers.
In this chapter I will extend my discussion of speakers’ intentions to
contents and contexts that are likely to feel more familiar to many of my
readers, namely, public speeches made during a political campaign in
1995–1996. But the point I will make is similar to what I discussed in previous
chapters: there are contexts and interactions where it is difficult for speakers to
control the meaning of their utterances. Audience members may have inter-
pretations that a speaker cannot predict or change. Speakers, in turn, can
“repair” what they just said or modify some of their message the next time
they happen to talk about the same topic. By giving us the opportunity to
examine speeches that were meant to be the “same” speech, political cam-
paigns make evident what also goes on in everyday life. We often repeat the
same story or make the same argument on different occasions. Each time, we
make some adjustments to our audience.
One of the most well-known studies of this feature of talk is Charles
Goodwin’s (1979) analysis of the ways in which the speaker changes the type
of speech act that he or she produces in the course of the same utterance and
within the same turn as his or her eye gaze moves from one recipient to the
next. This study was meant to illustrate several important characteristics of talk
in interaction, including the phenomenon that Harvey Sacks (1992) had called
“recipient design.” This is the recognition of the fact that we “design,” that is,
craft our utterances so that they can be made sense of by specific others. We do
this routinely, sometimes unconsciously sometimes with predetermination. For

135
136 Speaker intentions and the role of the audience

example, in identifying people, we use expressions that can be made sense of


by our audience, that is, we rely on our shared knowledge and the context of
talk. When someone says “the president,” he or she assumes that the audience
will be able to know which of the many possible people in the world who
could be called “the president” is being talked about at this particular moment.
At a meeting of a professional association, “the president” may refer to the
president of that association. When a politician talks at a rally, “the president”
may be the current President of the United States. When telling a story about
the past, “the president” may be someone who is no longer alive, e.g., in telling
an anecdote about Eisenhower or Kennedy. The “design” of our utterances is
not only about reference. It is also about what can be said to whom. For
example, we usually do not repeat the same piece of news to the same person.
As Goodwin (1979) showed, we “design” what we are saying to show that we
recognize the fact that one of the members of our audience already knows the
event we are reporting.
The fact that speakers shape their utterances taking into consideration the
knowledge that their interlocutors have and the occasion (e.g., a political rally
vs. a professional meeting) obscures the fact that in reality they might not fully
control some of the pragmatic implications of their speech. I believe that this
has to do with the fact that speaking (as opposed to reading a text) requires a
considerable amount of improvisation even when speakers believe that they
are following a script (in their head or from some notes in front of them).
Phenomenological philosopher Merleau-Ponty has captured this aspect of
speaking as a search for one’s own intentions.
For the speaking subject, to express is to become aware of; he does not express just for
others, but also to know himself what he intends. (Merleau-Ponty 1964: 90)

If speakers’ intentions are discovered through speaking, there is both creativity


and vulnerability in our use of language. We are both as inventive and as
vulnerable as our language and our interlocutors allow us to be. Leaving aside
the constraints put upon us by linguistic conventions, in this chapter
I concentrate on the role of the audience in helping and sometimes forcing
speakers make sense of what they are saying, pushing them to acknowledge
that their speaking might have a force or a directionality that they did not
expect (see also §6.3.2). Through their audience’s reaction speakers can
suddenly hear themselves speak or even discover themselves to have unknown
intentions, mixed with emotions, prejudices, or cultural preferences. Rather
than being restricted to the intimate exchanges that take place between a
therapist and a client/patient, these moments of revelation or self-discovery
can happen in public arenas as well.
The context of the political campaign offers us a rare opportunity to see how
speakers who are working hard at monitoring the content of what they are
When the audience has a different take 137

saying may discover that they cannot control the ways in which the audience
interprets their utterances. The audience, in turn, may not be concerned with
what the speaker actually meant.

6.2 The Capps-for-Congress Campaign2


In 1995–1996, I had the opportunity to closely follow and document the
political campaign of Walter Capps, then a Professor of Religious Studies at
the University of California at Santa Barbara, for the US Congress.3
Following one candidate across various situations gave me the opportunity
to document different versions of what can be considered the same stories,
assessments, promises, attacks, questions, answers, and introductions. This
opportunity, in turn, allowed me to evaluate the extent to which a candidate
like Walter Capps had to accept his audience’s interpretation of his words or
could instead resist such interpretation. As I show in this chapter, the live
interactions between the candidate and the voters provide us with some
telling examples of a struggle of interpretations that becomes particularly
conspicuous in political arenas, where candidates must balance their own
goals and moral values with the values of their actual or potential supporters.
As media coverage of local and national races have shown again and again, it
is common for candidates to move closer and closer to the ideal projected by
their audience responses and in so doing hide, ignore, or change their own
original intentions and lose track of the motivations for running for office or
espousing a particular cause with political ramifications. What one might
analyze as an issue of authorship – e.g., who is the author of this message? –
can be easily turned into an issue of coherence of the self or strategic lack
thereof. The instrumentally oriented political field ends up producing a type
of social persona that contradicts the individualistic ideology that is said to
permeate our social life (Sennett 1974) in the West.

6.3 When the audience has a different take


Our audience is important for the way we describe people, objects, and places
(e.g., Schegloff 1972) or we report an event (e.g., Goodwin 1979, 1981).
Audience members also actively participate in guiding the direction of talk,
even in a formal event in which they are expected to give minimal responses at
largely predictable points. Audiences have ways to display their approval or
disapproval of a candidate’s stance with respect to a certain issue or toward the
actions of another candidate. Steven Clayman (1992, 1993), for example,
showed that audiences may use laughter to display their disbelief in a particu-
lar claim that has just been made by a candidate. In such cases, the audience is
usually seen as taking a stance that is clearly in opposition to the candidate
138 Speaker intentions and the role of the audience

(e.g., refusing to believe the candidate’s claim to be a strong supporter of a


particular type of policy). In the data that I will analyze below, however, the
audience is manifestly approving and sympathetic. The laughter produced, in
other words, is meant to be affiliative, not disaffiliative. And yet, the audience
seems to be taking the meaning of the candidate’s words in a direction that he
did not seem to expect. In speaking in front of his supporters, Walter Capps
discovered his words to have a “force” that he did not expect and had a hard
time controlling.
To illustrate this situation, I will start by examining the verbal interactions
recorded during the first official day of the campaign, on November 14, 1995,
when Capps announced his candidacy to groups of supporters and potential
voters at several places in his district.

6.3.1 Controlling unintended humor


In the first speech of his 1995–1996 campaign, in Paso Robles on the morning
of November 15, 1995, Capps makes two consecutive statements that are taken
to be humorous by his audience despite the fact that, as I will show, no humor
seemed intended by the speaker.
(1) (November 15, 1995, Paso Robles)

W. Capps; but the second announcement is just as important. . . .


and that is we- that- that we will win this time.
we will win this time.
Audience; ((clapping)) yeah:::
((more intense clapping, cheers))
W. Capps; ((smiles, visibly pleased by the audience reaction))
W. Capps; and how do- how do- how do I know that? ((see Figure 6.1))
how do I know we’re gonna win?
Audience; ((laughter starts)) hehe ((increases)) HEHEHE!
W. Capps; ((smiles at the audience)) well, you know, I can see it in your
faces. (I mean-)
Audience; ((laughter)) haha//ha
W. Capps; and- and- and I- and I- and I mean that totally because-
because . . . uh, ((points to his wife Lois))
Lois and I . . . have lived. here,
in fact the first time we came in here in- August. of 1965 [. . .]

I want to show that Capps is surprised by the fact that the rhetorical question
“and how do I know that?” and the statement “I can see it in your faces” are
interpreted by the audience to be funny, even though, as shown in Figure 6.1,
he is actually smiling while delivering the rhetorical question (“and how do
I know that?”). The humorous interpretation by the audience is thus at first
apparently endorsed by Capps himself who is smiling with satisfaction at their
When the audience has a different take 139

Figure 6.1 Walter Capps speaking to a group of supporters in Paso


Robles, California, November 15, 1995, saying “How do I know
that?” – see example (1)

support – the clapping and cheers that greet his second pseudo-announcement
(“we will win this time”) – but quickly moves on to his planned next point: the
claim that he would be a good representative in the US Congress because he
knows the people of the 22nd district. When we examine the next stretch of
talk, we find a narrative of personal experience full of old memories and
nostalgia, as shown by the coda: “You are . . . the people with whom we’ve
lived our lives,” which constitutes a move away from the light mood estab-
lished by the laughter that accompanied his rhetorical question. An additional
sign of the fact that Capps is trying to reduce the force and potential implica-
tions of the audience response is his metanarrative statement (Babcock 1977)
“I mean that totally,” which he produces just before transitioning to the
personal narrative. “I mean that totally” makes sense only if we assume that
he is not fully (i.e., not “totally”) satisfied with how the audience thinks he
means it. Here is the passage all the way to the climactic end of the personal
narrative:
140 Speaker intentions and the role of the audience

(2) (November 15, 1995, Paso Robles, cont.)


W. Capps; that I- and I- and I mean that totally because- because . . . uh,
Lois and I . . . have lived here, in fact the first time we came in
here in- August of 1965, we stayed across the street. we- we
came out from- from uh, Yale University, uh to teach uh at UC
Santa Barbara. and we came down from Oregon. we stopped
across the street, had a- . . . had a-.. we were carrying a- trailer
with uh, our belongings. we didn’t have any children then¼that
was in nineteensixtyfour. . . . we’ve been here a:ll this time. . .
we’ve lived here a:ll these years. we know the people. . . of the
twenty-second district. . . . you know- . . . our. children were
born. in the twenty-second district. they’ve all gone to school
here. . . . uh so what I’m suggesting is, . . . not only suggesting
I know this to be the case: that I represent . . . majority. opinion.
in the twenty-second district. I mean¼I know what people in
the twenty-second district believe in because- these are our
people. . . . you are- . . . the people with whom we’ve lived our
lives.
The narrative of belonging reproduced in (2) is anything but humorous and, in
fact, could be seen as an attempt to move away from the humor evoked earlier
and to bring out some different emotions. After recounting events that prove
that he and his wife Lois have been in Santa Barbara for a long time (since
1965), Capps mentions the fact that their children were born and raised in the
congressional district in which he is running. The fact that they attended public
schools adds to the narrative of belonging while simultaneously hinting at the
importance of state-funded public education. All together, the experiences
mentioned by Capps in excerpt (2) are likely to be shared by the audience in
Palo Alto that morning, which was mostly made of senior citizens. They are
the people with whom the Capps have “lived [their] lives.” In making this
emotional connection with the people of the 22nd district, including those who
are standing in front of him at that early time in the morning, Capps is
definitely moving away from any possible humorous reading of his words.
In his second stop of the day, in San Luis Obispo, Capps speaks from behind
a podium on which he has placed the text of the speech he and his team had
worked on the night before. In this speech, when he gets to the “announce-
ment” about winning, he uses “we” instead of “I” and asks “how do we
know . . . we’re gonna win?” Since he had no text in front of him in Paso
Robles but he had the text at his podium in San Luis Obispo, it is possible that
the “we” was the original written version. It is possible then that Capps had
learned his lesson in Paso Robles and decided to stay closer to the prepared
speech precisely to avoid the humorous reading of his rhetorical question. He
returned to the first person singular a bit later in the speech (“thank you. How
do I know that?”), once he had avoided the humorous interpretation.
When the audience has a different take 141

(3) (November 15, 1995, San Luis Obispo)


W. Capps; and then the:: um- and the second announcement I think is
probably even more important. . . . the second announcement
is. that we. will. win. we will win.// in November.
Audience; ((clapping, cheers))
W. Capps; how do we uh- how- how do we know we will- how do we
know we’re gonna win?.. uhm- I- I- I can tell you. the reason
we’re going to win is that- uhm- that I have. every
confidence. that I represent . . . majority . . . opinion- the
majority (of) viewpoint . . . of the people in the 22nd.
district. of California. . . .
how do I know // I do that?¼
Audience; ((sparse clapping))
W. Capps; thank you. how do I know that? because we’ve lived here for
thirty-two years. we’ve, ou::- our children were born here,
I’ve taught here uh a:ll that time. we’ve lived among you. we
know what you think about things. ((looks at notes)) uh
((looks up)) we- we know your points of view, we know
your attitudes, we know your beliefs, we know your
convictions.

Figure 6.2. Walter Capps speaking in San Luis Obispo, November 15, 1995,
saying “How do we know we’re gonna win?” – see example (3)
142 Speaker intentions and the role of the audience

The restrained, serious “key” of the San Luis Obispo speech is particularly
striking given that after the Paso Robles speech, while he was going around
shaking hands with members of the audience, at least two of them encouraged
Capps to “keep up the humor” in his speeches. One possible explanation of
such a switch of interpretive key (Hymes 1972a) is that Capps had learned
from the Paso Robles performance that if he did not want the audience to
decide for him what was funny and what was serious, he ought to keep his own
joking under control. In San Luis Obispo, he never smiled throughout any of
his statements or rhetorical questions and the audience did not interject any
laughter.
One might argue, however, that the San Luis Obispo speech is not a good
indicator of how Capps conceptualized the rhetorical question “how do I know
that (I’m gonna win)?” precisely because of the overall lack of humor through-
out the speech (the audience will eventually produce some laughter after the
speech is over, when Capps, prompted by his wife, speaks again to acknow-
ledge some of the people in the audience). Given this possible objection to
what to make of the San Luis Obispo speech, the version delivered at his third
stop, at Hancock College, is potentially more revealing. On this occasion,
speaking inside a classroom where he had been invited by the instructor of a
political science course, Capps returns to a more informal tone without reading
from a text and does not shy away from humor. In fact, he starts out his speech
with what will become one of his favorite jokes throughout the campaign – the
“seventy-five minutes joke”:
(4) (November 15, 1995; Hancock College)

W. Capps; I’m- uh- I’m very happy to be here today and:- . . . uh- see the
problem with this is that I’m- I’m so used to this format . . .
that I ma- I may go on here for- . . . you know seventy five
minutes because-
Woman; ((laughter)) ha-ha-ha!
W. Capps; because the classes in Santa Barbara are an hour and fifteen
minutes. but, ((clears throat)) [. . .]

Even though Capps continues with a narrative that has a humorous punch
line (analyzed in Duranti 2003b), when he gets to the line about his confidence
in a victory (“we’re gonna win . . . we(‘ll) win”), he leaves out the “how do
I know that?” question altogether, thus removing the piece of talk that had
been followed by laughter in Paso Robles. In this context, the narrative of
belonging is also slightly altered, probably to fit the occasion where there are
mostly young people (the students in the political science course). The emo-
tional tone of the earlier version (in Paso Robles) is gone and the point about
knowing the people of the district is made explicit (“we belong here”.) In fact,
the people are talked about in the third person: “we know the people. So well
When the audience has a different take 143

that I know I can represent. their views in- in Washington.” This further
removes the possibility of the empathetic bond established in Paso Robles
with the senior citizens.
(5) (November 14, 1996; Hancock College, Santa Maria)
W. Capps; the second announcement is- . . . and I- I have total confidence
on this one, we’re gonna win. . . // we(’ll) win.
Audience; ((clapping))
W. Capps; and: uh, I think the reason we’re going to win is-.. becau::se
I understand the people. of the twenty-second district, . . . uh
I know their views, . . . I know what they want . . . uh- a
representative to do. in Washington. and I’m committed to
doing that because I’ve- we have- . . . lived our- lives with- . . .
these people- I’ve been here for- thirty one years as a professor
at- UCSB, . . . our children were born and raised here- so
um- . . . you know- we-uh, we belong here ((CL)) and we- and
uhm- . . . and-uh . . . we- we know the people. so well that
I know I can represent. their views in- in Washington.

Figure 6.3 Walter Capps speaking at Hancock College, Santa Maria,


November 15, 1995, saying “I think the reason we’re gonna win . . .” – see
example (5)
144 Speaker intentions and the role of the audience

The skipping of the planned question suggests that, by the third speech of the
day, Capps had found a way to have some control over the potential implica-
tions of his words. The insertion of humor earlier on in the speech had to be
balanced by the removal of verbal material that could be interpreted as a
continuation of that humor.

6.3.2 Can one attack without “being mean”?


In his presentation of his political persona, Walter Capps was faced with
problems – often leading to paradoxes – which are not uncommon for those
running for office in the US. One was the problem of having to explain why he
wanted to leave a respectable profession and a thirty-two-year successful
career at the University of California in order to embrace a career that, being
associated with moral compromises and even corruption, was not equally
respected by the public. Another problem was created by the pressure he felt
from his own political advisors to attack his main opponent, the incumbent
Rep. Seastrand (Republican). How could he criticize her without damaging the
image of fairness and compassion that he had acquired in his teaching –
especially through the highly popular and highly publicized course on the
Vietnam War? As shown by the transcripts of his speeches on the first day of
his new campaign, Capps was well aware of these problems and was working
hard trying to deal with them.
Capps dealt with the problem of the shift to a job that was perceived as
potentially less respectable by presenting himself as a candidate who was
following a call. This choice was inspired, in his own words, by the Jefferso-
nian model of the citizen-representative who goes to Washington as a form of
duty to his country and not for personal ambition.
The second problem – how to attack his main opponent without sounding
unfair or mean – was more difficult to solve. Capps’ strategy at first was to
explain to his audience that criticism was part of running for office. He was
hoping that he could do it in some straightforward fashion that would allow
him to avoid having to sound unfair or mean-spirited. In devising this strategy,
however, he had not taken into consideration his audience’s expectations,
grounded in highly antagonistic and hypercritical contemporary American
political discourse. He had not realized that the image of himself he was trying
to project through his utterances was going against deeply rooted assumptions
about political speeches and political debates. Contrary to his own anticipation,
his supporters did not always allow him or want him to be “Mr. Nice Guy.” On
the contrary, they expected and found “meanness” in his words, regardless of
his framing or his disclaimers. The transcript in exerpt (6) shows how Capps’
progressively intense criticism of Seastrand is met with laughter by the audi-
ence which exposes to him a possible interpretation of his intentions that he
When the audience has a different take 145

does not want to accept. His attempt at fixing the problem by reframing his
own words (“I’m not being mean here”) is greeted with even more laughter,
confirming the audience’s reading of his earlier words as purposely critical and
of his added disclaimer as sarcastic.
(6) (November 14, 1995, Paso Robles)
W. Capps; now, you know when you run for office, . . . it isn’t just like
applying for a job¼I mean you have to- you’ve gotta beat the
uh- . . . you’ve gotta beat the other candidate. . . . and this is
what I say about her . . . she doesn’t represent . . .// the
majority
Woman; no. she doesn’t.
W. Capps; she does not represent the majority . . . viewpoint. of the
people of the 22nd district. . . .
in fact, I don’t think she represents anybody. // in the 22nd
district.
Audience; ((laughs)) hehehe! hahaha!
W. Capps; (???) I’m not- . . . I’m not being mean here,
//I’m not at all being mean.
Audience; ((laughter)) hehehe. //haha

What leads the audience to laugh on both occasions? The first laughter is
probably produced by the contrastive pair of statements shown in (7) – a well-
known technique also used to evoke applause (Atkinson 1984: chapter 3):

(7)
A. she does not represent the majority . . . viewpoint. of the people of the
22nd district. . . .
B. in fact, I don’t think she represents anybody. // in the 22nd district.

By uttering (A) Capps sets up an implicit contrast whereby if the incumbent


does not represent the majority, then, she should be representing the
“minority.” But then in (B) he denies that implication by stating that she does
not represent anybody. The violation of the expectation has a humorous effect.
If we stick with the hypothesis that the second speech of the day, in San Luis
Obispo, is closer to what he had planned to say, in this part of the speech in
Paso Robles Capps was supposed to state that Seastrand “does not represent
the people of the 22nd district of California.” By mentioning “the majority,”
Capps brought up a contrast (between majority and minority) that did not fit
with the line about Seastrand not representing (all) the people in the district. In
addition, the use of “in fact” and the quantifier/pronoun “anybody” – with the
emphasis on the first syllable – contribute to the negativity of the statement that
provokes an immediate and affiliative response from Capps’ supporters in the
form of laughter. It is at this point that Capps seems to realize that the audience
146 Speaker intentions and the role of the audience

does not see his criticism as straightforward and definitely not as innocent. His
first attempt to fix the problem is the disclaimer “I’m not being mean” which
evokes additional laughter. And his repetition of the disclaimer with the
additional “at all” (“I’m not being mean here. I’m not at all being mean”) does
not solve the problem. A bit later, as shown in (8), Capps tries to regain the
high moral ground by claiming restraint during the first year of Seastrand’s
tenure in the House. However, the addition of the hypothetical “whether
anyone believes this or not” and the emphasized repetition “I mean I really
wanted her to do well” give away his fear of not sounding sincere even to a
group of supporters.

(8) (November 14, 1995, Paso Robles)


W. Capps; but now, I have not critiqued her in a- in a full year
because, . . . whether anyone believes this or not, . . . you know
she’s my representative too, . . . and I wanted her to do well,
I mean I really wanted her to do well because I want- people in
our district, . . . to be well represented. in Washington.

At the next stop, in San Luis Obispo, Capps sticks closer to the written speech,
avoiding introducing personal statements and going instead straight into a
series of rhetorical questions about her performance. This time there is a
metapragmatic cue (“I’ll just have to ask you”) that lets the audience know
that they are expected to participate in a question–answer drill that evokes the
“call and response” format found in many African American church sermons.
(9) (November 14, 1995, San Luis Obispo)
W. Capps; the reason I know we can beat my opponent is because our
opponent next fall does not represent the people of the twenty-
second district. of California.
I’ll just have to ask you,
does our representative represent the seniors of the twenty-
second district?
Audience; NO::!
W. Capps; does she represent . . . the students of the twenty-second
district?
Audience; NO::!
W. Capps; does she represent the children, of the twenty-second district?
Audience; NO::!
W. Capps; does she represent the women, of the twenty-second district?
Audience; NO::!
W. Capps; does she represent the people who care for the environment of
the twenty-second district?
Audience; NO::!
When the audience has a different take 147

W. Capps; does she represent the people who believe in local government
of the twenty-second district?.
Audience; NO::!
W. Capps; who does she represent? she- she campaigned- I heard it.
I heard it. she campaigned- she campaigned on the theme a
leadership- leadership that listens. well I can tell you the voice
to which she pays the most attention. and the voice- to which
she pays the most attention is the voice of Newt Gingrich.
Audience; YEAH!
W. Capps; and we didn’t elect-
Audience; boo:::!
W. Capps; we didn’t elect Newt Gingrich. We didn’t elect Newt
Gingrich. We elected a representative
Audience; (b)oo::!
W. Capps; and we need to hold the representative to the charge that the
representative has been given.

In this exchange Capps lets the audience co-author the criticism in a systematic
and orderly way that he can control. He will follow the same strategy on the
campus of his own university (UCSB) but will change it at Hancock College,
where the audience included people who could not be counted on as support-
ers. In this context, instead of inviting the audience to answer the rhetorical
questions about who Seastrand represents, Capps takes off from Seastrand’s
slogan “Leadership that listens” to ask a series of questions that he answers
himself, simultaneously avoiding a test of the audience loyalty.
(10) (November 14, 1996; Hancock College, Santa Maria)
W. Capps; I’ve been asking myself, has she listened to the seniors . . . in
our community? I don’t think so. . .
has she listened to the people who want to protect the
environment, . . . in our community?
I don’t think so. . . .
has she listened to the people who are advocates of education?
that one kills me because- [. . .]

This is a much more mitigated form of criticism, with the personalized “I don’t
think so” instead of the well-timed collective shout “NO!!!” obtained in San
Luis Obispo. Furthermore, Seastrand’s actions with respect to certain initiatives
like her alleged record on education are presented as having a direct impact on
Capps himself as a believer in education (“that one kills me because- . . .”). By
presenting himself as a victim of Seastrand’s voting record and beliefs, Capps
puts himself in the same category of the students he is trying to attract to his side.
A different form of mitigation of the attack is used by Capps on the campus
of UCSB, where the audience, which comprises some of his own students, is
148 Speaker intentions and the role of the audience

very supportive, making him seem more at ease and willing to insert new parts
in his speech. In this context, standing at a microphone on a platform (where a
rock band had just finished performing) and without the help of any notes,
Capps launches into a new story about his own inability to be “critical,” which
ends with Seastrand becoming the butt of an obviously planned joke.
(11) (November 14, 1996; University of California, Santa Barbara)
W. Capps; uh, uh but what I’ve been telling people is that running for-
running for office is not . . . quite like just applying for a
job¼you know, not, just that- that you’re qualified . . . and you
get the job. I mean it- to run for office really means . . . that
you have to beat your opponents. and the only way that this
can become successful, . . . is if I beat Andrea Seastrand in
November.
Audience; ((clapping, // cheers))
W. Capps; well, . . . people who- . . . people who know me well . . . know
that . . . I have had some difficulty in the past in being critical
of other people¼because it goes against my nature. uh-
((clears throat)) uhm last time around I would say things like-
. . . you know¼I’ve heard it said that- . . . that Andrea
Seastrand is a: . . . very warm- . . . human being, . . . but
I looked up warm . . . in the dictionary . . . and it said “not too
hot.” . . .
Audience; ((laughter)) haha//hahaha!
W. Capps; now- see- that’s about- that’s about as far as- . . . as I can go
with- . . . critique because . . . I think one of the things we want
to fight in our- resist in our culture is this- this invective- this-
. . . this super- charged political rhetoric that- [. . .]

This excerpt and the ones above give us a glimpse of what I saw as a recurring
pattern, namely, Capps’ hard work at finding a way of criticizing his main
opponent in ways that could be approved by his audience without making him
feel uncomfortable with the implications that such attacks would have on his
own persona. Although Capps, perhaps due to his academic background,
tended to verbalize some of the challenges that he was facing, I believe that
his dilemmas and struggles are the same as the ones faced by any candidate for
public office.

6.4 Conclusions
In examining the speeches recorded during the first day of the 1995–1996
campaign for the US Congress by Democratic candidate Walter Capps,
I identified a number of exchanges where we can see Capps at work trying
to control the audience’s reactions to his utterances. This is made particularly
difficult in moments when a candidate finds himself without a script and starts
Conclusions 149

Figure 6.4 Walter Capps speaking on the campus of the University of


California, Santa Barbara, November 15, 1995, saying “this super-charged
political rhetoric that . . .” – see example (11)

to improvise on a theme that he has planned to address. We have seen that the
audience is anything but a passive recipient of made-up slogans, jokes, or
criticism. Audience members may react in ways that are surprising to the
speaker because they reveal meanings that the speaker had not foreseen. This
type of situation creates a difficult moral dilemma for political candidates,
especially when the audience reaction suggests that they approve what they are
hearing the candidates say. Speakers must decide whether to go along with the
audience and capitalize on their unexpected reaction or try to keep control over
the meaning of their own words. In our case, we saw how one candidate edited
his own speech and reformulated his stances with respect to a number of key
issues in his speech to retain some control over his own message and present
himself as a type of person who could be critical without sounding “mean.”
This type of struggle between the speaker’s voice and the audience voice is
fought at many different levels at once, involving specific sequences of acts,
specific grammatical and narrative frames, and, in our case, specific cultural
expectations about what it means to be a political candidate for the US
Congress in the last decade of the twentieth century. Above all, however, this
150 Speaker intentions and the role of the audience

struggle fought over the right balance between pleasing others and asserting
oneself reminds us of the fact that as political candidates refine their message
to meet the challenges of a political campaign, they are often making claims or
promises whose meanings or implications are not fully worked out until they
are made public.
In the next chapter, Teun van Dijk will claim that the audience-as-co-author
model, which underlies the discussion in this chapter, has nothing to do with
intentions, but I think it does. When dealing with their audience’s reaction to
their words, political candidates are forced to rethink their message and thus
reshape what they are saying. At the same time, when confronting an audience,
as well as their opponents, political candidates are also forced to reflect on
what they really mean. This is, in fact, a publicly and privately played out
process of self-discovery. Responsibility does seem to play a role in defining
(social) action because it encourages and even obliges speakers to investigate
the truth conditions of what they are saying. The public quality of political life
reproduces a type of ethnomethodological experiment (Garfinkel 1967)
whereby the unspoken assumptions are pushed to the foreground and people
are confronted with consequences of their utterances that they might have
ignored under different circumstances. Running for office is thus a bit like
having dialogues in public with lots of people who play the role of Socrates,
challenging you to reconsider what you just said or believed and showing you,
in the end, that you are an imperfect human being struggling for coherence and
cut out from the light of reason.
7 A dialogue on intentions

With Teun A. van Dijk and T. Jason Throop

7.0 Introduction
This chapter reproduces a continuous series of fairly long email messages
initiated by discourse analyst and cognitive scientist Teun van Dijk, who wrote
to me on November 1, 2004 to ask me “a theoretical question,” as he put it,
about my position on the role of intentions in discourse. A number of factors,
including electronic mail as a medium for unedited and informal exchanges as
well as the long-term professional relationship between the two of us (over
several decades van Dijk has involved me in a number of his editorial
projects), made possible what I hope readers will appreciate as a frank
exchange of ideas and positions that is rarely found in the more polished and
peer-reviewed journal articles or chapters of edited volumes. I have also
included toward the end of this chapter a message by my colleague Jason
Throop, whom I invited to join the discussion. His response to the exchanges
I had forwarded to him captures his own original ways of thinking about
intentions and anticipates some of the themes found in the following chapters,
themes that since 2010 have been incorporated into our jointly taught seminar
on the culture of intersubjectivity at UCLA.
In reproducing this exchange of emails as a chapter of this book, I have
resisted the urge to provide here a summary of the entire discussion for the
simple reason that almost each message contains some attempt to clarify both
the points of agreement and those of disagreement. I believe that any add-
itional, post hoc summary of mine could not do justice to the friendly tension
of the arguments whose main value is both the genuineness of the claims
contained in it. Hence, with minimal cuts (of greetings at the beginning and the
end of each message) and a few added references, here is what we wrote to
each other over a few days, in the fall of 2004.

7.1 November 1, 2004: from van Dijk to Duranti


May I bother you with a theoretical question?
In your paper for the (interesting) Hill & Irvine book on responsibility and
evidence in oral discourse [Duranti 1993b] you reject an “intentionalist”

151
152 A dialogue on intentions

approach to discourse, and so do (more or less explicitly) also other contribu-


tors, thereby rejecting the traditional pragmatic approach to the accounts of
speech acts.
Although I guess I understand the reasons for this rejection, I wonder what
the exact implications are. That is, the notion of intentionality is not invented
by Searle, but a property of the philosophy of action more generally.
Does your paper imply a rejection of the notion of intention for any kind of
action analysis, in the sense that you claim that ALL actions are performed
without intentions? And if so, do you reject any kind of concept that involves
conscious or unconscious mental preparation, planning or goals of action? Or
do you accept it, theoretically, but not as a researchable notion? Or do you
rather claim that whatever the intentionality of actions (speech acts, discourse),
it is not a notion used in all cultures – but rather, as you seem to suggest, that
intentionality is not individualistic but can be shared with others? Or, even
more restrictedly, do you claim (as others in the book) that some form of
discourse (performance, etc.) is being carried out without conscious, explicit
intention, i.e. more or less “automatically” (in trance, etc.).
More generally I wonder where this rejection (or skepticism) against inten-
tions comes from? Is this a general feature of current linguistic anthropology?
Or just a skepticism about the universality of traditional theories on illocution/
speech acts?
I ask this because in my book on context (van Dijk 1977, 2008), one of the
(cognitive) aspects of context are – in my theory – the intentions of the
participants (although formulated in a way that is different from speech act
theory), in line with the dominant theories of discourse production in psych-
ology, that is, as intentions accounted for in terms of (subjective, conscious or
unconscious) mental models of future activity. I would like to be able to refer
to alternative approaches, like those defended in the Hill & Irvine book, and
understand what exactly they imply.
There is no hurry to reply, but hope you can take a few minutes to explain
your own position and more generally the one by others in the field as
represented in this book.

7.2 November 2, 2004: from Duranti to van Dijk


I have often thought in the last few years that I should write again and more
extensively on intentionality, in part because what I wrote a long time ago (the
original version of the paper you read was written in 1984 when I was a post-
doc fellow at UCSD with Michael Cole and the more theoretical piece
published in Cultural Anthropology was written in the late 1980s) is ambigu-
ous enough to leave too much room for interpretation including the idea that
I reject intentions altogether (which maybe I did at some point). Recently
November 2: from Duranti to van Dijk 153

I have been working on “agency” in language, which I find a more interesting


notion, especially because it has a bit more empirical backing (e.g. my paper in
A Companion to Linguistic Anthropology Blackwell 2004). But I basically
avoided the notion of intentionality. Perhaps the following paragraph from my
entry on “Intentionality” in Key Terms in Language and Culture might clarify
a bit my current stance and help with some of your questions:
Grice’s and other rationalist accounts of human action typically rely on a commonsense
view of people’s beliefs and desires. For anthropologists, the crucial issue is whether it is
possible to separate intentional acts from the cultural context in which they are produced.
Since the route to interpretation relies on conventionality as much as on intentionality, it
is possible and in fact quite common that an audience respond to what they judge to be the
contextually relevant conventions, ignoring the issue of the speaker’s intentions.
Ethnographers also showed that the view that one can know what goes on in another
person’s mind is not shared by all cultures and a difference in power or authority,
sometimes defined in terms of expertise, may grant some individuals or groups the right
to interpret while depriving others of the right to reclaim their original intentions.
Furthermore, when we abandon imagined exchanges and look at actual interaction, we
learn that participants cannot always know what their actions or words are meant to
achieve, as shown by John Heritage for the particle Oh! in English. (Duranti 2001: 130)
In other words, I believe that
1. a great deal of what philosophers write about intentions is folk-theory, with
little empirical support;
2. the notion of intentionality can have different interpretations, which are not
usually clarified in the literature, from the most basic (say, in Husserl’s
view of intentional acts) to the most complex (e.g., the notion of planning).
Hence the tremendous confusion in the literature and in the claims made.
3. The relationship between agency and intentionality needs clarification
(Giddens started to do this in the late 1970s), considering the fact that in
natural languages the issue of doing something on purpose is not always
implied in the representation of human action, see the use of adverbs like
“intentionally” or “on purpose” for actions that philosophers would nor-
mally take to imply a strong sense of intentionality (a point made by
Jackendoff back in 1972).
4. A great deal of what is assumed as individual intentionality is in fact more a
kind of distributed intentionality, i.e., among participants, actions and
artifacts (a notion that I strongly believed in in the early 1990s but then
I abandoned because I felt it was hard to distinguish at the time from
distributed cognition) (see Chapter 10 of this book).
5. If you start with (a certain notion of) intentionality, how do you account for
the phenomenon of the audience as co-author?
6. I agree with Wittgenstein that an intention is embedded in its situation, at
the same time.
154 A dialogue on intentions

7. I think there is an anti-cognitivist bias in anthropology (and linguistic


anthropology), which tries to compensate for the cognitivist bias of so
much research on human action and language. The two groups hardly
communicate. It’s a real paradigm difference.

7.3 November 2, 2004: from van Dijk to Duranti


I had read (of course) your article on agency, and was just rereading it a few
days ago when I went again through the (excellent) Companion searching for
any work or references on context. I agree that a discussion of agency is more
interesting (more general, more basic) than on intentionality. I also agree with
many of the things you say – e.g., the intuitive (folk) nature of the notion in
much philosophy, etc.
I also see that, in general, also in linguistic anthropology, there is a tendency
to move away from cognitive notions and analyses. I see this most radically in
discursive psychology, which although they are psychologists they don’t want
to deal with the ‘mind’ at all. But also in conversation analysis, ethnomethod-
ology, much sociolinguistics, ethnography of communication and linguistic
anthropology – and the last despite (or because of?) the importance of cogni-
tive anthropology (D’Andrade, Quinn and Holland, Shore, etc. etc.).
I am afraid that I regret this movement away from cognitive analysis –
whatever critique one may have – because often it is throwing away the baby
with the bathwater. In my opinion we should do better cognitive analysis, at
the same time as better interactional and other social and cultural analysis, and
really integrate them. There is no way a good cognitive analysis can be
interesting without (a) a social, culturally shared dimension of various cogni-
tive notions – e.g. knowledge, and (b) without a detailed account of the
interactional aspects of the acquisition, uses, etc. of these cognitive notions.
But the reverse is also true. There is no explicit account of society or culture
without a cognitive dimension (e.g., of shared knowledge or other beliefs or
representations), and no detailed account of interaction without many cogni-
tive notions, such as personal knowledge and beliefs, plans, goals, and
interpretations.
Indeed, this is more generally true for “meanings” – and no account of
conversation or interaction dispenses with some notion like “meaning”: what-
ever they are, they are not “observable” either, and when they are social, or
interactionally ratified, this means that they also have a social dimension, but
no less also a cognitive one.
I guess the same is true for “intentions.” Whatever the perspective, one
cannot escape a combined interactional-social and a cognitive account. It is
not, as is often assumed in various of the directions of research mentioned
above, a question of “observability.” No abstract structures (from phonemes to
November 2: from van Dijk to Duranti 155

syntax) and no meanings are – strictly speaking– “observable.” In that sense,


intentions are no different from meanings. So how can we study them? In the
same way as meanings: (a) introspection, (b) verbal accounts, (c) inferences
from interaction and discourse given a theory of action. How do we know what
A intends when saying P? By examining (interactional) accounts of what they
want to say, why they are saying it, etc. How do we study “meanings”
attributed by interlocutors to what people do or say: in the same way: by
asking them, or studying the interactional consequences. The only difference is
that the first is an analysis of self-interpretation, and the second an analysis of
the interpretation of another person. The first is the meaning I attribute to my
activity, the second the meaning others attribute to my activity. These usually
are similar and overlap, so that communication is possible. There may be
reasons to give privilege to the interpretation of the Others, for social reasons,
but that is a fallacy, because the speaker is no less a participant than the
recipient. One may take the interpretations of the recipients as more valid,
because based on social rules and conventions – because they have no direct
access to “private” intentions. But that is also a fallacy, because the speaker
uses exactly the same rules and norms to link intentions (or knowledge, or
meanings etc.) into socially understandable utterances. In sum, whether deal-
ing with intentions or with meanings assigned to recipients, one basically deals
with the same processes of speaking, communicating, understanding, etc. but
only described from different perspectives, one from the point of view of
production, the other from the point of view of reception. Moreover, what
recipients try to do when understanding (participating in) discourse is not
simply to apply the social rules, etc. to get to the “public” meaning of what
is being said, but basically they want to know what THIS speaker is NOW
trying to say – as many conventional expressions show: “What do you mean?,”
“What are you saying?” etc. In sum, much (though not all) of discourse
understanding involves making hypotheses about the meaning (¼ intention)
assigned to the utterance by the speaker.
And instead of simply copying what philosophers say, I think it is perfectly
acceptable – though not therefore necessarily true – to follow what contem-
porary psychologists and other cognitive scientists say about the processes,
representations, or structures involved. NOT (only) to describe what is going
on, but to explain it.
Also in your quote you emphasize that intentions cannot be separated from
the cultural context. I fully agree. But who says one has to do that? In the same
way, the social or cultural context should not be separated from real, flesh and
blood members who have their own biography, beliefs and knowledge, and
personally engage in the discourse or other practices we are studying. Much of
what they do and believe is shared with others of the same group or culture.
Other things more personal and individual. Hence the pervasive individual
156 A dialogue on intentions

differences in ALL our data, of any culture. Intentions are only ONE aspect of
the whole complex process of speaking, acting, and interacting. Of course,
there are social and cultural conventions. But where are these? What is their
precise nature? Obviously, these are also sociocognitive: shared by members
of a community. How “shared”? Well, as forms of knowledge or beliefs, and
hence (also) as forms of cognition. And with other knowledge and social
beliefs, these conventions thus are able to influence the (personal) intentions
of individual speakers in concrete situations. I see no other way how sociocul-
tural notions such as norms, values, conventions can influence discourse or
action without this cognitive connection. There is no magic device (black box
etc.) that links society, culture, or interaction with talk or text – the only way
we can explain this in terms that are more or less empirically founded is
through some kind of brain-based ‘cognitive’ processes. Hence, we need
BOTH dimensions in the analysis, and any reduction in any one direction
would be theoretically unproductive.
Of course – as you say – audiences do interpret discourses often independ-
ently of the intentions of the actor(s), e.g., on the basis of contextually relevant
conventions. But what is the precise nature of these “interpretations”? As
suggested – are these more social, more observable, more relevant than the
action interpretations (intentions) of the actors? Why?
Also “western” legal principles take both views. On the one hand, usually an
infraction will count as such, whether or not a person intended to do so or not.
Also racist or sexist discourses will count as such, whether or not the speaker
intended them as such or not. Why? Because also the speaker is supposed to
know the conventions and intend and produce a discourse or action in this
way. On the other hand, the law distinguishes clearly between murder and
manslaughter, and hence whether or not an act was voluntary, intentional,
planned, controlled, could have been avoided, etc.
It is of course interesting, as you say, that not in all cultures one finds
evidence that people can know the mind of others. Of course, this is as such no
more proof that intentions do not exist, no more than the intuitive (folk)
distinctions we (or our philosophers) make in “our” culture. There are many
notions in cognitive theory that have or have not a correspondent “folk”
equivalent (beginning with the distinction between short-term and long-term
memory – where we only have the folk notion of “memory”). The question is
how to account for discourse and action generally, not only in terms of how
participants describe it. There is little doubt in my mind (!) that although in
some cultures there is no explicit way of describing intentions, or that dis-
courses or acts are attributed to some kind of “joint production,” we need some
kind of notion that is the “meaning” equivalent of an activity as assigned to this
activity also by the actors, and not only by the recipients or observers, etc. One
may use such vague notions as intentions, purposes, or goals, as long as one
November 2: from van Dijk to Duranti 157

knows theoretically what they are (e.g., mental models, with such and such a
structure, etc.).
I also would find it very surprising that in other cultures people would have
not any form of “introspection” in the sense of (a) “wanting” or “planning” to
do/say something, and (b) assuming that others do more or less the same thing.
In sum, there must be some idea about “other minds” in order for any
interaction to be possible. There are many proofs for that, and the most obvious
one is that of the presupposition of (culturally shared or previously communi-
cated knowledge): in all languages and all cultures speakers need to make
explicit all they know, but assume that the recipients already know a lot. This
is one way in which speakers already have a hypothesis about the beliefs and
hence the minds of others. Of course, this may be described and conceptual-
ized in many ways, also in our own culture, but the fundamental processes of
communication and interaction are the same.
Of course, as suggested, this does not mean – as is the case in our own legal
and normative system – that sometimes “good” intentions, etc. do not count (as
excuses, legitimations, etc.) in some cultures. But that is no proof that people
have no intentions.
Of course, it is also true that people do not always know how their discourse
will come across, be interpreted, etc. Of course not. Recipients will form mental
models based on the discourse structures PLUS their own personal and socially
shared beliefs, and hence production models will of course be different from
reception models. Again, this is a good reason to study misunderstandings, etc.,
but this is no proof that speakers do not have intentions. On the contrary –
a debate or conflict precisely presupposed such different interpretations of the
discourse, and hence also that of the speaker/actor, that is, the intentions.
True too that the everyday and philosophical terminology is confusing. In
the philosophy of action, there is a distinction between intentions and pur-
poses. An intention is the representation of the act itself, e.g., menacing or
kissing someone. The purpose is the representation of the state of affairs
(consequences) one wants to bring about by means of the act: e.g., that
someone will do or not do something.
True too that people may “do” things sometimes without intentions, pur-
poses, etc. (“without thinking”). These, however, are not usually described as
actions but as “doings,” e.g., when one has no control over one’s action
(because of drugs, sleep, etc.). Or when someone forces you to “do” some-
thing. Hence the degrees of diminshed responsibility provided by the law.
The notion of “audience as co-author” is a metaphor that has little to do with
intention. Of course an actor may take into account, and will usually take into
account, what recipients or co-authors want or believe. This will be part of the
construction of “intentions” (mental models of action). And if actions are co-
produced then this also is the case: there will be an explicit pre-agreement (and
158 A dialogue on intentions

negotiation etc.) about “what will be done together,” and this intention may be
shared like any meaning. The problem in that case is more a question of how to
coordinate the action – as is the case in conversation.
An intention is of course embedded in its situation (although this is again a
metaphor), but the point is that these two notions can be combined only when
a situation is not something “objective” or “real” etc., unless interpreted and
represented (constructed etc.) by participants or social members. For situ-
ations to be able to influence discourse and other social actions they first need
to be (subjectively) interpreted – and hence represented. And it is this
representation (model) of the situation that influences intentions and hence
actions.
In other words, however one formulates the issues, in my view they come
down to the same thing. Whether you call them meanings, or interpretations, or
intentions, or representations, or constructs, etc. – they are all personal or
socially shared mental models of some kind, and any discourse or act is
intended or understood by relating them with such “entities.”
True finally what you say: there is still a lot to be done, and I wished that
neither in the cognitivistic, nor in the sociocultural-interactionist direction,
people would ban the ideas of the others. I think we can learn a lot from each
other, and I think a serious theory of socially and culturally situated discourse
and interaction needs both.

7.4 November 8, 2004: from Duranti to van Dijk


There are several themes, and several claims, embedded in your message. Let
me try to list them (so that you can also see what I read into your writing and
correct me if I misinterpreted something):
1. There should be BOTH cognitive analysis and “the other kind,” that is, the
social, cultural, interactional . . . (the fact that one can characterize the first
type, “cognitive,” with one word and one instead needs several terms to
characterize the other kind is something interesting in its own right).
2. We should figure out (ideally agree on?) the empirical bases for talking
about intentions and more generally for any analysis of meaning (what you
call at one point the question of “observability”). You question the wisdom
of privileging one type of data, e.g., what Others think, over other types,
e.g., what we, analysts-westerners, think, or privileging production over
reception, context over cognitive processes. You also said that even though
the audience may have its own take on what someone says/does, you don’t
see why privilege it. More specifically the “audience as co-author” idea, in
your view, has little to do with intentions.
3. There is no proof that other people don’t have intentions and they don’t
have ideas about “other minds.”
November 8: from Duranti to van Dijk 159

4. Intentions are representations. Representations (in turn) influence intentions


and hence actions.
5. No matter what we call these things (intentions, meanings, interpretations,
representations, constructs, etc.), they are all “personal and socially shared
mental models of some kind.”
Let me try to quickly go through them one by one.
1. I know that politeness should encourage me to say “Of course” we need
both “cognitive analysis” and “the other kind.” But I am afraid that I would
be cheating if I said it without at least some qualifications. The point is not
that I don’t think people should be doing cognitive analysis if that’s what
they want to do and are good at. It would be like saying that people
shouldn’t be doing statistics simply because usually I and my students
don’t. There are things that people can “see” with statistics that they can’t
(easily) see otherwise, e.g., gender discrimination in the work place. But
I would need to be shown that there are phenomena that can either ONLY
be seen with “cognitive analysis” or they can be seen BETTER with such
an analysis. In other words, I would not want to start out with the idea that
everything is equally amenable to cognitive or social-cultural-interactional
analysis. For one thing, I have a hard time with the distinction period. To
the extent to which I understand what people mean by “cognition,” I see it
as something very cultural, social, interactional. Take problem-solving.
I find the analysis by people like Ed Hutchins (1995) and Michael Cole and
his associates at the Laboratory of Comparative Human Cognition
(LCHC) very convincing. People come to an understanding of a problem
(e.g., how to navigate safely and efficiently, how to solve math equations,
how to cook a meal) by relying heavily on interactions with other social
actors and with artifacts. Yes, there are things happening in their minds,
I am sure, but when we look at the whole activity, what we call “problem
solving,” it seems very ad hoc to say that we can just talk about it as
something that happens in the mind or that the individual mind has
everything there is to have to problem-solve. Now, you might agree with
this, but not everyone does. I would want to suspend judgment on what
cognition is and “look” at what happens in order to provide a first descrip-
tion of the phenomena in need of an explanation before starting to build
mental models, whose ontological status is often unclear or at least under-
analyzed. Again, this is not a general dismissal of mental models. I think
they can be very interesting tools. But they should not be seen as having
privileged status. In this I am probably close to the position that you have
with respect to the issue of the audience interpretation, when you say that
you don’t think that it should have privileged status – by the way, my point
about the audience “counting” was meant to be based not on some a priori
160 A dialogue on intentions

principle but on the observation that in specific circumstances the audience’s


interpretation is an important factor in controlling people’s understanding of
what goes on (but I’ll get to the audience issue below). I also want to mention
that the dichotomy between “cognitive” and “social-cultural-interactional” is
problematic for me because I think that there are important considerations
that are left out of this dichotomy. For example, it is not clear to me where to
fit morality and aesthetics. And yet, any human act is potentially (and often
de facto) subject to moral and aesthetic canons and evaluations (I tried to
raise this, briefly, in my agency in language paper; Duranti 2004). That
should be obvious to anyone doing work on development or socialization
(see Piaget’s work). But is it obvious to cognitive scientists? It’s not even
obvious to some of the interactionists I hang out with. So to me the solution,
“a” solution, is not to say that we need “both” but that we need as many tools
as we can to make sense of what people do. I am for a COMPLEMENTAR-
ITY APPROACH, which is different from an eclectic approach. I do believe
that different analytical tools, like different artifacts, allow you to see, hear,
feel, know, etc. different aspects of what we call (imperfectly) cognition,
social interaction, cultural practices, etc. I find it crucial to be able to say
exactly what it is that a given set of analytical techniques can give us that is
peculiar, specific, and, yes, illuminating. Let me also add that it could very
well be the case that for you “cognitive” refers to something very specific that
is NOT covered by what other people do using other labels and, more
importantly, analytic techniques. If this were the case, I would start from
listing what those things are. I feel pretty confident that I know at least some
of the things that we miss if we leave out of our accounts of human activities
their understanding in terms of people coming together and having to take
each other’s actions into consideration. I would want the same kind of
revelation from any “cognitive” account, which I am sure you are quite
equipped to do.
2. Empirical bases. This is a huge issue. I’m glad you brought it up, but it ain’t
easy to address without a couple of conferences and lots of help from
others. Let me mention that I see the question of how to get data (or rather
“good” data) as related to the question: how do we know we are talking
about the same thing? Because, as you and I know, there’s no methodology
for getting “data” without some basic theoretical assumptions on what
could constitute such an entity. Let me try to expand/comment on a couple
of points you made.
You said that the fact that in some cultures (by which I think you mean
communities) members do not like to talk about intentions or internal states
is no proof that intentions do not exist or are not relevant for them. I agree
with this statement if we mean that we can’t build a theory of human action
and its interpretation based on what Boas called “secondary explanations,”
November 8: from Duranti to van Dijk 161

i.e., members’ rationalizations about their own doings (and I would insist
that we should apply this measure to some of the explanations used in the
literature on intentionality). I would even go a step further and say that just
like it is in certain members’ interest to have an intentionalist theory of
meaning, it is in certain members’ interest to have an anti-intentionalist
view of interpretation or an anti-personalist view of meaning. Such a view
allows people to ignore what some other people (e.g., children, servants,
lower-ranking individuals, outsiders) think, feel, or (in our vernacular
sense) “intend.” Highly hierarchical systems, like the military in the US,
seem to be this way and so might be the case with some stratified societies
studied by anthropologists. One of the ways in which power relations
display themselves and become active is precisely by reclaiming or ignor-
ing the will, desires, and feelings of others – and these are all terms that
somehow figure or are relevant to some version of what cognitive scientists
seem to refer to when they use the term “intentionality” (the comparative
perspective is complicated by the fact that we seem to also have examples
of egalitarian anti-intentionalist communities, e.g., the Kaluli of Mt. Bosavi,
in PNG, studied by Bambi Schieffelin). Finally, it is true that in the western
philosophical tradition, we have different perspectives as well, Plato vs.
Aristotle, . . . Husserl vs. Heidegger, Carnap vs. Wittgenstein, etc. And
authors change over time as well or rather they bring out hidden dimen-
sions, see Husserl’s Logical Investigation vs. his Cartesian Meditations or
the Crisis of the European Sciences.
One could say that one should be careful about how to use ethno-
graphic accounts because most anthropologists tend to favor the native’s
point of view and the natives can be wrong. Granted. But there is another
side of the native’s point of view, namely, that other people’s perspective
on something that we take for granted can make us rethink about our own
conceptualization and understanding of our reality (that’s the old idea that
anthropologists make the strange familiar and the familiar strange). It’s
not that the Other is always right and We are always wrong or that they
are better than we are at knowing themselves and what goes on in their
mind(s) – this would be some kind of romantic or politically correct version
of cultural anthropology (and actually a very understandable reaction-
antidote to the superiority complex of some members of our respective
societies). Rather, the ethnographic perspective (a form of cultural relativ-
ism, but of the best kind in my view) is that we should stop assuming tout
court that Our view is the objective, universal view and Others’ view is the
particular, native, emic view. We have our own reasons to believe what we
believe and our own problems clearing our vision (e.g., we still haven’t
figured out why acupuncture works). For one thing, people around the
world are not THAT DIFFERENT. It really depends on how we talk about
162 A dialogue on intentions

it and what we are looking for. We can find lots of similarities and therefore
we can also find evidence of the fact that intentional states of course do
matter even for the people who do not seem to care much about them or do
not even have a language to talk about them. People around the world seem
to have a term for what we call “lie,” which is a form of recognition of the
fact that someone is saying something different from what they believe-
know (although the set of situations that would define “lying” vary cross
culturally and cross-contextually).
3. This might be the right time for me to say that I never said that “intentions
do not exist” or that “they never matter.” I might have said, though, that (i)
I am not sure what “intentions” are, by which I mean that I am not always
sure what people mean when they use the term and a clarification would
help a lot in the discussion, (ii) there are situations in which intentions
matter less (e.g., for sure in certain institutional settings), and (iii) people
around the world vary in terms of their concern for reconstructing the
intentions of a speaker-actor; and (iv) there are, at least in some cases,
alternative accounts to the intentionalist account of a given action and
therefore, (v) pace Dennett, the intentionalist stance might not be the best
solution under all circumstances.
4. I’m afraid I do not agree with the idea that intentions are representations.
Maybe we do not agree with the meaning of representations, or I just don’t
understand what you mean, in which case you’ll have a chance to teach me,
but I think of representations as things that stand for something else. For
example, I can have a material representation of a room, e.g., a map, or
I can have a mental representation of a room (I can remember some of its
features from memory or from imagining something that does not exist in
reality – I could even visualize it) – this type of representation has a lot to
do with temporality as described by Husserl. Language represents in very
imperfect and yet quite efficient ways through symbols and icons. I can
represent through language by providing descriptions or by symbolically
embodying my feelings, attitudes, sensations, etc. through particular lin-
guistic acts, often relying on other media and conventional resources (e.g.,
gestures, material artifacts, etc.). But I can’t quite think of intentions as
anything remotely similar to these cases. Intentions would be something
closer to an experience or to an act. I can see that a cognitive perspective
would want to try to “represent” intentions (which is different from saying
that they ARE representations), although, in some cases, it would be quite
difficult to do, except in some simple way, e.g. creating abstract predicates
that use ordinary language but are meant to be not-language-specific, e.g.,
the primitive notions of “INTEND” or “BELIEVE” or “ENVY.”
5. Mental models. I touched a bit on this notion earlier. Models can be fun to
play with. And I agree that there are personal-biographical aspects of
November 9: from van Dijk to Duranti 163

interpretation as well as social-shared aspects. But again, I would find it


difficult to accept that “intentions, meanings, interpretations, representa-
tions, constructs” are all “models.” I can see that a representation can be a
model, but I am not sure how an intention is a model, except for the fact that
anything can be made into a model by making it into something with
generative power, in which case we lose the specificity of the concept of
intention, which is one more thing to think about.

7.5 November 9, 2004: from van Dijk to Duranti


I am glad that some of my comments are helpful to represent also the more
cognitive approaches to discourse or those who have the ambition (like me) to
combine cognitive, interactional, and societal aspects of text and talk. I am sure
you already knew all the things I have said (and also noted that we agree on
most issues, even when we place different accents and may have different
preferences in our research). In any case, I am learning and have learned a lot
from you and your recent books (e.g., A Companion to Linguistic
Anthropology).
Also our debate is very useful for me to better understand another perspec-
tive on discourse (I have always believed that of all disciplines anthropology is
ideally placed to combine the best of the other disciplines and approaches in
the study of discourse; it is not surprising that the modern study of discourse
originated there: Propp, Hymes, Gumperz, etc.). This is also essential for me to
be a better editor of the journals, in which I want to understand and welcome
approaches from different perspectives, and promote mutual integration and
influence.
[In what follows] I shall try to be as concise as possible, by enumerating
briefly the points I agree with, and then reformulating some of the points where
I do seem to have a somewhat different (or complementary) view.
1. Most importantly, I agree with you that “interactional” and “cognitive”
approaches are or should be complementary. I regret any kind of reduc-
tionism in either direction, and am fervently in favor of integration: this is
better for cognitive as well as interactional approaches, since in my
opinion the one cannot really be understood without the other – even
when one can go a long way just studying one of these fundamental
dimensions of discourse or language. In sum, I think that all psychology
should (also) be social psychology, and all social science should also be
cognitive science.
2. Yes, we should not privilege any kind of data (“ours” or that of the
“Others”) – even when we obviously know “ours” best: which may also
be a handicap because it may lead to bias of extending our knowledge as
being universal.
164 A dialogue on intentions

3. True: Doing cognitive or interactional analysis depends on the kind of


aims or data one has. Below I shall mention some examples of issues that
in my view cannot be handled without a “cognitive” approach (one
of many).
4. Yes: there are cognitive issues (such as your example from Michael Cole’s
lab – I was at UCSD on sabbatical in 1985, so know them well) that
should be studied in an interactional framework (as there are interactional
studies that require a cognitive perspective, see below).
5. Yes, in my view a (broad) cognitive approach should include the dimen-
sions you raise of “appraisal” – moral or esthetic. I guess there are people
in cognitive science doing this, but this is not my area at all. My closest
topic is that of ideology, which also involves evaluation, norms, values,
etc. The same is true, incidentally, for questions of emotion, which is now
a very popular topic in both cognitive and social psychology.
6. Yes, in general, for what you say about “intentionalist” vs. “anti-
intentionalist” communities, scholars, or perspectives, but doubtful
whether the notion of “intentionalist” is correct here. I guess we need a
more general term (you also mention feelings etc.), in terms of whether or
not we attribute any kind of “thought” or “feeling” to others, or whether
we describe and explain our own actions in any kind of “mental” or rather
in social terms. I guess that I would tend to be in favor of one of your
conclusions, namely, that besides obvious cultural differences, people are
not that different. I would add that I think that the differences might be
socially more pronounced than cognitively: I doubt whether the basic
principles of cognitive processing involved in engaging in discourse and
other everyday actions is fundamentally different between cultural com-
munities. For instance, our knowledge and beliefs (their contents) may be
very different, but not the way such knowledge is organized and used in
discourse and interaction.
7. True, intentions are only one aspect of action and it very much depends on
the kind of actions or collectivities studied whether it makes much sense to
talk about them. Obviously, it makes less sense to talk about intentions
when speaking about collective acts of groups, organizations or states.
I have no doubt at all, though, that for the description of the (inter)actions
of individual actors, and hence for language users, intentions (or whatever
we want to call them) need to be at least part of the theory of action – as is
the case for related notions such as aims, plans, or goals, whether or not
one wants to describe them in a particular project. One may want to
describe intentions in a more abstract sense (as we do with “meanings”
in linguistics and the humanities and social sciences more generally), or in
terms of their inherent nature as properties of the mind, as forms of
thought. Of course, for radical antimentalists, this is no solution, but
November 9: from van Dijk to Duranti 165

I have no reply to those who deny the existence of “minds,” and hence do
not talk about such things as knowledge, beliefs, wishes, thoughts, etc. as
mental phenomena. I would only be able to say that I am sure that radical
antimentalism is inconsistent and contrary to compelling human experi-
ences (thought, feelings, beliefs) and to massive evidence of many kinds.
I don’t think it is very fruitful to discuss such a position – which in my
view is as silly as denying “society,” e.g., by radical mentalists who would
say that all social things are mental constructs (which in a way they are)
and nothing else. In sum, the question is, as you say: It depends on the
issue or problem or goal of research whether a more cognitive or more
interactional (or more societal) approach – or a combination, is more
interesting or fruitful. For the study of, e.g., racism or sexism one obvi-
ously needs all dimensions. For a study of strategies of self-presentation,
one would need a cognitive and interactional perspective, and for a study
of the power of the news media, one would rather take a societal and less a
cognitive and interactional perspective, even when locally news is pro-
duced by (interacting) journalists, and although the power of the news
media also depends on the knowledge, beliefs, prejudices, and ideologies
they help (re)produce in society.
8. Yes, there are different kinds of models: personal and social ones, more or
less abstract or more concrete (analog etc.) ones etc. I use the notion only
as it is used in current cognitive psychology, namely as subjective repre-
sentations of specific events (acts, situations) in episodic memory, i.e., as
some kind of experiences – involving both personal and social aspects but
as unique (and hence changing, dynamic) “interpretations” of events. One
never forms exactly the same mental model of an event, and no two
persons ever have exactly the same model of an event: their experiences
always will also depend on unique life histories.
9. Although I use the general label “cognitive,” obviously that description
covers a huge domain (that of the cognitive sciences), which is as
diverse as, if not even more complicated than, that of interaction, and
involves such diverse areas of study as cognitive processing of lan-
guage and discourse, the study of knowledge, beliefs and ideologies,
memory, thought, problem solving, the study of metaphor, and so on –
and many (sub)disciplines, such as psycholinguistics, AI, cognitive and
social psychology, hermeneutics, epistemology, etc. etc. So, yes, I
would know more or less what I would be examining or leaving out
when doing a “cognitive” analysis – and obviously cannot do every-
thing at the same time. My book on ideology (van Dijk 1998) (and the
new one on context [van Dijk 2008]) are examples – although only a
first step: in the future to be followed by more detailed (also empirical)
studies.
166 A dialogue on intentions

10. Intentions are “representations” because they subjectively represent our


“doings” as human actors. Intention þ doing ¼ action, in the same way as
meaning þ utterance ¼ discourse. Such a representation is subjective, but
under social control: we may “mean” many things by doing X, but we
know also that whatever we may “mean” with doing X, it will probably be
understood as U(X). When we “understand” an action we do so by
attributing a meaning to a doing (activity, conduct), and this usually means
attributing an intention to an author. Hence the standard question: “What
do you MEAN?” (when interpreting discourse or actions), or the standard
excuse “I did not MEAN it.” Such representations are formed in short-
term memory and stored in episodic memory, and as such constitute our
(personal) experiences, and our autobiography. One may intend to do X,
but never be able to actually execute such an intention. Subjective repre-
sentations of events (and doings) in episodic memory are mental models,
which are constructed by (i) previous mental models (experiences), (ii)
general, shared social or cultural knowledge and beliefs, as well as (iii) the
subjective representation (¼ model) of the current situation, i.e., the
context.
11. Finally, what uses of a “cognitive” approach to discourse are (more) useful
than an “interactional” approach? First of all, I would say that ideally they
should always go together, because any approach to discourse has both
cognitive and interactional dimensions. Also, I prefer to speak of a socio-
cognitive approach, so as to emphasize that cognitions always also have a
social (cultural, etc.) basis. For some questions/issues/problems a socio-
cognitive approach is indispensable in my view, such as (several of which
I base on my own experiences and research):
a. obviously, but not quite trivially, for all studies of “mental” processes
of discourse production and comprehension;
b. for all studies of language learning and the acquisition of various kinds
of discourse “competence”;
c. in order to account for local and global coherence of text and talk,
based on shared knowledge;
d. for the account of semantic and pragmatic meaning and interpretation
in general. . .
e. . . . and for the account of implicit meanings and implicatures in
particular;
f. for the general account of the role of knowledge in discourse (and the
study of the reproduction of knowledge and other beliefs) through
discourse;
g. idem, for the study of discourse and ideology – including for instance
the study of prejudice and its expression and reproduction through
discourse;
November 16: from Throop to Duranti (and van Dijk) 167

h. for the study of the role of context (and various context factors) defined
as subjective interpretation (mental models) of relevant aspects of the
social situation, and hence as interface between situation and societal
structures on the one hand, and discourse structures on the other. In
other words: there is no DIRECT link between situation/society and
structures of text or talk without the intermediating role of subjective
definitions/interpretations (models) of communicative situations. Thus,
also social (macro) structure is only relevant for discourse when it is
“made” relevant by participants, that is, through their subjective, but
socially constrained, interpretation. In other words, also the theory of
“relevance” has such a sociocognitive basis;
i. to explain (and not only to describe) interaction, in the sense that
participants are only able to act relevantly (coherently, etc.) with
respect to previous acts of others if they understand these acts, i.e.
represent them subjectively, and use this subjective representation as
one of the conditions for their own current, ongoing action – besides
general knowledge about norms, etc., actions and their possible
consequences.
This is only a selection of possible topics. It can be extended at will. I believe
that none of these topics can be “better” formulated or understood in inter-
actional terms, although many of them also have interactional dimensions.
E.g., contexts (as subjective interpretations of ongoing communicative situ-
ations) are produced, negotiated, etc. by interaction.
Of all these topics and issues, I am personally most interested in those at the
interface of discourse, cognition, and society (involving both micro – inter-
actional – as well as macro dimensions), such as the (re)production of racism,
the role of knowledge and ideology in discourse (and vice versa), etc.
[Duranti forwarded the above exchange with van Dijk to a number of
graduate students and postdoctoral fellows in the Department of Anthropology
at UCLA. Jason Throop responded with the following message.]

7.6 November 16, 2004: from Throop to Duranti (and van Dijk)
Personally, I have always felt that the concept of intentionality should be
rigorously restricted to its most basic Husserlian sense – the directedness of
a constituting consciousness toward its objects so constituted. Otherwise, as
you have so perceptively noted in both your Cultural Anthropology article
[Duranti 1993a and see Chapter 5 of this book] and again here in the context of
this dialogue, what individual thinkers mean by the term is decidedly unclear,
even at the best of times. Are individuals referring to plans, goals, desires,
motives, acts of willing, agency? Are they referring to a stance toward the
168 A dialogue on intentions

possibility of transparently accessing such phenomena in others? Are they


referring generally to theory of mind, i.e., that regardless of the relative opacity
of “intentions” and whether or not such phenomena are indeed intersubjec-
tively accessible, that we believe/intuit that others in some way still “have”
them? Are they referring to questions concerning an individual’s ability to
translate plans, goals, desires, etc. into effective action? Are they referring to
explicitly conscious phenomena or can there be unconscious motivations/
intentions (à la Freud)? Or are they trying to highlight their allegiance to a
particular theoretical orientation to social action (i.e., mentalist vs. anti-
mentalist)? Etc.
That said, if we must stick to a more general rendering of “intentionality”
I cannot help but wonder how temporality fits into the picture. Schutz, in his
critique of Weber’s theory of meaningful social action, raises a number of
interesting issues here (I think). In his Phenomenology of the Social World,
Schutz challenges Weber’s a-temporal assumption that motives can be simply
equated with what can be construed to be an adequate explanation brought to
bear by either actors or observers with regard to a given sequence of conduct.
He asserts, as I highlight in “In the Midst of Action” (Throop 2010b) that
such a perspective hypostatizes human motives (or, as some would prefer,
“intentions”) as reflections of, or orientations toward, an already completed
act, while failing to explore how motives are configured in the context of
ongoing activity. As he notes, this a-temporal perspective fails to take into
consideration distinctions between “in order to” and “because” motives; the
former being motives that are articulated as directed toward a future antici-
pated goal, and the latter being those that are organized around present
behavior that is understood to arise from past experiences. And what is
interesting about this distinction between “because” and “in order to” motives
is the fact that Schutz highlights the ever-present potential for differences in
social actors’ temporally configured orientations to the motives informing
what would otherwise appear, on the pure behavioral level, to reflect similar
sorts of activity (i.e., the difference between “the act of chopping wood
because . . .” and “the act of chopping wood in order to . . .”). (I should note
that “in order to” motives overlap, at least in part, with the latter of what
Dr. van Dijk explained to be a distinction in the philosophy of action between
“intentions” as representations of the act itself and “purpose” as representa-
tions of the state-of-affairs that are meant to be brought about through the
activity. Also, I would add, that Weber’s perspective also importantly fails to
recognize the representation as the “act itself” in its unfolding, which in
drawing from Ed Casey, I argue in “In the Midst of Action” (Throop 2010b)
is based on individual actors’ abilities to not only “imagine-that” a given
completed action is embedded in a certain state-of-affairs, but also to
“imagine-how” a given act unfolds in time.
November 16: from Throop to Duranti (and van Dijk) 169

Turning to Husserl, Schutz distinguishes Weberian “motives” from the


notion of “intentionality” which is held to be a synthetic activity that is
predicated upon a sedimented series of constitutional act phases. In Schutz’s
perspective, depending on how we orient to intentional act phases and their
corresponding intentional objects, we are thus equally able to engage motives
as already constituted, objectified, justifications for action or as emergent from
the constitutive intentional acts that (in part) structure the dynamic flow of
subjective experience (which is confronting an equally dynamic and consti-
tutive field of social interaction). As he puts it,
On the one hand, I can look upon the world presenting itself to me as one that is
completed, constituted, and to be taken for granted. When I do this, I leave out of my
awareness the intentional operations of my consciousness within which their meanings
have already been constituted . . . On the other hand, I can turn my glance toward the
intentional operations of my consciousness which originally conferred the meanings.
Then I no longer have before me a complete and constituted world but one which only
now is being constituted and which is ever being constituted anew in the stream of my
enduring Ego: not a world of being, but a world that is at every moment one of
becoming and passing away – or better, an emerging world. (Schutz 1967: 35–36)

(Here, Schutz is drawing explicitly from Husserl’s insights into the phenom-
enology of internal-time consciousness and the phenomenological reduction.)
Advancing a distinction between “objective” and “subjective” meaning, he
asserts that whereas the former (which he claims Weberian interpretive soci-
ology is restricted to investigating) is concerned with attending to the phenom-
ena of the external social world as they are presented to an actor as pre-given
“indications of the consciousness of other people,” the latter is predicated upon
an orientation to those processes through which the constitution of these
“external” indications of the consciousness of another being are dynamically
achieved. It is important to note the difference here between what may be
prevalent commonsense renderings of “intended meaning” and Schutz’s per-
spective, however, since for him “intended meaning” is necessarily subjective
meaning, but not necessarily overtly consciousness meaning. As he puts it,
[W]hen we speak of subjective meaning in the social world, we are referring to the
constituting process in the consciousness of the person who produced that which is
objectively meaningful. We are therefore referring to his “intended meaning,” whether
he himself is aware of these constituting processes or not. (Schutz 1967: 37)

Significantly, in distinguishing between meaning-contents that are already


constituted and those that are in the process of constitution, Schutz points
out that the “meaning of my action consists not only in the experiences of
consciousness I have while the action is in progress but also in those future
experiences which are my intended action and in those past experiences which
are my completed action” (1967: 39). It seems that in light of Schutz’s insights
170 A dialogue on intentions

that we should thus recognize an ever-present complexity entailed in even the


most habituated of everyday activities, wherein which orientations to ongoing,
elapsed, or anticipated phases of activity may differentially implicate present,
recollected, or imagined audiences whose reactions (whether actual or
imagined) play a role in configuring an actor’s course of action and the
imputation of purpose (and thus accountability) to specific acts – and of
course, actors often work to mitigate or highlight their accountability in re-
casting elapsed acts as “intended” or “unintended” depending on the reaction
of an audience to those acts (an insight that is, as you note, occluded when the
possibilities for an audience’s role in co-authoring the “intentional” basis of a
given act is not adequately taken into account).
Of particular interest to me in the context of this more general usage of the
concept of “intentions,” however, concerns how what constitutes the “context”
or “situation” toward which individual “intentions” are arrayed may vary for
different participants, who may otherwise be working to coordinate their
actions in light of putatively shared interpretive frames. To this end, while
I agree that it is certainly a crucial insight to recognize that “the audience’s
interpretation is an important factor in controlling people’s understanding of
what goes on,” thus problematizing any simplistic notion of intentions as
confined to an individual actor’s stream of experience, the audience to which
any given actor may be orienting in the context of the ongoing flow of activity
may also be an imagined future audience, or recollected past audience, whose
imagined or recollected responses may or may not align with the reactions of a
given social actor’s present interlocutors. Last year, during one of the Sloan
workshops, Doug Hollan mentioned the idea that just as we must be attentive
to the dynamics of the social context within which social actors are enmeshed,
it is also important to consider the dynamics of the subjective contexts within
which social configurations take on particularized articulations (here I recall
him specifically using the phrase “thinking of the person as context”). I think
that when we take into account the fact that the texture of subjective experience
is temporally structured, and that the articulation of this structuration in terms
of past, present, or future orientations may differ for particular actors at
differing points in the flow of ongoing interaction with their interlocutors, that
it can become at times quite challenging to emplace social actors within a
shared interactional field, even when they are engaged in what might otherwise
seem to be co-constructed, present focused forms of mutual-attunement
through talk and interaction. All of which, I think, raises some interesting
questions regarding how it is that we should go about determining the contexts
from which, or toward which, “intentions” arise.
I am not sure exactly where this gets us, except that it sheds light, in my
mind at least, on the importance of considering how inter-subjective and intra-
subjective processes, which are both temporally stratified and configured, may
November 16: from van Dijk to Throop and Duranti 171

align or dis-align in very complex ways, and how all of this raises interesting
questions as to the possible variety of ways in which interlocutors may
highlight, impute, background, disavow, ignore, and/or contest the “inten-
tional” nature of their own and others’ experienced activity.
And of course, all of what I have said so far has not (at least explicitly)
touched upon questions of privileging particular methodological or analytic
approaches, concerns relating to the degree of methodological privacy that is
associated with such states/processes, issues concerning whether or not inten-
tions can be considered “representations,” problems tied to reconciling actors’
vs. (scientific) observers’ rationalizations for explaining action, ideas concern-
ing the ways in which other “non-intentionally” based explanatory frames can
be brought to bear in interpreting social action, or issues of power relations and
how they may serve to pattern attentiveness to, or the ready display of, a
particular actor’s motives/intentions.

7.7 November 16, 2004: from van Dijk to Throop and Duranti
Since “intentions” and similar notions are obviously too complex, and yes, too
vague, to be satisfactorily dealt with in an informal email message, let me just
briefly comment with a few points:
1. I can only agree that the notion of “intention” (from Husserl to Searle, via
Schütz, and others) is very ambiguous. In order to be useful in a more or
less explicit theory of discourse, we must, indeed, severely restrict it. The
same is true for such vague general terms as “meaning” or “understand-
ing,” among others.
2. Yes, intentions, however defined, are part of a complex of related notions,
such as purposes, motives, plans, aims, objectives, targets, and, indeed,
meanings, all of them related to action or agency.
3. Yes, this relation, indeed, is – among others – one of temporality: inten-
tions etc. are usually conceived of as “preceding” acts. More specifically,
they may be construed as “causes of,” or more appropriately, “reasons” for
action, in the sense that they make acts “rational,” “reasonable,” and hence
understandable.
4. Among the many analytical, theoretical distinctions that need to be made,
however, is that between actions as a combination of intention þ doing
(behavior, conduct), and all the temporarily and rationally “preparing”
notions such as motives, purposes, aims, etc. that are invoked (before,
during, or after the action) by agents or attributed by co-agents or obser-
vers, to explain, justify, or understand actions.
5. This means that one can in principle engage in actions by intentionally
doing something (moving a pencil etc.) but without any further purpose,
172 A dialogue on intentions

motive, etc. In real life, however, most actions have a complex and fuzzy
set of more or less conscious aims, purposes, goals, etc. Sometimes, and
informally, these various “conditions” of action are also described as
“meanings” or “reasons,” in the sense of what we mean BY an action,
what we want to accomplish or “get done” BY (as a consequence of) an
action. For these latter notions we may need a host of further descriptions
and analyses, ranging from cognitive psychology and psychoanalysis to
shared sociocultural beliefs of various kinds. There is nothing mysterious
about them – they only need careful, systematic analysis in terms of some
theory.
6. Intentions, however, do not in this way “explain” etc. actions, but merely
distinguish actions from unconscious etc. doings or bodily events. In other
words, they are the conscious basis or counterpart that constitutes human
action, as distinct from, e.g., conditioned reflexes, or activities engaged in
under drugs, hypnosis, or force. Intentions are thus essentially definable in
terms of self-control over activities.
7. This also has moral, cultural, and legal consequences at least in many
communities: If I misbehave without control over my conduct (when
forced, drugged, sleeping, etc.) I usually am not found guilty. This is not
true for the other notions (motives, goals, etc.): one may not be guilty of
“premeditated” murder, but at least of manslaughter. Hence, all the other
notions are forms of “pre-meditation.” Intentions on the other hand frac-
tionally precede but further practically co-occur with an act. They control
the process of being engaged in activity.
8. Intentions in this sense are also described as the agent’s “meaning” of the
doing itself – and not of the whole action and its consequences. Like the
speaker-meaning of a word. Hence the usual observer’s question when an
act is ambiguous: “What do you mean?,” which may be glossed as “What
was your intention when engaging in this conduct?”
9. All this is mere analytic philosophy of action, but hardly empirical theory.
Since we are dealing with mental objects of some kind, we had therefore
better devise a proper cognitive theory for these states and processes. In
such a theory, intentions are (part of) mental models, represented in
episodic memory, and hence part of our everyday experiences. These
mental models control discourse and other actions and provide the
speaker’s “meaning” for them. Others may interpret our doings (as our
word-sounds) and attribute meanings to them by constructing their own
mental models. (See van Dijk and Kintsch, Strategies of Discourse Com-
prehension, 1983, and a host of later work in cognitive psychology for
details.)
10. Obviously mental models are not just constituted by current experiences
but also by previous ones, as well as by general, socially or culturally
Conclusions 173

shared beliefs. This is why and how we are able to understand words and
acts in terms of similar mental models. However, as such, mental models
are contextually unique, subjective, etc. They are tokens, not types. We
can never have the same intention twice. NB. These models are NOT the
same as the cultural models of Shore etc. – which are socially shared, etc.
There is much more, but (at this late hour;-)) I don’t have much more to say.
Some of this is consistent with what Jason says. Some of this tries to go
beyond the philosophical confusion of a host of related notions, by devising a
more explicit, more analytical theory of it. Whatever the exact nature of
intentions as defined, I have no doubt that an explicit theory of action and
agency needs at least a notion like this – a notion that not only should account
for pretty strong intuitions, but also for a range of phenomena and data in
different cultures as referred to above (reasons vs. meanings of action, being
guilty or not guilty of some doings, etc.). That not all cultures have a notion
like “intention” does not as such warrant that we should not have it in a theory,
as is also true for a theory of atoms or molecules as components of matter.
So what do we need to advance?
a. Comparative cultural studies of concepts and uses of words such as inten-
tions, goals, etc.;
b. systematic analysis of spontaneous discourse using these notions in differ-
ent communities;
c. more theoretical analysis distinguishing these various forms of “premedi-
tation” of action;
d. a detailed theory of the structures of mental models for action;
e. a theory of the relationship between mental models for action and actual
conduct (how to translate mental intentions into the complex “bodywork”
of doings);
f. experiments to test this theory;
g. brain observations to find out the neurobiological basis of intentions and
related notions, etc.
Lots of work. At the moment I only take one of these options: devising a
theory of context as mental models that will do some of the control of
discourse/interaction, and hence needs intentions and “premeditations” to
make discursive conduct rational or reasonable or meaningful.

7.8 Conclusions
The discussion went on for a few more days, but the contributions I have
reproduced above constitute the exchanges that more closely and consistently
focus on some of the core issues of this book. In fact, the ideas and arguments
174 A dialogue on intentions

found in the sections above motivated me to go back to the topic of intention-


ality after a hiatus of almost fifteen years. The brief article that I later published
in a special issue of Discourse Studies edited by van Dijk (Duranti 2006b) was
the beginning of a new exploration of the Samoan materials as well as a return
to Husserl’s writings, which have continued to inspire me ever since. Almost at
the same time cultural anthropologists had returned to the issue of intentional-
ity within the context of the debate about “theory of mind.” The next chapter
presents my reflections on this new wave of contributions about what people
can know or think they can know about what others really think, feel, or mean.
8 Opacity of other minds: local theories revisited

8.0 Introduction1
Starting in the late 1980s, a number of developmental psychologists and
philosophers (e.g., Astington, Harris, and Olson 1988; Carruthers and Smith
1996) recast the issue of the role of intentions in interpreting people’s actions,
words included, in terms of the so-called “theory of mind,” a term originally
introduced by David Premack and Guy Woodruff (1978) in their work on
primates’ cognition. In the first decade of the twenty-first century, linguistic
and cultural anthropologists entered this tradition of studies not so much to
claim cross-cultural differences in the capacity to read other minds, which they
took to be a human universal, but in the need and willingness that members of
different communities express and demonstrate to engage in introspection or in
guessing what others might be thinking, wishing, or feeling. As shown in the
following quote by Joel Robbins and Alan Rumsey, the doubts that Michelle
Rosaldo and I, and others, had raised in the 1980s about whether we need to
involve individual intentions to interpret human action were reframed as issues
of ideological control.
While it may seem at first blush as if the work on intentional communication and
“theory of mind” that we have referred to in the last two paragraphs is completely at
odds with the anthropological critique of personalist/intentionalist theories of meaning,
this is not necessarily the case. For even if it is true that the capacity for inferring the
mental states of others is a generically human one, and plays a part in communication
everywhere, it does not follow that all language ideologies will give it equal promin-
ence, or even allow it to be openly recognized or actualized in speech. (Robbins and
Rumsey 2008: 414)

Starting from the observation that, according to existing ethnographic


accounts, people in a number of societies in the Pacific have been said to
claim – or to imply through their behavior – that it is impossible to know what
goes on in another person’s mind, Robbins and Rumsey (2008) put forward the
hypothesis that members of these societies subscribed to what they called
“doctrine of the opacity of other minds” and that such a doctrine “matter[s]
to how people operate socially” (Robbins and Rumsey 2008: 414). Evidence

175
176 Opacity of other minds

of this local “doctrine” included conversations with native consultants who


refused to provide motivations for their own actions or resisted any kind of
intentional reading of what they had just done.
In this chapter I review some potential problems that emerge in the data used
in this line of argument and provide some examples that suggest that, while a
certain degree of “mind-reading” is at work in Samoa like everywhere else in
the world, whether or not we treat these examples as evidence of people using
intentions to infer the meaning of a person’s actions is debatable.

8.1 Levels of argumentation


A careful reading of the literature shows that the opacity claim has been dealt
with on two main levels, each of which relies on a different order of general-
izations and implies different kinds of analytical categories and types of
argumentation. The first level – which we might call “observational” – draws
from fieldwork situations in which ethnographers try to get their consultants,
field assistants, hosts, and acquaintances to tell them about the intentions,
motivations, or emotions that might be used to explain other members’ actions
or their own. These attempts, as we know from the ethnographic literature (see
Keane 2008; Robbins 2008; Schieffelin 2008; Stasch 2008), are sometimes
met with surprise, suspicion, or flat rejection by local consultants, as in “How
would I know?” or “Why should I know?” An early account of this stance is
provided by Margaret Mead in her Coming of Age in Samoa, where, after
introducing the concept of musu (see §3.1 in this book) she provides further
evidence of her claim about Samoans’ “odd incuriousness about motives” by
citing the ways in which people reacted to her questions about motivation.
This lack of curiosity about motivation is furthered by the conventional acceptance of a
completely ambiguous answer to any personal question. The most characteristic reply to
any question about one’s motivation is Ta ilo, “search me,” sometimes made more
specific by the addition of “I don’t know.” This is considered to be an adequate and
acceptable answer in ordinary conversation although its slight curtness bars it out from
ceremonial occasions. So deep seated is the habit of using this disclaimer that I had to
put a taboo upon its use by the children in order to get the simplest question answered
directly. When this ambiguous rejoinder is combined with a statement that one is musu,
the result is the final unrevealing statement, “Search me, why, I don’t want to, that’s
all.” Plans will be abandoned, children refuse to live at home, marriages broken off.
Village gossip is interested in the fact but shrugs its shoulders before the motives.
(Mead 1928: 123)2
I would like to suggest that an analytical shift takes place when these types of
responses from informants are interpreted as signs of a particular “local” theory
of interpretation: the “observational” level morphs into a first-stage “explana-
tory” level. With the confidence that comes with the access to local taxonomies
Levels of argumentation 177

made available by the linguistic labeling of actions and moods, ethnographers


then begin to search for more evidence to support their hypothesis about the
local theory of interpretation. At this point of their inquiry ethnographers also
face the question of whether they themselves are creating the problem, and thus
supporting a particular theoretical stance, by asking certain kinds of questions.
Could it be that, as suggested by Charles Briggs (1986) three decades ago, a
different way of asking about introspection would bring about a different kind of
answer? This line of argumentation was pursued by Eleanor Gerber, who, in her
research on Samoan emotions, criticized Mead for taking the above-mentioned
response to her questions as evidence for Samoans’ lack of interest in motiv-
ation. As shown in the following quote, Gerber claimed that she instead found a
way to get Samoans to show an interest in motivations.
I myself asked many similar questions of informants, and unerringly got the same reply,
“I don’t know.” But I was investigating emotions, and knew from other data that there
existed a complicated framework by which explanations of other people’s behavior
could be, and were, constructed. After a period of confusion, I discovered a way into my
informant’s interpersonal assessments. If I brushed aside their initial plea of ignorance,
saying, “Yes, I know – but why do you think, in your own mind, that he did this?”
suddenly my informants’ eyes would light up. “Well,” they would say, “in my own
mind . . .” and then would follow an explanation that (to my disappointment) frequently
had nothing whatever to do with emotions. (Gerber 1985: 133)

Unfortunately, Gerber ends up undermining her own claim. Not only does she
admit in the passage above that she was given explanations that “frequently
had nothing whatever to do with emotions,” but, in a more damaging way for
her criticism of Mead’s position, in the next paragraph she confirms what
Mead and, later, Shore had written, namely, that Samoans simply do not
believe one can know what another person thinks.
What prevented my informants from answering the first question was not lack of
interest in motivation, but a belief about knowledge. Samoans frequently say, with
the full force of self-evident conventional wisdom, “we cannot know what is in another
person’s depths,” or “we cannot tell what another person is thinking.” Given this idea, it
becomes very difficult to state with assurance what another person’s motives in fact are,
especially since the question “why” is apparently interpreted as calling for nothing but
the truth. As a consequence, the only possible answer to this question about matters
internal to another person is “I don’t know.” (Gerber 1985: 133)

The professed lack of knowledge of a person’s thoughts and feelings is


captured by the Samoan saying e le iloa e le tagata lona sesē, lit. ‘a person
does not know his/her3 (own) mistake,’ that is, a person might not be able to
monitor his own action or recognize his own wrongdoings (Shore 1982: 176).4
As shown in example (1) below, this expression is not just used in talking to
the anthropologist. It is spontaneously used to justify improper or disrespectful
178 Opacity of other minds

actions. In (1) Chief Tevaseu uses it to explain why an older orator from the
village next to his – whose name is not mentioned here but implied by the
discourse context – might have said some inappropriate words during the prior
exchange (see Duranti 1994: 94–97).
(1) (Fono April 7, 1979; Chief Tevaseu is wrapping up his conciliatory speech)
1599 Teva; ia` po`o lea fo`i se `upu
‘so if there are any (wrong) words’
1600 `ua- . . . mākau mai ai `oulua Makua ma- `Āiga-
‘(which) have been noticed (by) you two senior orators and
the chiefs’
1601 ma lo kākou gu`u, (. . .)
and (the rest of) our village, . . .
1602 ia` e le iloa e le kagaka loga sesē. SAYING
‘well, a person doesn’t know his own mistakes.’
1603 (1.0)
1604 Loa; mālie!
‘nicely (said)!’
1605 Teva; (e) vagagā lava le Akua,
‘except the Lord,’
1606 `o lo`o iai le kogu. (. . .)
who has the truth . . .

The above example shows that, in addition to directly asking questions, as


attempted by both Mead and Gerber, we can get glimpses of native theories of
interpretation by recording what people actually say to one another. More
importantly, as I will show below, by investigating what people tell one
another in the course of everyday interactions, we can even see what inferences
are made from words to thoughts and feelings (or vice versa).
Another opportunity to look for further evidence of the local theory of other
minds is presented through inter-cultural contact situations and the introduction
of new activities that typically accompany contact. For example, it makes sense
to hypothesize that the opacity doctrine would inhibit or in some way affect
participation in recently introduced Christian practices such as confession,
which relies on the assumption of the existence of an inner self that is accessible
to introspection (Cary 2000). This indeed seems the case among the Urapmin, as
described by Robbins (2008), and the Bosavi, as described by Schieffelin
(2008). In the first case, confessing in public made people very uncomfortable
and even suicidal in some cases – we can imagine the same effects on Catholics
in the West, who are accustomed to confession as a private activity. Among the
Bosavi, Schieffelin documented people resisting confessing altogether and
holding on to the privacy of their own thoughts, intentions, and desires.
Once a family resemblance among different practices is identified, some
ethnographers are tempted to move to an even higher level of explanation,
proposing to subsume the opacity doctrine under another, more general
Contextual variation and inconsistencies 179

principle. For example, Robbins (2008) suggests a general tendency among the
Urapmin to think of people’s minds as “private places” and calls this phenom-
enon “psychic privacy.” Stasch (2008) suggests that the opacity doctrine may
be linked to an egalitarian ethos. This particular generalization, however, does
not hold cross-culturally given that opacity doctrine phenomena have been
documented in highly stratified societies like Samoa (Ochs 1982; Duranti
1988). A more promising line of explanation is Stasch’s observations
regarding the potential relationship between opacity of mind and local attitudes
toward authority and responsibility, two dimensions rich with political and
ethical implications that can potentially have universal application.

8.2 Opacity of other minds as a problem


In a classic example of anthropological reversal in which, to use Spiro’s (1990)
terms, the “strange” is turned into “familiar” and the “familiar” into the
“strange,” some of us started to argue in the 1980s that there was nothing
really exotic about the reluctance to go into introspection or the avoidance of
intentional readings of other people’s actions, words included. Rather, as
suggested by Clifford Geertz (1983: 59), it was the western view of interpret-
ation and its associated concept of self or person that was the problem
(Rosaldo 1982; Duranti 1984, 1988; Rosen 1995). The only thing we had to
do was get rid of the philosophical bias represented by the views of philos-
ophers like Austin, Grice, and Searle, who had built a theory of meaning that
was too dependent on intentions and sincerity (see Chapter 2). If we could
recast intentional meaning as simply another “local” theory – this one origin-
ating, say, in Oxford, England, instead of a Polynesian or Papuan community –
then the anti-intentionalist or anti-personalist view of meaning documented in
a number of Pacific cultures (and in some other places) would no longer be a
“problem.” In fact, some of us hinted, reducing the role of intentions in
meaning-making might be beneficial, providing a theory that is closer to what
people actually do, think, and feel when they go about their daily life.
The “problem,” however, has not gone away. One of the reasons is that as
ethnographers continued to look for evidence of the non-intentionalist theory
of interpretation, some inconsistencies began to emerge. Even the people who
were said to be reluctant to explicitly “read others’ minds,” in fact, were doing
it sometimes.

8.3 Contextual variation and inconsistencies in the


ethnographic accounts
Throughout the ethnographic data that specifically address this issue, it is not
uncommon to find that the same people who in some contexts seem reluctant
to read others’ minds do not seem to have a problem doing so in other contexts.
180 Opacity of other minds

For example, Rumsey (2008: 465) tells us that the Ku Waru are not reluctant to
read others’ minds in courtship situations or when describing sexual attraction.
They also seem comfortable with the idea of being able to spot deception on
the basis of observations of others’ behavior. Among the Korowai, according
to Stasch (2008: 448), it is recognized that people may think silently. But this
recognition comes with the suspicion that silent contemplation may hide
malevolent intentions – I will return to this finding below because it offers
some important hints about how introspection is perceived.
The documentation of the adoption of the various Christian practices of
confession also suggests that they are not completely rejected as inappropriate
by some of the same groups that have been said to exhibit the opacity doctrine.
In fact, in Rumsey’s article we learn that, prior to the introduction of Christian
confession, the Ku Waru (as well as the Melpa and other groups in Papua New
Guinea) had their own indigenous practice of group confession that was used
in preparation for war or during war activities – a practice that was connected
to the belief that “pent-up anger and concealed wrong-doing can cause bodily
illness” (Rumsey 2008: 460). We also learn from Robbins that the Urapmin
people at first embraced the practice of public confession and claimed that they
were “able to see in the minds of others” (Robbins 2008: 425).
These inconsistencies suggest that our research strategies should be revised
so that we can be in a better position to document whether the same group of
people who seem reluctant to engage in speculating about what goes on in their
own mind or in the minds of others might, in fact, exhibit mind-reading
behavior under certain circumstances.
Having been socialized in a sub-discipline, linguistic anthropology, where
recording what people normally do is expected, I am always surprised that so
many of my colleagues, including sociocultural anthropologists, do not do that
as much or at all. The recording of what people spontaneously do should be
routine, especially now that digital recorders are so cheap and small and our
“subjects” use them too, thereby allowing for an ecological validity that was
not quite there a few decades ago. Armed with the hypothesis that the natives
have a way of thinking about what we can know about others that is different
from what is usually assumed in “western” theories of meaning, ethnographers
should be able to search for naturally occurring situations in which the same
reluctance to read other minds might manifest itself spontaneously, that is,
without any outsider’s prompting.

8.4 Examining spontaneous interaction


The recording of spontaneous interactions provides us with an opportunity
to examine whether participants make inferences about what others are
thinking, feeling, and wishing or, instead, resist making such inferences.
Examining spontaneous interaction 181

Samoan conversations are potentially telling in this respect precisely


because of the claims made by Mead (1928), Shore (1982), and me in
earlier chapters.
I first went to look for the uses of the expression ta ilo, which Mead
mentioned as a common response to her attempts to look for motivations.
She translated it as ‘Search me’ and an alternative to ‘I don’t know.’
Literally, the expression is a combination of the first person singular
affective pronoun ta ‘I, me,’ which carries the pragmatic implication of a
negatively affected subject, as in ‘poor me’ – Milner (1966: 220) glosses it
as “denoting self-abasement together with an appeal for sympathy, love, or
help” – and the verb ilo ‘perceive, be aware’ (Milner 1966: 84), often used
with the suffix –a to mean ‘know’ (e le iloa e a`u ‘I don’t know’ or ‘me,
I don’t know’).
The transcripts of my recordings of spontaneous conversation contain a
number of examples of the expression ta ilo (pronounced /ka ilo/ or /ka `ilo/),
for example in responses to yes-or-no-questions and wh-questions.5 But even
the ‘why-questions’ in my corpus tend to be about causes or reasons for things
being the way they are or reasons for having to do something in a particular way,
as in (2) below, where Father asks why the son (Sefo) needs to quickly get to
school. The question in (2), like the other ones I have found in my transcripts,
do not seem to request or elicit a response about someone’s intentions or
motivations.
(2) (“Dinner # 3”, 1988; Father (Fa) and Mother (Mo) are having dinner with
their children; a number of neighbors have stopped by to chat)
Fa; o le ā le kaimi e alu ai Sefo pe`ā alu i Apia?
‘what time does Sefo leave if (he) goes to Apia?’
Mo; e alu pe`ā uma ga `oki le ulu
‘he’ll leave after he has a haircut’
(pause)
Mo; se`iloga e fai mai e vave aku
‘only if they say (that he has) to get there quickly.’
Fa; vave aku e ā?
‘why would (they want him to get) there quickly?’
Mo; ka`ilo ‘How would I know?’
‘How do I know?’
lē iloa pē fai se fogo a kamaiki. Sefo! Rephrasing
‘(I) don’t know if the kids are having a meeting. Sefo!’

A different strategy for looking at possible examples of mind-reading in


interaction is to look at cases where a speaker makes a prediction about what
someone else was, is, or will be thinking, feeling, or wishing. Such examples
are found in my data, often expressed with the verb fia, a modal verb with a
range of related meanings, including ‘wish, want, like, need,’ in combination
182 Opacity of other minds

with a wide range of other verbs or adjectives to form predicates that express
physiological needs such as ‘be hungry’ (fia`ai, lit. ‘wish, want, like, need to
eat’) or ‘be thirsty’ (fia inu, lit. ‘wish, want, like, need to drink’), as well as
attitude or personal characteristics.
The reading of others’ needs and wants might be part of the cultural script of
particular situations, like when guests are around. As pointed out by Adam
Harr (2013: 319), “Hospitality . . . demands that hosts actively look after
guests’ needs without those needs being openly expressed. I quickly learned
that if my eyes flitted to a bowl of food for even a moment, it would be offered
to me immediately.” This is a kind of anticipation that in a place like Samoa
might be also motivated and explained by means of social hierarchy. Younger
and lower-status individuals are expected to anticipate what older and higher-
status people might need or want, regardless of whether a wish has been
expressed. In addition, the attribution of wishes can be used strategically to
compel others to offer food or services. This is shown in example (3) below,
taken from a videotaped interaction among four Samoan women who have
been cleaning a communal house belonging to their church congregation.
While M is instructing her son to go and get two spoons from the people in
the next house, T attributes a want of bread to Vaega (Vg), another woman in
the group. Immediately following T’s utterance, Vaega reinforces the request
by attributing the same wish to Vaetolu (here/Vaekolu/), the oldest of the four.
(3) (“Women of the Congregation”; August 1988)

M; kamo`e- sh:::- alu e ku`u le `apa i lalo o le paipa . . .


‘run okay- go and leave the bowl beneath the tap . . .’
ga `ē kamo`e mai lā`ia ma gi sipugi se lua
‘and then run back with two spoons’
e kausami ai kama`ika`i la`a kuai
‘for the ladies to eat with. it’s getting late’
T; fia `ai falaoa o Vaega`au `ua leva ga fa`akali
‘Vaega`au really wants to eat bread she’s been waiting a long time’
Vg; mm:: lea fa`apegā fo`i Vaekolu
‘yes Vaetolu is like that also’

The modal verb fia is also used in all kinds of compounds that express desires
that are seen as antisocial or reproachable. These are negative traits of some-
one’s personality that can be inferred from the way a person behaves with
others or in front of others. Their encoding by means of a fia-compound carries
a negative connotation and implies that such ways of being should have been
repressed.
One example of this type of negative attribute is found in (4) below, where
the expression fia tama matua, lit. ‘wish, want, like, need to be old (matua)
boy (tama)’ describes someone who acts in such a way that presupposes that
he feels more important than he really is given his age.
Examining spontaneous interaction 183

(4) (“Dinner # 3”; August 1988)


Fa; pei pea ai.
‘leave him alone.’
kalia ā ga ia ma alu ifo e fai le mea`ai
‘do let him go and make food (by himself).’
Mo; fia kama makua ‘wants to be a big boy’
‘(he) wants to be a big boy’ or ‘acts like an older boy’
Fa; gā la`a fia kama makua le ali`i gā
‘that guy wants to be a big boy’
ma kīgāiga ai `oulua e fai aga fe`au.
‘and you two have to suffer for it (and) do his chores.’
The next example is taken from the same video that gave us example (3)
above, but a bit earlier in the interaction. While they are gossiping, one of
women realizes that they are being watched by the children of someone they
all know, a woman named Falafala (a pseudonym). At this point the four
women speculate that Falafala’s children will go and report what they saw to
their mother, who will inevitably come later to find out what the women were
talking about and to add her own (annoying) stories.
(5) (Women of the Congregation, August 1988)

Vg; va`ai fo`i lā i le fāgau a Falafala ((name))-


‘look there at the children of Falafala’
lā e kilokilo mai lā ua koe ō
‘they are looking here, now they’ve gone back’
Vt; e ō mai e kilokilo mai po`o i ai gisi `i`igei
‘they’ve come to check out if there’s anybody here’
Vg; va`ai- va`ai
‘look- look’
M; mm::
‘mh’
((three turns left out))
Vt; lili mai ā e va`ai po`o iai gisi
‘they’ve come to see if anybody’s here’
T; e ō aku gei fa`apea- lā e `a`ai- o ai?
‘now they go (and) say- they’re eating- (and they will be asked)
who?’
Vg; Vaekolu `o Kamae.- `o Malue
‘Vaetolu Tamae- Malue’
Vg; e lē mafai gei ga lē sau- // auā o le mea fia iloa
‘(she) cannot not come // because the thing is (that she) wants
to know’
[
Vt; mm::
‘mh’
T; sau gei kīgā kaliga
‘(she’ll) come here (tonight) and our ears (will) hurt’
184 Opacity of other minds

One could argue that this is an example of reading another person’s mind.
I would not disagree with that conclusion but I would also suggest that if we
want to invoke “mind-reading,” we need to acknowledge that it is a limited
case of mind-reading because it is based on expectations grounded in know-
ledge of a particular type of person. The women know that Falafala tends to act
in a certain way. Her likelihood of wanting to come to see them and gossip is
not that different from other highly predictable events, including that her
children will report what they have seen – this is what children are expected
to do – or that certain people will go to church on Sunday. This is a type of
inference that is based on repeated, generalizable, and even routinized behav-
ior. The person whose wants are being assessed is being treated as a type more
than as an individual. The case is similar to the situation in which we guess that
someone might be wanting to eat simply because that person is in our house at
dinner time and there is smell of cooked food coming out of our kitchen.
The next example is a more direct challenge to the opacity of mind claim
about Pacific societies like Samoa. It is taken from an exchange during a
family dinner that I video-recorded in the same village where the other
interactions reproduced in (1)–(4) took place. In (5), six-year-old Onosa`i
(a pseudonym), the youngest girl in the family, has been begging her mother
to give her a bigger piece of banana (see Duranti 1994: 154 for the earlier part
of the interaction). Little Onosa`i’s actions are witnessed by her older siblings,
including her nine-year-old sister, Roma (a pseudonym), who first verbally
displays her disapproval of such selfishness (in line 516) and then ascribes to
her younger sister the specific wish not only to get a bigger piece but “the very
big one” (or “biggest”) (in line 519).

(6) (Family Dinner #2, August 1988)

514 O; ((to her mother)) `aumai le mea fo`i gale!


‘give me that one (piece) over there!’
515 ((The mother complies with O’s request by switching the piece
of banana on O’s plate))
516 R; `ē! `o lo`u igoigo ia i amio a le la`ikiki // lea!
‘Hey! I’m so disgusted at the behavior of this little one!’
517 O; leai, lea, e lē:-, e lē (le) kipi::.
‘No, this, it’s not- not the one that’s cut.’
518 ((Once more the mother switches the piece of banana with
a bigger one))
519 R; maga`o ā e fa`akelē aga ia.
‘(She) wants hers to be the biggest.’

What is important in this example is that little O has not said that she wants the
biggest piece. She has merely rejected, rather forcefully, the pieces that her
mother has tried to give her. For this reason R’s statement that O wants the
Examining spontaneous interaction 185

biggest piece available must count as an inference about what O is thinking but
not saying.
These two examples demonstrate that some kind of mind-reading obviously
goes on in Samoa, as in any other place in the Pacific or elsewhere in the
world, and that in some situations mind-reading is also made explicit through
language. In this case we have a nine-year-old girl voicing her interpretation of
the (morally “bad”) intentions of her younger sister. This means that the public
reading of intentions constitutes an explicit condemnation expressed through
moral indignation in line 516 (‘I’m so disgusted at the behavior of this little
one!’). This contextual information is important because R’s utterance in line
519 could be read as a potential counterexample to Ochs’ (1982) claim –
discussed in Chapter 3 – that Samoans do not expand children’s unclear
utterances or engage in clarification sequences. Older sister R is de facto
guessing what little O wants. This could be explained in a number of ways.
One is that, at nine, older sister R might be engaging in a type of interpretive
action that adults usually avoid. This explanation would go along with the
hypothesis that Samoan children are more likely than adults to publicly engage
in this type of mind-reading and only later would “grow into” or more fully
adopt a cultural dispreference for clarification and explicit mind-reading.
Another hypothesis is that older sister R is stepping in on behalf of her parents
to control and reprimand younger sister O’s improper behavior. The mind-
reading in this case, as in the other cases mentioned above, is based on sister
O’s visible actions. Such actions display a stance or disposition that calls for an
immediate corrective action.
All of this means that (i) a certain amount of figuring out what others are up
to is always going on and is necessary for people to manage their daily life; (ii)
whether or not this type of mental activity should always be glossed as
“reading other minds” depends on the specific situation as well as on our
theory of human action, including our view of intersubjectivity; (iii) conscious
and explicit reading of other minds is one of the possible routes to understand-
ing a situation retrospectively and prospectively; (iv) communities (and indi-
viduals) vary in the extent to which reading other minds is recognized,
verbalized, and justified.
Because of its unique research conditions and interests, anthropological
fieldwork can help us refine each of these generalizations. In particular, it
can clarify the role of ideology in how people see and practice introspection.
Although explicitly reflecting upon one’s own thoughts and feelings as well as
upon the thoughts and feelings of others is considered necessary if not
innocuous by most philosophers, anthropologists have shown that such activ-
ities are seen as highly suspicious by ordinary people in the Pacific (and
elsewhere). For example, Stasch (2008) tells us that in those cases in which
introspection is acknowledged among the Korowai it is cast as a morally
186 Opacity of other minds

dubious activity. This is an important hint because it can help us explain why
people feel uncomfortable about trying to speculate about others’ thoughts and
desires. Our asking people to tell us what they imagine that others are thinking
might be like asking them to spy for us. The ethnographer’s curiosity might
imply that there is something wrong or devious that needs to be uncovered.
A number of ethnographic accounts show that the very act of bringing out in
public one’s speculations on the mental activity of others makes speakers
worry about potential retaliation. Hence, from a sociocultural point of view,
the phenomenon of the opacity doctrine might be seen as a defense strategy
against the accountability that comes with making claims about what others
think or want. It is perhaps not by accident that the three Samoan examples
briefly discussed above involve moral indignation. It could turn out that the
opacity doctrine hides or at least implies a pan-human preoccupation with
reducing one’s accountability. The study of the conditions that might increase
or decrease such a concern is a worthwhile research project.
9 Intentions and their modifications:
a lesson from Husserl

9.0 Introduction1
In this chapter I return to Husserl’s notion of intention, which I briefly
introduced in Chapter 2 (§2.4), and combine it with his notion of “modifica-
tion” to describe the role that directing novices’ attention plays in socialization.
I also use Husserl’s conceptual apparatus to hypothesize that, even though
children and novices have ways to modify their habitual way of being and the
ability to assume what Husserl called a “theoretical attitude,” the accumulation
of earlier intentional and phenomenal modifications into a well-established
habitus might prevent certain changes and thus certain types of learning from
occurring.

9.1 Husserl’s intentionality and the relationship with the


modification of attention
For Husserl, intentionality covers a wide range of acts that go from transform-
ing raw sensations into a meaningful entity (e.g., a sequence of sounds into a
song) to combining several such experiences into more complex meanings
(e.g., the remembering of the time when we heard that song for the first time).
An important element of Husserl’s phenomenology is the key role of attention.
We speak of intending . . . in the sense of specially noticing, or attending to something
[im Sinne des auf etwas speciell Achtens, des Aufmerkens]. An intentional object need
not, however always be noticed or attended to. Several acts may be present and
interwoven with one another, but attention is emphatically active in one of them. We
experience them all at the same time, but we get absorbed by this particular one.
(Husserl 1901: 557; translation, here slightly altered, in 1970b: 562)
By bringing some objects to the foreground while leaving others in the
background, attention is one of the ways in which the temporality of our
being-in-the-world gets realized (Arvidson 2003, 2006; Depraz 2004; Husserl
2004). To better capture the different ways in which the phenomenal world
changes for the perceiving, thinking, acting, and interacting Subject, Husserl at
times used the notion of “modification” (German Modifikation) – a term that

187
188 Intentions and their modifications

bears a family resemblance with other terms also used by Husserl, including
Wandlung, Abwandlung, and Änderung – combined with a number of adjec-
tives that express different experiences as well as different theoretical points of
view. The English translations of his writings render these various combin-
ations with such terms as “intentional modification,” “phenomenal modifica-
tion” (Husserl 1970b), “phenomenological modification” (Husserl 1989),
“noetic modification,” “logico-categorical modification” (Husserl 1931),
“retentional modification” (Husserl 1991, 2001), and “modification of atten-
tiveness” (Husserl 2001), among others.
In this chapter, I bring together Husserl’s theoretical apparatus with some
insights from developmental pragmatics and language socialization ([Ochs]
Keenan and Schieffelin 1976; Ochs and Schieffelin 1984). A number of
studies from these two closely related traditions have shown that adults and
experts are routinely engaged in getting children and novices to see and hear
what is culturally relevant. I propose to rethink the acculturation of attention in
terms of Husserl’s notions of intention and modification. I will argue that
socialization can be conceptualized as a series of attempts to shape children’s
and novices’ phenomenal world through a series of “modifications” of what
they see, touch, smell, hear, or taste. In some cases, the modification is more
complex, consisting of a change of perspective whereby a person moves out of
what Husserl called “the natural attitude” to assume a “theoretical attitude.”
I will show that both adults and children are capable, under certain circum-
stances, of making these more radical modifications of their way of being in
the world.

9.2 Phenomenal modifications: the constitution of linguistic acts


For Husserl, language provides an excellent example of how we are constantly
engaged in the process of experiencing modifications of the phenomenal world
even though we are usually not aware of it.
Even though for Husserl we are normally aware of our sensations (Zahavi
2002: 56–58), we are not aware of the role that we have in making sense of or
constituting our everyday world (Sokolowski 1964; Zahavi 1992). We are
usually too busy living our lives to realize that we are the ones who give
meaning to objects, people, and events – whether real or imaginary – through
intentional acts, and that such acts are in turn constituted by modifications of
our ways of seeing, hearing, smelling, thinking, etc. Believing that language
provides the clearest cases of our active role in meaning-making,2 Husserl used
linguistic signs to explain how constitution works (see also Sokolowski 1964:
57). In Logical Investigations he asks us to carefully examine the process
whereby a physical “sign phenomenon” (Zeichenerscheinung) becomes an
expression (Ausdruck), that is, something meaningful. In a few lines, Husserl
Intentional modifications 189

sets up for us what amounts to a phenomenological experiment (Husserl 1901:


40–41; English translation in Husserl 1970b: 282–283). He asks us to consider
“a sign in itself” or – as translated by J. N. Findlay – “a sign qua sign” (Husserl
1970b: 282), that is, as a sensible thing that does not have meaning yet.3 We
are supposed to focus on a printed word without the “meaning-intention”
(Bedeutungsintention) that we normally associate with it. As we do this, we
realize that the entity (the visible marks on a piece of paper) is still present, that
is, “there” for us to see, but it no longer means what it meant a few seconds
earlier. We apprehend it but we no longer understand it as something endowed
with linguistic meaning. It has acquired a new meaning-less, intention-less
status. We no longer relate to it as a “word.” Only after we have made
ourselves experience the modification of the “essence” of the word as a
word-phenomenon and changed it (for us) from meaning-full to meaning-less,
can we come to see that ordinarily the direction of the modification works the
other way around. Before the object “printed mark” or any other object of the
same kind can be a unit of meaning – an “expression” in Husserl’s terms – it
typically “undergoes an essential phenomenal modification” (“eine wesentlich
phänomenale Modifikation” [Husserl 1901: 41]) whereby, even though “what
constitutes the object’s appearing remains unchanged, the intentional character
of the experience alters” (Husserl 1970b: 283). The notion of “phenomenal
modification” already in the 1900–1901 edition of Logische Untersuchungen
shows Husserl’s early use of the notion of modification.
Something similar to what happens in our understanding of a written word
goes on when we encounter the spoken word. Just as for written language we
must assume that the marks on paper are more than a “brute physical datum”
(Sokolowski 1964: 57), for linguistic sounds to be interpreted as such, we need
to make them into something more than “mere sounds” (bloße Laute).4
According to Husserl, this “something more” includes a purpose for communi-
cating about something, the imparting of a “sense” (Sinn) that comes from the
speaker’s mind through an intentional (or meaning-imbuing) act that can be
understood by the hearer.
Husserl’s theory of the constitution of linguistic meaning is meant to be part
of a more general theory of how humans make sense of the surrounding world.
As I will show in the rest of this chapter, this aspect of his work makes it easy
to connect language with other meaningful human experiences.

9.3 Intentional modifications


As we live our everyday lives, we experience continuous shifts in the ways in
which we understand the world, that is, in the ways we think of, feel about, or
act toward objects, people, animals, and the events in which we participate or
that we hear about through narrative. At one moment, while we are typing or
190 Intentions and their modifications

reading in our study, we might be barely aware of the books that are on our
desk even though they are in front of us and within our peripheral vision. But
as we decide that we need to consult one of those books, we start to relate to
them in a different way. As we attentively and quickly glance at the books, we
match each one against the memory that we have of the book that we are trying
to find. We find ourselves quickly examining each book in terms of the color of
its cover or in terms of its thickness. We also consider its place on the desk and
its proximity to other books to assess their content or relevance to our current
needs. We might also stop and think about the fact that one of those books is
the one that is overdue at the university library and that another one is the book
that someone gave us for our birthday and we haven’t read yet. We might
notice the picture on the front cover of one of the books and think that it is
poorly designed. We might pull one of them out, hold it in our hands and
wonder what made it into a book that so many people buy but very few read.
Throughout these moments, as our gaze moves from one book to the next and
our hands grab one to then place it back where it was or put it aside for a future
occasion, it is not just our attention that is continuously shifting. The way we
are disposed toward what we see or touch also shifts. At any given moment,
each of those books is the same object that was in front of us a few seconds
earlier; in other words, that is, our perception of it as a physical object has not
changed (for example, its color, weight, or smell has not changed in any
perceivable way), but our consideration of it, the way we direct our attention
to it has changed. These shifts in our ways of thinking of, feeling about, or
coming in contact with the same object is what Husserl called “intentional
modifications” (Husserl 1931, 1989). A modification in our consciousness –
often accompanied by or realized through a modification of our embodiment –
may affect entities that were in the background and suddenly become the
object of our attention (e.g., the book that was lying on our desk and now is
examined to see whether it is the book we want to consult) as well as entities
that we attend to in a different way (e.g., the cover of a book may change from
source of information about its contents to target of an evaluation about its
design).
The concept of intentional modification is important because it makes
evident the role that human subjects play in meaning-making through inten-
tionality. It is in this sense that Husserl (1931, 1970b, 1989) uses the verb
“constitute” as well as its nominal and adjectival derivations (e.g., “consti-
tution,” “constitutive,” “constituting”). Our way of relating to entities in the
world, whether real or imaginary, does not “create” them out of nothing, but it
“constitutes” them, that is, it “objectivates” them – makes them acquire
objectivity – through distinct intentional acts, with distinct meanings (Soko-
lowski 1964: 46). As briefly discussed in Chapter 5, we can attend to the same
“object” or “referent” as something (or someone) that we need, want, admire,
Intentional modifications 191

feel curious about, despise, miss, judge, are repelled by, feel comfortable with,
and so on (see Figure 5.4).
When the type of intentional act we entertain toward something or someone
changes, e.g., from admiration to fear, from disapproval to approval, from
seeing it as something alien to seeing it as a member of a familiar group, we are
experiencing an intentional modification, that is, the “phenomenon” – in the
sense of what it appears to be for us – changes as a result of our way of relating
to it. We can schematically represent this event with a time variable t next to
the two intentional acts schematically represented in Figure 9.1.5
Some classic examples of this type of change in the domain of vision are
provided by Gestalt psychology and have been discussed by other philoso-
phers, among them Ludwig Wittgenstein. In his Philosophical Investigations,
he used the example of the picture in Figure 9.2 (“the duck-rabbit” picture) to
discuss the concept of “aspect seeing” or “seeing as . . .” This is a kind of
seeing that is different from mere perception.
I might have been looking at the picture in Figure 9.2 for years and only seen a
rabbit. Suddenly I see a duck or, rather, I see it as a duck. The “it” as a picture

Subject A Object G

Figure 9.1 The same Subject (A) is directed toward the same Object (G) at
time t1 in the intentional act of admiration and at time t2 in the intentional
act of fear

Figure 9.2 The rabbit-duck figure discussed in Wittgenstein (1958: 194)


192 Intentions and their modifications

(corresponding to what Husserl, following Aristotle, calls the “hyle”) has not
changed but what I see it as has changed. Wittgenstein compares the ability of
“seeing an aspect” to having a “musical ear” (1958: 214) and to “experiencing the
meaning of a word.” In all these cases, we are relying on an ability to go beyond
mere perception. The similarity between Wittgenstein’s “aspect seeing” and
Husserl’s intentional modification finds support in a terminological convergence.
At one point, to express what he means by “aspect seeing,” Wittgenstein uses the
German word Einstellung – translated as “attitude” by G. E. M. Anscombe (see
below). He says that to “see as” is the same as to have an “attitude” toward
something. For example, to see “an animal pierced by an arrow” is the same as
saying that “this is my attitude to the figure” (“dies ist meine Einstellung zur
Figur”) (Wittgenstein 1958: 205). Husserl also used Einstellung in his 1913
Ideas (English translation in Husserl 1931) to describe the way in which we relate
to the world, e.g., in his notion of natürliche Einstellung “natural attitude” (see
below). Both authors were trying to capture how meaning-giving consists of a
change in attitude or disposition toward something.
Below I will give some examples of the ways in which language can be used
to get novices to modify their attitude toward something. In this case, the
intentional modifications that are being promoted do not concern vision but
aural perception. Drawing from the world of music education, instead of
discussing cases of “seeing as,” I will discuss cases of “hearing as.” The cases
in question will also give us an opportunity to appreciate the range of modifi-
cations that language is expected to encourage in order to socialize students to
a professional way of hearing.
The discussion of music is here partly motivated by the fact that it was very
important for Husserl’s theory of time consciousness (Husserl 1991). Stimu-
lated by the work of Franz Brentano and of his thesis advisor Carl Stumpf, who
wrote on the psychology of music and sounds and to whom Logische Unter-
suchungen is dedicated, Husserl tried to answer the question of how we arrive
from a succession of tones to hearing a melody. The succession must be united
in our consciousness but, as Husserl wrote, “[w]e obviously do not have the
tones all at once” (Husserl 1991: 22). The problem of how we hear a melody
became a model for understanding not only “temporal objects” like tones, that
is, objects that have duration, but also the very constitution of time in our
consciousness, what Henri Bergson (1908) had called “la durée.”

9.3.1 Instigating intentional modifications in jazz students


It is well known that musicians’ ways of hearing music are distinct from non-
musicians’. What is perhaps less well known is that each genre of music
requires special kinds of “listening.” In the jazz tradition, musicians playing
Intentional modifications 193

in a band are expected to respond on the spot to any change in rhythm,


tempo, harmonic structure, or melodic lines initiated by the band leader
(for large ensembles) or by one of the other musicians (in small ensembles)
(e.g., Berliner 1997; Monson 1997; Berger 1999; Reinholdsson 1998; Sawyer
2001; Duranti and Burrell 2004). As documented by Paul Berliner (1994),
jazz students “must depend greatly upon their ears” (1994: 93) for a number
of tasks, including the fact that “much of the jazz repertory remains part of
the community’s oral tradition” (1994: 93) and the need “to apprehend the
unique features of each rendition as they unfold during a performance,
instantly adapting their parts to those of other players” (Berliner 1994:
93–94).
In what follows I suggest that, in trying to socialize their students to develop
a “jazz way” of listening to music, jazz instructors are asking those students to
engage in “intentional modifications” of their ordinary or previous ways of
listening. Further research is needed to document the effects of these experi-
ences on students over developmental time.
The three examples that I discuss are taken from my video recordings of two
jazz performance classes (called “Jazz Combo Classes”) at UCLA.6
In the first example, the four students in the class7 have just finished
playing together “Lament,” a “ballad” (that is, a song that is meant to be
played at slow tempo) written and performed by J. J. Johnson, one of the
most famous trombone players of modern jazz. The instructor, George
Bohannon, himself a renowned trombonist, has quietly listened to the
students play and now is providing feedback. He is talking in front of
everyone but specifically addressing the student who played the trom-
bone. Bohannon just mentioned the importance of following the chord
structure of the song during the improvised part (the player’s “solo”), so
that the notes played would fit in properly. Then, after conceding that
playing “ballads” is hard, Bohannon reminds everyone that teaching
alone cannot do it. Playing good solos comes from hearing the solos of
great masters.
(1) (Jazz Combo Class, February 10, 2003)8
GB; [. . .] As uh we have said over and over. the mo:re you listen. the
mo:re you’ll understa:nd. you know because i- it’s. you can teach a
lot of it but- a lot of it comes from your hearing. hearing good solos.
hearing what J. J. ((Johnson)) has played on his:: tune or- listening
to::- you know some of the other great- trombone players you
know- or not just trombone players anybody you know I’ve- I’ve
listened to more of the trumpet players and piano players than I did
to- [. . .] trombone players although I love what J.J.- his playing.
[. . .]
194 Intentions and their modifications

Figure 9.3 George Bohannon (on the right) interacting with the students in his
Jazz Combo Class on February 10, 2003, saying “Hearing what J. J. has
played on his tune” – see example (1)

The students here are told that they have to attend to the music of great players
like J. J. Johnson and others in such a way as to detect what they are doing in
their solos, that is, when they stop following the original melody and impro-
vise, creating new melodic lines (and more). This kind of advice is commonly
directed to younger players in the jazz community, as documented by Paul
Berliner (1994), among others.9 What I am interested in underscoring here is
that this type of attending to jazz recordings constitutes a change in the ways in
which one may otherwise attend to the recorded sounds. Bohannon’s choice of
two distinct verbs, hear and listen, within a stretch of discourse that seems to
be focused on the need for one particular activity (to learn from the ways in
which recorded artists perform their solos), calls for reflection. Whereas
“listening” (as in “listening to . . . some of the other great trombone players”)
is here reserved for the willful disposition to pay attention to how certain
musicians play, “hearing” (as in “hearing good solos”) seems to refer to the
ability to differentiate among those sounds and absorb them. The contrast is
captured in the difference between expressions like “I’m listening but I can’t
hear it” and “I’m hearing it but I am not listening.” In the first case I am
making an effort but I cannot get to the phenomenon I am looking for. In the
Intentional modifications 195

second case, I am being exposed to the sounds but I am not concentrating,


paying attention.
Instructors may also give hints about a specific quality of playing that the
students must be tuned into. In the second example, the students in drummer
Sherman Ferguson’s Jazz Combo have been rehearsing the song “Cain and
Abel” written and performed by Branford and Wynton Marsalis, saxophone
and trumpet player respectively. After some discussion on how it should be
played, they decide to listen to the recording by the Marsalis brothers’ band.
While they are listening to the CD playing, Ferguson first gives some hints to
the two students who play the saxophone and the trumpet, then he turns to the
drummer and tells him to listen to the drummer Jeff “Tain” Watts in the
recording. This time the listening must focus on how the drumming is done
in an “interactive” way.
(2) (Jazz Combo Class; April 23, 2003, tape 1; SF ¼ Sherman Ferguson)

SF; that’s Tain ((motions with his hands as if playing the drums)) . . .
(listen to how-) the way he plays . . . he’s in- interactive with them.

Figure 9.4 While listening to the song “Cain and Abel” with his students, on
April 23, 2003, Sherman Ferguson points out that the drummer in the recording,
Jeff “Tain” Watts, is “interactive with [the other players]” – see example (2)
196 Intentions and their modifications

Musicians’ attentive listening to other musicians in their ensemble (no matter


its size) is expected in any kind of music performance. In each tradition,
however, attentiveness to other players is subordinated to higher-order goals
that are specific to its aesthetic canons and socio-historical context. It is
therefore with respect to the expectations specific to jazz music, and its
improvisational qualities, that the instructor’s request in example (2) must be
interpreted. In the language of Wittgenstein’s notion of “aspect seeing,” we
can say that Ferguson is asking the drummer in his class to listen to Jeff “Tain”
Watts not just as a drummer, but as an interactive drummer. To better
understand what this might entail, we can go back to the jazz combo class
from which example (1) was drawn. In this case (example 3), we find the
instructor, trombonist George Bohannon, addressing the drummer in his class.
The issue is how to achieve a collective clarity of sound. First, one needs to be
listening in such a way as to recognize the problem of the lack of clarity. To
convey this idea, Bohannon uses the metaphor of a crowded room where
everyone speaks so loudly that it is impossible to hear what people are saying.
(3) (Same setting as example 1, but earlier; GB ¼ George Bohannon)

GB; [. . .] It’s a good idea when you know when . . . ((the bass player))
Aaron is playing to- . . . bring it down ((in volume)) you know ’cause
he’s not amplified but he’s pulling. real well. [. . .] we don’t wanna
drown him out. so the contrasts are really what’s- . . . what’s so-
important. when everything is right here ((motions with hand)) [. . .]
it’s like- when you’re in a room full of people and everybody’s talking
and you don’t hear- you can’t hear anything. but you hear a lot of
chatter but there’s no definition. [. . .] but soon as- you drop it down a
notch. Then, you know that he- he’s speaking and you’re
accompanying him. and then y- you know- even if-. there’s more
chance of doing some interplay together you know, . . . [. . .] be
aware of that.

In this case Bohannon is telling the student drummer in his class to adjust his
playing by (a) listening to the sound that the band is producing as a whole,
(b) evaluating his own contribution to that sound (e.g., whether it is too loud),
(c) listening to a particular instrument (in this case the bass), and (d) evaluating
whether that instrument is being heard not only by the potential audience but
by the other members of the band. Implicit in these recommendations are
certain assumptions about the expected role that each instrument has, e.g., the
role of the bass lines to provide the rhythmic and harmonic foundations on
which the horn players can improvise (Monson 1996), but the important
point in each of these three cases is that the students are being encouraged to
attend to the music of others as well as to their own music in new ways. These
are activity-specific ways of paying attention. For Husserl, attention
The “natural attitude” 197

(Aufmerksamkeit) is nothing other than a kind of intentional modification, a


characteristic that, in his view, the psychologists of his time were missing
(Husserl 1931: 250 fn).
Music students can actually try to engage in this kind of listening because
they already have the ability to engage in other kinds of modifications,
including the “phenomenal modification” necessary for understanding lan-
guage. We could say then that our speaking and listening in everyday life is
always a speaking and listening in particular ways, or “speaking and listening
as” in the sense of Wittgenstein’s notion of “aspect-seeing.” We speak or listen
as friends, lovers, teachers, colleagues, employers, parents, children, neigh-
bors, customers, homeowners, and so on. All of these ways of speaking and
listening require particular kinds of modifications of our way of relating to the
elements in our surrounding world. We are usually not aware of such modifi-
cations. All together, they constitute what Husserl called the “natural attitude.”

9.4 The “natural attitude”


For Husserl, our everyday experience of the world is characterized by what he
called die natürliche Einstellung, translated as “the natural standpoint” by
Boyce Gibson in Ideas: General Introduction to Pure Phenomenology (Hus-
serl 1931) and as “the natural attitude” by subsequent translators of Husserl’s
works (e.g., Husserl 1970a, 1989).10 In this type of attitude, our understanding
of the world is practical rather than conceptual and reality is not problematized
(Husserl 1931: 91).
[T]his world is not there for me as a mere world of facts and affairs [Sachenwelt11], but
with the same immediacy, as a world of values, a world of goods, a practical world.
Without further effort on my part I find the things before me furnished not only with the
qualities that befit their positive nature, but with value-characters such as beautiful or
ugly, agreeable or disagreeable, pleasant or unpleasant, and so forth. (Husserl 1931:
92–93, emphasis in the original)

It should not be difficult for anthropologists to recognize this attitude as the


kind of experience of the everyday world that we try to capture through
ethnographic fieldwork. Even though we would call it “the cultural attitude”
rather than the “natural attitude,” Husserl’s use of the term “natural” is not
foreign to anthropologists. For example, we use it when we say that people
“naturalize” their ways of acting, thinking, and feeling. By this expression we
mean that they take those ways to be universal ways of being, instead of
recognizing them for what they are: the product of a complex process of
socialization in which “nature” (e.g., evolution, biology) plays a role but is
always transformed and interpreted through human praxis. This explains old
and new views of certain Others as “savages” or “uncivilized.” They are
198 Intentions and their modifications

considered so not because they are different from us, but because they are
different from the ways we expect a “person” to be. The concept of the
“natural attitude” captures the practical, moral, and aesthetic stance that we
ordinarily take toward the surrounding world, human beings included.
Husserl also points out that as human beings we do have the capacity to step
out of the “natural attitude.” When we do step out, we have the possibility of
taking on other attitudes. One class of such other attitudes is what Husserl
called the “theoretical attitude.” As we shall see, to shift to this attitude we also
need to engage in particular kinds of modifications.

9.5 The “theoretical attitude”


By reading Husserl’s writings we learn that the capacity to step out of the
“natural attitude” and enter the “theoretical attitude” is not something restricted
to intellectuals or scientists. It is part of our everyday experience of the world
(Husserl 1989: §4). We transform a perceptual or emotional experience into
the object of a theoretical attitude. This is made possible by what Husserl
called “a phenomenological modification.”12
This characteristic change of attitude belongs, as an ideal possibility, to all acts, and
accompanying it is always the corresponding phenomenological modification. That is,
all acts which are not already theoretical from the outset allow of being converted into
such acts by means of a change in attitude. (Husserl 1989: 10; emphasis in the original)

This phenomenal modification is different from the phenomenal (or intentional)


modification because in the former we maintain the same “attitude” (e.g., the
“natural attitude”) but the “object” changes meaning for us (e.g., from a noise to
a person’s voice, from a baby’s crying to a cat’s meowing). In the latter we
change our stance or positioning. For example, we go from being a participant in
a situation (e.g., a teacher, a guest, a parent, a mechanic) to an observer of
ourselves and others participating in that situation. This might happen when a
guest who is a writer is struck by what someone just said and decides to use it for
one of the characters in the novel she is writing or when a language learner
realizes that he has a hard time understanding a particular accent.
A phenomenological modification produces a “theoretical attitude” when
we transform a particular experience into an object of our reflection (e.g., the
object of an evaluation). Before examining how this concept can help us
understand what happens between adults and children, I provide an example
taken from a vast corpus of video recordings of ordinary and extraordinary
moments in the life of someone who is running for political office. The
example I have chosen is meant to be “ordinary” and yet, as I hope to show,
revealing of the difference between being “in” an experience and stepping
“out” of it to make it, through language, into the object of our reflection.
The “theoretical attitude” 199

Example (4) below reproduces a short portion of a verbal exchange on


March 2, 1996, between two people who, together with the present author,
are driving north-bound on Highway 101 along the California coast on the way
to a political rally. The man at the steering wheel is Walter Capps (WC), a
Professor of Religious Studies at UCSB, who was then also a Democratic
candidate for a seat in the US Congress.13 Next to him is the researcher, quietly
filming, and in the back seat is a member of Capps’ family (here named “FM”
for “family member”). Capps has just finished talking about the time he got
tenure in relation to other relevant events in his family life (e.g., when his
children were born, when his wife started to work) – a topic originally raised
by the researcher and then further pursued by FM. Then follow seven seconds
of no talk with Capps looking at the road ahead and on his left side (see
Figure 9.5). What follows the seven seconds’ pause is an example of a change
of attitude with a corresponding phenomenological modification.
(4) (“Walter Capps for Congress”; March 2, 1996; on the road from Santa
Barbara to Lompoc, California; WC ¼ Walter Capps; FM ¼ Family
Member)

WC; [. . .] I got tenure before that. . . . I think that’s when I became full
professor.
FM; Seventy-three.
WC; oh something like that.
(7 sec.)
WC; It’s a very nice ride over here.
FM; Yeah.
WC; m:[::
FM; [This is the best. I love this ride.
From the point of view of its illocutionary force (Austin 1962), that is, in terms
of what the speaker is conventionally aiming at, Capps’ utterance “It’s a very
nice ride over here” is an assertion and, more specifically, an assessment.14 If
we look at it from the point of view of its placement in a sequence of turns, as
we have learned to do from conversation analysts, this type of speech act calls
for a response by the co-participants, with the preferred response being
agreement either immediately followed by or in the form of a second assess-
ment (Pomerantz 1984; Sacks 1987; Goodwin and Goodwin 1992). FM
follows this script perfectly by first agreeing with Capps’ assessment (“Yeah”)
and then by producing a second assessment in the form of an “upgrade”
(Pomerantz 1984: 65): “This is the best.”15
Capps’ initial assessment takes as an object of evaluation the so far
unspoken-about but perceptually available sensorial environment or, rather,
makes the sensorial environment (in this case, what is seen from the driver’s
point of view – see Figure 9.5) into an object of evaluation. With his assess-
ment of the experience of the “ride,” Capps moves out of the practical
200 Intentions and their modifications

Figure 9.5 While driving, Walter Capps turns slightly to the left to look
at the side of the road

relationship with the road and the surrounding. They are no longer understood
and utilized as a means to an end, or as “affordances,” in Jerome Gibson’s
(1986) terms.16 That is, they are no longer or not only what allows Capps to
easily drive to his destination, the town of Lompoc, where he will join a state
senator and some Democratic supporters. The orientation toward such a
practical goal and the practical relationship with the equipment (e.g., the car)
and the built environment (e.g., the paved road) is in this segment momentarily
abandoned in favor of the recognition and evaluation of the experience itself,
which includes the pleasure one gets from looking at the place through which
one is driving. The road itself and what defines it as a road, including the
“nature things” on the sides (i.e., trees, bushes, the skyline), are recognized as
having a value of their own, independent of their practical utilization. For all of
this to happen, Capps first and then FM must engage in a modification of their
attitude toward their surroundings. The change from the “natural” to the
“theoretical” attitude is thus both expressed and realized (or “performed”)
through language, that is, in this case, English grammar and vocabulary.
The linguistic shape of Capps’ assessment (“It is a very nice ride over here”)
objectifies the pleasure of the ride (which otherwise could have remained a
The “theoretical attitude” 201

lived but unexpressed and private “affective attitude”) and presents it as a


potentially universal judgment that is explicitly about the here-and-now and is
only implicitly linked to the speaker’s own pleasure. The same attitude
remains in the first of FM’s following statements, namely, the upgrade “This
is the best.” The second utterance by FM, “I love this ride,” with the first
person singular pronoun I and the verb love, introduces a change in syntax and
semantics that claims an assumption of responsibility. But, even with the
personal touch, the “theoretical attitude” remains, just like when a guest says
to the host “I love to come to your house,” or someone watching a Woody
Allen movie says “I love Allen’s movies.”
These phenomena could also be described by using Bateson’s concept of
“frame.” Bateson suggested that messages such as “this is play” or “this is a
therapy session” (whether or not they are realized through language) are used
to define the relevant frame for a given interaction. They are meta-
communicative in the sense that they tell us how to communicate and more
generally how to behave in a given situation. For Bateson, certain changes are
made possible by the practice of moving in and out of frames. For example,
therapists exploit the potential ambiguity between “therapy” and “life” by
allowing patients to see that sometimes “fantasy contains truth” (Bateson
1972: 192).
We could then say that in providing an assessment of his experience while
driving, Capps has moved out of the frame established by the immediately
prior narrative account and into another, in this case, meta-frame (that is, a
higher-order frame), in which the ongoing activity of driving through a
particular part of the country is being talked about.
There are, however, subtle differences between Bateson’s and Husserl’s
theories. The notion of “frame” suggests a cognitive activity, whereas Hus-
serl’s notion of Einstellung, imperfectly but effectively translated by “atti-
tude,” evokes a way of standing or taking position with respect to our
surrounding world (Umwelt), people included. Not surprisingly, given its
intellectual sources including Bateson’s interest in cybernetics, the notion of
“frame” seems to avoid reference to the Subject of the framing and concen-
trates on the presuppositions and effects of the encoding. Husserl, instead, is
more concerned with what the perceiving-thinking Subject is achieving and the
conditions of such an achievement.

9.5.1 The theoretical attitude in talk to and by children


If the “theoretical attitude” is indeed available to all human beings and not just
to scientists involved in abstract argumentation, we should expect children to
be exposed from an early age to phenomenological modifications that involve
a shift from the “natural” to the “theoretical attitude” and we should also
202 Intentions and their modifications

expect language to play an important role in these shifts. One place to look for
cases in which children’s attention is being oriented in such a way is directives.
Transcripts of verbal interactions between children and adults in the litera-
ture on language acquisition and language socialization are full of directives, a
class of speech acts that includes “imperatives” such as English Come here!
Go there! Give it to me! Stop! (e.g., Ervin-Tripp 1976; Ervin-Tripp and
Gordon 1986; Clancy 1985; Rumsey 2003; Wootton 1997: chapter 3). In these
transcripts, parents and children are shown to be involved in “request
sequences” of varying length and complexity. By the time they go to school,
where they are instructed on how to sit, stand, talk, read, and write, young
children have had a protracted experience of being the targets of requests.
However, not all requests are the same from the point of view of the attitudes
that they instantiate. For example, a request for information like the one in (a)
requires a different kind of attitude from the request for information in (b):
a. What’s that? (Bloom 1970: 113)
b. That machine scare you? Hm? (Peters and Boggs 1986: 90)
In (a), the child is asked to focus on a given object as identified by the deictic
term that and to respond by providing a description, a requirement that can be
satisfied by the use of a linguistic expression, often a single word. In (b), the
child is also asked to focus on a particular object characterized through an
expression that contains both a deictic (that) and a common noun (machine).
But this time the child must do something different from providing a descrip-
tion. She must provide information on whether the object caused (or is causing
now, not clear because of the verb morphology) a particular emotion (fear). In
order to answer, the child must engage in an evaluation of what is going on.
Even though this does not mean that the child must have an awareness of the
reflexive quality of the act, the “attitude” – in Husserl’s sense of Einstellung
(see above) – that is required for understanding and answering (b) is different
from the “attitude” that is required for understanding and answering (a). In
addition, the request in (b) provides a model for self-reflection.17 In implying a
child’s ability to both reflect and report on her emotional state, the request
provides a model for what is reportable, in this case emotional states caused by
encounters with machines.
All of this means that over time, in being directed, children are not only
oriented to attend to objects or to perform certain actions on them, they are also
exposed to ways of reflecting on their own experience of such objects and
more generally on their life experiences. This in turn entails an ability to move
from the “natural attitude” of the here-and-now to the “theoretical attitude” of
evaluating the type of ongoing activity or the type of person that such an
activity entails or invokes. In some cases, the “theoretical attitude” appears in
situations in which children challenge or resist a particular request that is being
The “theoretical attitude” 203

made from them. It is, in other words, while they are being confrontational or
argumentative that children also practice or experiment with stepping back
from the here-and-now to provide a typification of it.
To illustrate how children can engage in this kind of behavior and support
the claim that it can be an occasion for engaging in “theoretical” action, I have
chosen to reexamine an interaction between a Samoan mother and her child
discussed by Martha Platt (1986). The example comes from a large corpus of
transcripts of family interactions originally recorded in 1978–1979 by Elinor
Ochs and Martha Platt in a (then Western) Samoan community as part of a
longitudinal study sponsored by the National Science Foundation in which
I also participated (see Duranti 1994: 20–26).
Example (5) below corresponds to example (5) in Platt (1986: 135–136).
The layout of the example has been altered to be consistent with the format of
the other examples in this chapter and the orthography for representing
Samoan speech has been slightly changed to conform to (consistent) standard
Samoan orthography. The English glosses in (5) have been left as in Platt’s
original example. A few minor revisions are suggested in the discussion that
follows. Scene: Niulala (Niu), who is three-and-a-half years old, is inside the
house with his mother (whose name is here given in the abbreviated form
“Ak”), his younger brother, Fineaso (Fine), and another small child, To`o. Ak
is in the central room of the house and Niulala and To`o are in a side room.
Throughout the whole interaction Ak remains seated while the children move
back and forth between the side and center rooms.
(5) (Western Samoa, 1979)

Ak; ko`o mai ska masi.


‘To`o bring poor me a cracker.’
Se maiga ke a`u kago vaelua
‘So that I can divide it in half.’
((To`o gives cracker to Ak; she breaks off a piece and gives the rest to To`o))
((Niulala comes out of the side room))
Ak; Niulala.
‘Niulala.’
Niu; uhh
Ak: sau mai sa`u masi.
‘Come bring me a cracker.’
Niu: leai!
‘No!’
Ak: ai se ā?
‘why?’
Niu: laga ke `ai`ū.
‘because I’m selfish.’
Ak: ke `ai`ū iā a`u?
‘Are you being selfish toward me?’
204 Intentions and their modifications

Niu: uhh.
Fine: ((laughs)) uhh.
((Niulala and Fineaso are running around the house))
Ak: se Niulala, (.) Figeaso
‘Please! Niulala, Fineaso!’
Niu: ei!
‘hey!’
Fine: o!
Ak: sau lā `oe mai ska masi ē!
‘Come here now and give me a cracker!’
The interaction starts with Ak (who is seated) asking a small child (presumably
from a neighbor family) who is in the next room, To`o, to bring her a cracker.
To`o complies and gives her the cracker, which she breaks in two, giving one
piece back to him. Then Ak makes the same request to her (older) son,
Niulala,18 who instead refuses with a straight ‘No!’ (leai). When she asks
him why (ai se ā?), her son replies: laga ke `ai`ū, which Platt, probably trying
to stay close to a gloss given by a Samoan native speaker, translates as
‘because I’m selfish.’ A more literal translation would be ‘because I am
unwilling to share food,’ as suggested by the fact that the word translated as
‘selfish’ in the example, `ai`ū, includes the verb `ai ‘eat’ (Milner 1966: 10). At
this point the mother asks ‘Are you being selfish towards me?’ (ke `ai`ü iā
a`u?) or, more literally, ‘Are you being unwilling to share food with me?’
After receiving no answer to her question and seeing that instead of complying
with her request Niulala has started to run around the house with To`o, she
uses another directive: ‘Come here now and give me a cracker’ (sau lā `oe mai
ska masi ē).
As is always the case with communicative events, there are potentially many
different ways of discussing this verbal exchange. Platt, for example, chose to
highlight the mother’s use of the affective marker /ska/ (from /siþka/19), the
combination of an affective article (/si/) and an affective first person possessive
(/ka/) (pronounced /ta/ in other contexts, see Appendix B) that is meant to
make the listener feel sorry for the speaker (see also Ochs 1986; Ochs and
Schieffelin 1989). The mother first uses the affective form /ska/ with To`o but
does not use it with her son Niulala. After Niulala refuses to give her his
cracker, she uses the /ska/ with him as well (as shown in the last line of
example 5).
Another way of analyzing this example is from the point of view of
Niulala’s defiant behavior. By bringing up a description of himself as someone
who is unwilling to share his food with his own mother, Niulala is testing his
mother’s patience and simultaneously asserting his own will. Within an inter-
action that is ostensibly about exchange and sharing, a common locus of
socialization in Samoa and elsewhere (e.g., Schieffelin 1990), we see language
The “theoretical attitude” 205

being used also to reflect on and actually constitute the type of moral persons
that participants are. Niulala’s verbal actions in example (5) are a clear case of
what Samoans themselves would characterize as “cheeky behavior” (tautala-
laitiiti) (Milner 1966: 257; Ochs 1988: 154).
As described by Ochs (1988; Ochs and Izquierdo 2010), Samoan children
are socialized (and expected) to respect the local hierarchy by being attentive
to and cooperative with older people, including their older siblings and their
parents. If they do not comply, they are subject to shame or punishment.
However, there are always exceptions to this amply documented cultural
preference for being cooperative and respectful. As recently remarked by Ochs
and Izquierdo, “adults supported occasional cheekiness, noting that later in life
untitled youth are sometimes called on to be ‘bad’ in ways unbecoming to
older titled persons” (Ochs and Izquierdo 2010: 397).
In what follows, I want to focus on what Niulala’s response entails from the
point of view of his “attitude” toward what he has being asked to do, that is, to
share his cracker with his mother. First, it should not be difficult to see that
Niulala’s explanation for his refusal to comply with the request qualifies as an
example of what Husserl called “the theoretical attitude.” In the response
reproduced here below in example (5)’, Niulala provides his mother (and
whoever else happens to be listening, e.g., the researcher who is tape recording
the interaction and taking notes) with a typification of his own behavior.
(5)’ (Two-turn exchange between Mother, Ak, and son, Niu)
Ak; ai se ā?
‘why?’
Niu; laga ke `ai`ū.
‘because I’m selfish.’
Niulala is here able to “step out” of the flow of ongoing events and the world
they constitute, which is in this case the “lifeworld” of kids running around the
house eating crackers, parents making requests for them to do something, and
kids complying or refusing to comply. Niulala’s stepping out is as good an
example of “theoretical attitude” as we will ever find.
Second, when we look at Niulala’s utterance from the point of view of his
sequential placement, we realize that it is the second pair part of an adjacency
pair consisting, grammatically, of a question and an answer (Schegloff and
Sacks 1973).
(5)’’

Ak; ai se ā? First Pair Part (QUESTION)


why?’
Niu; laga ke `ai`ū. Second Pair Part (ANSWER)
‘because I’m selfish.’
206 Intentions and their modifications

Hence, Niulala’s turn here translated as ‘because I’m selfish’ is an answer to


his mother’s why-question. This means that in fact the shift to a “theoretical
attitude” that Niulala makes here had been (just) set up by his mother, that is, it
is Ak’s why-question that lays the (interactional) ground for the phenomeno-
logical modification verbally performed by the child. This in itself is as
remarkable as, or more remarkable than, the lack of respect entailed by
Niulala’s response because Samoans have long been characterized by ethnog-
raphers (e.g., Mead 1937; Shore 1982) as not interested in motivations or
psychological explanations. When prompted for such explanations through
why-questions people often respond with the expression ta`ilo ‘who knows?’
or ‘how should I know?’, a phrase that includes the above-mentioned affect-
loaded pronoun ta (often pronounced /ka/) ‘poor me,’ which is meant to make
the listener feel sorry for the speaker. By answering ta`ilo, speakers simultan-
eously excuse themselves for not answering and discourage any further
attempt to pursue that particular kind of questioning (see Mead 1928: 123;
and §8.1 in this book).
The fact that a mother asks for a reason that would explain her child’s
refusal to satisfy her request is telling. Seen in this cultural context, Niulala’s
“cheekiness” seems less of a surprise. We could hypothesize that he can be
cheeky at least in part because his mother gives him a chance to be so.
Furthermore, from the point of view of the attitude that he embodies and
displays, his cheekiness is also evidence of cognitive and interactional abilities
that are quite sophisticated for a three-and-a-half year old.

9.6 Modifications and their role in socialization


On the basis of the previous discussion, I propose to think of socialization as
the accumulated effect of a number of recurrent modifications – in the sense
given by Husserl to this term – in the ways in which novices are expected to
relate to a particular phenomenon. These modifications give children and
adults not only the power to act in novel ways, but also to adopt, reflect upon,
question, reject, and revise what is being perceived, reported on, proposed, or
ordered.
In the world of music instruction, over time the sum of the different ways of
hearing and listening that are encouraged and modeled by the instructor and
experienced by the students become part of their musical competence and
eventually contribute to the definition of their professional identity not just as
musicians, but as musicians of a certain type, in our case jazz musicians
(Jackson 2000, 2002; Dortier 2002; Black 2008). If they succeed, they will
reach a point where they will share what we might call a “professional ear” just
like archaeologists share what Charles Goodwin (1994) called a “professional
vision.” When archaeology students are asked to identify a “change of slope”
Modifications and their role in socialization 207

in the ground surface (Goodwin 1994: 613–614), they are being socialized to
see the ground in new terms, according to a profession-specific categorization
scheme. Similarly, when jazz students are asked to identity the “interactive”
character of someone’s playing (see discussion of example 2 above), they are
also being socialized to hear music (whether recorded or live) according to a
jazz-specific category system, which includes emic paradigmatic oppositions
such as interactive vs. non-interactive or emic gradient differences, e.g., from
“more” interactive to “less” interactive. In addition, since, as shown in
example (3) above, the fact that one is “listening” must be demonstrated by
the way one plays (Black 2008), any kind of categorization scheme must be
activated and reproduced through specific actions, which in turn contribute to
the constitution of particular kinds of moral persons (Day 2000; Duranti and
Burrell 2004).
Language socialization research has shown that caregivers (of different ages
depending on the social and cultural organization of families and communities)
organize children’s attention to particular linguistic sounds, acts, and routines
(Ochs and Schieffelin 1984; Schieffelin and Ochs 1986b). Children and
novices are constantly being “oriented” through talk and other semiotic
resources toward specific tasks and specific Others, with noticeable cross-
cultural differences (e.g., Ochs and Schieffelin 1984: 297). Children are told
to call out to certain people (e.g., Demuth 1986: 58–59; Watson-Gegeo and
Gegeo 1986; Ochs 1988) or to ask certain kinds of questions of others (e.g.,
Schieffelin 1986). They are also teased and instigated to defend themselves
against present or future threats posed by imagined or real others (e.g., Eisen-
berg 1986; Schieffelin 1990). Language use, therefore, starts within a locally
available set of activities where the child is monitored and directed to move
and speak in certain ways. This means that when, later in life, children or
adults participate in new activities, they have already had considerable experi-
ence speaking, acting, thinking, and feeling in specific ways; they have, in
Bourdieu’s terms, acquired a “habitus.”20
Language socialization experts have argued that these earlier experiences
may affect a person’s ability to change in the ways required by a new environ-
ment or a new activity (e.g., the classroom, the playground, the shop) (see
Heath 1982, 1983; Kulick and Schieffelin 2004; Ochs, Solomon, and Sterponi
2005).
By relying on Husserl’s notion of modification, we can hypothesize that at
least some of the problems encountered not only by students, but also by
educators (or therapists, as in the case discussed by Ochs, Solomon, and
Sterponi 2005) in adapting to new ways of being may be due to the accumu-
lated effects of modifications experienced earlier in life which make it difficult
if not impossible to retrieve earlier, pre-modificational ways of being. As
infants, we can assume body postures that are no longer possible when we
208 Intentions and their modifications

grow older. As young children, we can develop a musical ear that is harder to
achieve later in life as shown by the positive correlation between early expos-
ure to music training and perfect pitch (Pressing 1984; Sacks 2007). After
years of socialization, the way we do things has come to feel as if it is the only
possible way (Dortier 2002: 5; Duranti and Black 2012: 446). We cannot
imagine another way of being and our mind and body just cannot “go there.”
What we have been socialized to think, feel, and do has become part of what
Husserl called “the natural attitude.” In addition to suggesting that we shoud
read Husserl’s use of “natural” as corresponding to what anthropologists call
the “cultural,” I have proposed that we adopt Husserl’s concept of modification
as an analytical tool for the documentation and theorization of how the process
of “naturalization” is realized in the lives of children and novices across
situations and communities.
10 A sense of the other: from intentionality
to intersubjectivity

10.0 Introduction
In this chapter, I examine in some detail Searle’s (1990) notion of “collective
intentionality” and the distinction he makes between I-Intentions and we-
Intentions to argue that, contrary to what is nonchalantly claimed by Searle
himself, collective intentionality is not synomymous with and cannot replace
the phenomenological notion of intersubjectivity, which is, in turn, a funda-
mental quality of all kinds of human experiences, including the apparently
most individual or private. By reexamining some of Searle’s examples and
providing some new ones taken from my own research on jazz instruction and
jazz performance, I show that, under closer scrutiny, Searle’s notion of col-
lective intentionality provides a limited and sometimes flawed understanding
of how people carry out tasks that involve coordination and collaboration with
others, including those tasks that might not seem to qualify as collective.
Searle’s realization that the notion of intentionality he presented in his
1983 book needed to be expanded to take into consideration that there is
something more than individual actions and individual intentions could be
seen as an answer to the anthropological critique of his model (see Chapter 2).
But the dichotomy he proposes between individual and collective intentions
turns out to be both theoretically and empirically problematic. In the final part
of this chapter I suggest that an alternative and better model of both individual
and collective action can be built by integrating a phenomenological under-
standing of intersubjectivity with the empirical study of how people learn to
and eventually succeed at coordinating their actions in complex joint activities
such as playing in a jazz band.
The theoretical perspective I am favoring in this chapter is built on the
assumption that the world of the individual is always a social world. We
anticipate the actions of others and need to coordinate with them on a regular
basis whether or not we are engaged in activities that are culturally recognized
as “collective” or “joint.” Conversely, even those situations that require a high
level of coordination and can qualify as collective in Searle’s sense of the term,
e.g., playing in a band, require individuals to alternate between attending to

209
210 A sense of the other

their own actions and those of others in situation- and culture-specific ways
that can result in smooth interaction.

10.1 Beyond individual intentions in human action


As discussed in earlier chapters (especially Chapters 2 and 5), in the 1980s
some anthropologists, including me, were very critical of the use of intentions
in speech act theory and of the implications of that theory for the conceptual-
ization of human action, which appeared to some of us to be too detached from
social life and from the more nuanced ways in which individuals are intercon-
nected through language and other symbolic systems. The anthropological
critique of speech act theory was based on a conceptualization of meaning
as public (Geertz 1973: 13), that is, as located and available in public events
and in the identities and relationships that those events make apparent. This
perspective on culture as a “text” constituted by publicly performed actions
was at the core of Geertz’s program for an “interpretive anthropology.” To
those of us who embraced this program it seemed that if we wanted to make
sense of speech acts such as promising, accusing, lying, apologizing, etc., we
should be looking beyond what individual speakers might have meant, known,
or understood at the time of producing those acts and focus – as well as or
primarily – on social contexts, social relations, and consequences of effects
of particular acts (see Ochs 1982; Rosaldo 1982; Rosen 1985, 1995). We
believed that a more socially oriented approach to meaning-making could help
us understand cultural differences while giving us a richer model of how
people in any society make sense of what is being done and said. All of this
is to say that anthropologists’ (and especially linguistic anthropologists’)
criticism of speech act theory in the 1980s was directed toward the major role
played by individual intentions in Searle’s model. Such critical stance was not
affected by Searle’s (1990) essay on “collective intentions,” with which he de
facto expanded his earlier model to account for group behavior and, as we shall
see, attempted to provide his own version of intersubjectivity. Unfortunately,
anthropologists who wrote about intentions after 1990 largely ignored Searle’s
1990 essay – although I suspect that some of them were not aware of it due to
the paucity of potentially useful safe grounds “trading zones” (Galison 1999)
for interdisciplinary conversations between anthropologists and analytic phil-
osophers.1 As I briefly mention in my 2004 email exchange with van Dijk,
here reproduced in Chapter 7, in the early 1990s I had considered proposing a
notion of “distributed intentionality,” which I saw as a possible alternative to
Searle’s “collective intentionality,” but I abandoned the idea because it seemed
hard to distinguish between “distributed intentionality” and the notion of
“distributed cognition” that was being discussed at the time by Jean Lave
and Ed Huchins (see below).2
Beyond individual intentions in human action 211

By the time I returned to the issue of intentions (Duranti 2006a), I had


become less interested in engaging with Searle’s writings and more interested
in understanding Husserl’s discussion of both intentionality and intersubjec-
tivity. A couple of years later, the issue of the cross-cultural relevance of
intentions in interpreting human action reappeared in the anthropological
literature within the context of the effects of Christian missionization in the
Pacific and the introduction of such practices as confession. This time, how-
ever, it was recast in terms of the universality of “theory of mind” (Robbins
and Rumsey 2008) and Searle was barely mentioned. This was, in a way,
another missed opportunity given the fact that Searle himself had just pub-
lished an article in an anthropological journal (Searle 2006), where he had
finally acknowledged – and, predictably, rejected – Rosaldo’s criticism, albeit
missing her main theoretical objections (see below).3
In this chapter I will try to remedy this collective forgetfulness and mutual
ignoring by presenting a more detailed evaluation of Searle’s notion of col-
lective intentionality and its relation to individual actions. This will also be an
opportunity for extending and putting into practice my own interpretation of
Husserl’s notion of intersubjectivity (Duranti 2010). I will argue that some of
the interactions that Searle describes as typical examples of “collective inten-
tionality” in fact have an organization that calls for a different type of
conceptualization.
As we shall see, despite Searle’s later explicit rejection of phenomeno-
logical methods (e.g., Searle 2001/2b: 282–283), he saw “collective inten-
tionality” as his own version of what phenomenologists have theorized under
the concept of “intersubjectivity,” and he admitted that we need to include “a
pre-intentional sense of ‘the other’ as an actual or potential agent like
oneself” (Searle 1990: 413). This type of statement shows that through the
concept of collective intentionality Searle was trying to address the difficulty
of accounting for “the other” in a model of the human mind that conceptual-
izes individual intentional action as the default mode of being in the world.
One problem he encountered, and did not solve, is that once we say that
collective intentionality needs others, we need to figure out where the others
are or, more appropriately, where they have been (temporality matters).
Searle accepts the existence of others organized in social groups, which
“are, so to speak, ready for action” even though “they are not yet engaged
in any actions” (Searle 1990: 414). This means that we, as individuals, have
“a sense of others as more than mere conscious others, indeed as actual or
potential members of a cooperative activity” (Searle 1990: 414). In an
unexpected choice of words that evokes Heidegger’s Being and Time, it
seems that the introduction of the notion of collective intentionality brings
Searle closer to the idea that others have always already been there (see, e.g.,
Heidegger 1962: 152; 1996: 109–110). But if that is the case, the question
212 A sense of the other

then becomes whether or how we can distinguish between clear cases of


collective intentionality and cases in which participants can be said to have a
readiness to participate, each in their own ways, in collective enterprises. An
observationally informed answer to this question must start by recognizing
the fact that both the conceptualization and the habitus of collective actions
exhibit variation. Different communities might hide or highlight the role of
certain participants, going as far as erasing the agency of individuals who
play the main role in getting something done – see, for example, the discus-
sion of how mothers in white middle-class families in the United States may
become “invisible” (Ochs 1992). This fact raises the question as to whether
we should start with individual intentions, as proposed and practiced by
Searle in his 1983 book on intentionality and by other philosophers who
think in terms of “folk psychology,” or whether we should accept that
individuals always attend to their own mind and body while also attending
to and anticipating the actions, thoughts, and feelings of others, even when
these others are not visibly there.
This chapter is dedicated to discussing these issues and proposing some
generalizations about the intersubjective side of individual intentions and the
cooperative acknowledgment of action. Even though other analytic philoso-
phers, most notably Raimo Tuomela (1984, 1989, 1991, 2007, etc.), have
written explicitly and extensively about collective intentionality, I will be here
focused on Searle’s theory. I have a number of reasons (and excuses) for this
choice. The first and least important reason is that to do a proper review of the
work of other authors would take too much space within the scope of this
book. The second is that anthropologists interested in intentionality have
directed their criticism almost exclusively toward Searle’s theory – Tuomela’s
writings, for example, are probably too formally oriented to enter the discourse
of anthropological inquiry. Third, it is Searle’s notion of collective intention-
ality, often rephrased as “shared intentionality,” that has become popular
among cognitive scientists and developmental psychologists interested in an
anthropological approach to cognition (e.g., Tomasello et al. 2005; Tomasello
2008; Colle and Grandi 2007).

10.2 Searle’s notion of “collective intentionality”


A few years after publishing Intentionality (1983), Searle decided that even
though he still believed that society must be conceptualized as an association
of individuals, each with their own separate mental realities, it was necessary
to account for the fact that individuals could coordinate their actions in
something bigger than and different from each individual action. In approach-
ing this issue, Searle acknowledged the existence of something qualitatively –
and ontologically – different from what he had previously described through
Searle’s notion of “collective intentionality” 213

his 1983 model of prior intentions and intentions in action. As shown in the
following passage, Searle had discovered “collective intentional behavior.”
It seems obvious that there really is collective intentional behavior as distinct from
individual intentional behavior. You can see this by watching a football team execute a
pass play or hear it by listening to an orchestra. Better still, you can experience it by
actually engaging in some group activity in which your own actions are a part of the
group action. (Searle 1990: 401)

To account for such phenomena, Searle introduced the concept of “collective


intentionality” (Searle 1990), which he characterized as born out of an “intui-
tion”: “Collective intentional behavior is a primitive phenomenon that cannot
be analyzed as just the summation of individual intentional behavior” (1990:
401). For Searle, this meant that we can use “we-intentions” to explain
behavior (through causation) and that we-intentions are a special kind
of intention that cannot be understood as or reduced to a collection of
“I-intentions.” When involved in a collective action, the individual intention,
expressed by something like “I am doing act A,” is considered to be dependent
upon the collective intentionality “we are doing B.” This means that “the
individual component of the collective actions plays the role of means to ends”
(Searle 1990: 410). For Searle, this means that the end causes the means. If
Jones and Smith are preparing a hollandaise sauce, they are engaged in
cooperative behavior. “Jones is stirring while Smith slowly pours the ingredi-
ents” (1990: 410).4 Causation is at work in this example in the sense that the
collective (and prior) intention “we are making the sauce” causes the individ-
ual intention-in-action “(I) mix the sauce.”
Searle has used the notion of collective intentionality for his social ontology,
arguing that we need collective intentionality for making sense of how society
is possible at all and for clarifying what makes it into what it is (Searle 1995,
2001/2a: 179). Thus, for example, his definition of “social fact” requires
collective intentionality of two or more human or animal agents.
Collective intentional action is especially important in any theory of society. In such
cases, I am doing something only as part of our doing something. For example, I am
playing the violin part as part of our playing the symphony. I am pitching the ball as
part of our playing a baseball game. Collective intentionality is the intentionality
that is shared by different people, and just as there can be shared intentions to do
things, so there can be shared beliefs and shared desires. The church congregation,
for example, reciting the Nicene Creed, is expressing a shared belief, a
common faith. (Searle 2006: 16)

It would be hard to disagree with Searle that as members of society individuals


engage in actions that are part of collective endeavors. In a sense, society itself
is a collective endeavor and Searle (1995, 2010) acknowledged this by using
the notion of collective intentionality for his theory of social institutions. But,
214 A sense of the other

as discussed below, we might have to disagree with the criteria by which


Searle distinguishes between individual and collective behavior or with the
weight that he gives to collaboration only under special circumstances.
It should be noted that by the time Searle’s chapter on collective intention-
ality came out (in an edited volume dedicated to “Intentions in communi-
cation”), the notion of “we-intentions” was not new. It had been discussed in
great detail, for example, by Raimo Tuomela (1984), who defined social action
as “a joint action performed by several agents who suitably relate their
individual actions to the other’s actions in pursuing some joint goal or in
following some common rules, practices or the like” (Tuomela 1984: 12).
Even though there was no mention of Tuomela’s book, Searle’s 1990 essay did
include a discussion of a later article by Tuomela and Miller (1988) explicitly
entitled “We-intentions,” which Searle called “the best [he had] seen” but
criticized for keeping the I-intentions as components of the we-intentions
instead of considering them primitive and thus unanalyzable notions (1990:
404). Other authors in the 1990 book where Searle published his essay on
collective intentionality also included discussions of group actions. For
example, Grosz and Sidner (1990: 433–435) provided an explicit “shared
plan” for the joint activity of lifting a piano (see also Hobbs 1990). Beyond
the conceptual differences, the existence of this literature suggests that by the
early 1980s analytic philosophers had become aware of some problems in their
previous conceptualization of human action and were trying to address them
with some extensions of earlier models. The notion of “we-intentions” was one
proposed solution, even though it was presented as solidly grounded within
“conceptual individualism.” In proposing the notion of “we-intentions,”
Tuomela was quite explicit: “Our view is individualistic in the sense that we
accept a principle of conceptual individualism which entails a partial reduc-
tion of holistic social concepts to individualist ones” (Tuomela 1984: 17;
emphasis in the original). This theoretical position is also made clear by Searle
in a number of statements, including: “society consists entirely of individuals
and no facts about any individual mental contents guarantee the existence of any
other individuals” (1990: 406), “collective intentionality, and therefore . . .
collective behavior, must be consistent with our overall ontology and metaphys-
ics of the world, an ontology and metaphysics based on the existence of
individual human beings as the repositories of all intentionality, whether indi-
vidual or collective” (Searle 1990: 407), and “we are not required to suppose
that there is any element in society other than individuals” (1990: 407).5
Conceptual individualism is also implied by Searle’s flat-out rejection of
concepts like “the collective unconscious” (1990: 404) and “group mind or
group consciousness” (1990: 407). At the same time, Searle as well as
Tuomela and Miller recognize the need to presuppose not only the existence
but also the agency of others. This is stated in clear terms by Tuomela and
Searle’s notion of “collective intentionality” 215

Miller, who accept the role of the other in most actions, that is, even in actions
performed by single agents:
Most of our actions are social in the wide sense that they conceptually presuppose the
existence of other agents and of various social institutions. Of actions that are social in
this sense, some are performed by single agents. The rest are either performed jointly by
several agents or performed by collective agents. (Tuomela and Miller 1988: 369)
For Searle, it is “collective behavior” (as opposed to individual behavior) that
calls for recognition of what he calls “a sense of ‘the other’.” This experience
of the other is “pre-intentional,” which I understand as meaning a type of
experience that is not subject to (what he calls) “representations.”
[I]t seems to me that the capacity to engage in collective behavior requires something
like a pre-intentional sense of “the other” as an actual or potential agent like oneself in
cooperative activities. (Searle 1990: 413)
Here Searle seems to be saying that to be able to participate in collective
intentionality we need to be potentially connected. The use of the term “pre-
intentional” also suggests that this connection would be made possible by a
way of being with others that is unsaid, non-conscious, and undetected by our
rational thought. The question is: Where is this region of being? Where is
human connectedness recognized but not (yet) cognized? Searle’s answer is
“the Background.” As we saw in Chapter 2 (§2.1.4), the Background is the
conceptual as well as practical knowledge acquired by individuals as members
of a given society. The Background is what allows us “to stir” (when we are
making a sauce with someone else) or “play football” (Searle 1990: 413).

10.2.1 The “Background”and the “Horizon”


Evoked as a source of wisdom and expertise for both individual intentionality
(in Searle 1983) and collective intentionality (in Searle 1990), the role of the
Background in Searle’s theories seemed to grow over the years in importance
and scope. It started out as a label for the cultural and practical knowledge
necessary to be able to function (or survive) in a given community – something
akin to what Alfred Schutz (1967: 78) called “the stock of knowledge at hand.”
It also covered something that can only be made explicit a posteriori, once it is
activated in some particularly obvious way or is reflected upon. Despite
Searle’s wish to distinguish himself from phenomenologists, the Background
resembled quite closely what Husserl (1931: 92) called the “horizon” of our
interactions,6 a spatial and temporal context of potentialities, which are dor-
mant until something attracts our attention or we change our point of view –
biophysically or mentally – and become suddenly interested in something that
we had barely noticed up to that point. It could be a tree, a house, a book, or a
216 A sense of the other

fellow human being. These are “all real things which at any given moment are
anticipated together or cogiven only in the background” (Husserl 1973: 33)
and include “human and animal subjects as subjects in the world . . . products
of culture, useful things, works of art, and the like” (Husserl 1973: 34). This is
the domain of “passive synthesis” (Husserl 1973, 2001), that is, a wide range
of experiences in which consciousness is absorbing information and perceiv-
ing entities in the surrounding world (Umwelt) but is not actively attending to
them. It is in this experiential domain that one would expect to find those who
are “potential members of a cooperative activity” (Searle 1990: 414).
Given such a manifest interest in the presence of others in our everyday
world, it is at first difficult to explain Searle’s avoidance in the 1990 essay of
the vast phenomenological literature on intersubjectivity. One justification for
such a noticeable absence, however, was made explicit sixteen years later in a
terse statement, in which Searle revealed that he had found the term “inter-
subjectivity” and the existing literature unclear and therefore of “no use.”

I have never seen a clear explanation of the concept of intersubjectivity, and I will have
no use for the notion. But I will use “collective intentionality” to try to describe the
intentionalistic component of society; and I suspect that if intersubjectivity is a legitim-
ate notion at all, it must amount to collective intentionality. (Searle 2006: 16)

Over the years that followed his 1990 introduction of the concept of collective
intentionality, Searle continued to criticize phenomenological approaches
(Searle 2000), discouraged any phenomenological reading of his work,7 and
went as far as explicitly denouncing some of “the weaknesses of
phenomenology” in his “Reply to Dreyfus” (Searle 2001/2b: 282). This was
unfortunate not only because Searle (and his many readers) could have actually
benefited from an explicit engagement with prior work on intersubjectivity, but
also, and perhaps more importantly, because those who found inspiration in his
concept of collective intentionality took it as equivalent to intersubjectivity,
even though it was not (see below and Duranti 2010).
Searle’s dismissal of phenomenology may be a casualty of the long, unsolv-
able, and unwinnable debate with his Berkeley colleague Herbert Dreyfus (see
Dreyfus 1991, 1993, 2001/2; Searle 2000).8 In an apparent attempt to bring
closure to what he perceived as unjustified attacks by Dreyfus,9 Searle claimed
that Dreyfus’ interpretation of Husserl, Heidegger, and Merleau-Ponty had
shown that “they do not have much to contribute to the topics of the logical
structure of intentionality or the logical structure of social and institutional
reality” (Searle 2001/2b: 284). This is a preposterous and defensive assertion
that is hard to make sense of. As already shown in Chapter 9 and in the few
among the many available quotes from Husserl’s discussion of the “horizon,”
phenomenologists have provided innumerable insights into all the issues that
Searle is trying to address, including, of course, the relationship between Self
Searle’s notion of “collective intentionality” 217

and Others. I will give some examples in the rest of this chapter that show how
one might learn about collective action by combining phenomenology with the
empirical investigation of spontaneous human interaction. But first we need to
get a better sense of the ways in which Searle distinguishes between individual
and collective intentions. To this aim, I will here examine the most detailed
example of collective intentionality that Searle provides in the 1990 essay.
I will show that, by leaving out key dimensions of what it means to be together
in a public space, Searle misses the subtle distinction between intersubjectivity
and planned behavior.

10.2.2 People in the park


To explain his notion of collective intentionality, Searle (1990) presented a
contrast between what he saw as two sets of behaviorally similar but inten-
tionally distinct acts involving groups of people, with one of them illustrating
individual behavior and the second illustrating collective behavior.
Imagine that a group of people are sitting on the grass in various places in a park.
Imagine that it suddenly starts to rain and they all get up and run to a common, centrally
located shelter. Each person has the intention expressed by the sentence “I am running
to the shelter.” But for each person, we may suppose that his or her intention is entirely
independent of the intentions and behavior of others. In this case there is no collective
behavior; there is just a sequence of individual acts that happen to converge on a
common goal. Now imagine a case where a group of people in a park converge on a
common point as a piece of collective behavior. Imagine that they are part of an outdoor
ballet where the choreography calls for the entire corps de ballet to converge on a
common point. We can even imagine that the external bodily movements are indistin-
guishable in the two cases; the people running for shelter make the same types of bodily
movements as the ballet dancers. Externally observed, the two cases are indistinguish-
able, but they are clearly different internally. What exactly is the difference? (Searle
1990: 402–403)

The answer given by Searle is that in the first case each person has an intention
that can be expressed without reference to the other whereas in the second case
each individual “I intend” is derivative of a “we-intend” and therefore only in
the second case do we have collective behavior and collective intentionality.
On close examination, we discover that there are some aspects of the two
sets of actions that complicate the distinction drawn by Searle and at the same
time raise some questions about the conceptualization of what constitutes
“collective behavior.”
There is no question that the example is meant to illustrate in a simple and
unambiguous way the difference between the behavior of a group of people
who simply happen to be in the park at the same time and a group that shares a
reason (or motive) to be there, whose members had a chance to practice their
218 A sense of the other

moves together, might have been given some instructions by a choreographer,


and thus could be said to have a “plan of action” (even though Searle does not
mention such a plan10).
I will ignore the assumption that we can expect that “the people running for
shelter make the same types of bodily movements as the battle dancers.” This
is a questionable assumption given the fact that dancers behave in very distinct
ways and are typically recognized, especially but not only when they dance, as
more graceful, agile, athletic, etc. than most people. I want instead to focus on
the fact that, regardless of their previous preparation or goals, the people in
both (hypothetical) groups who are converging toward a particular point in the
park must engage in a certain level of collaboration and coordination with each
other while monitoring their action with respect to the natural environment
(e.g., ground, grass, trees) and the built environment (e.g., paths, fences,
benches, sidewalks). These considerations raise the issue of whether in fact
the difference between the two groups who seem to be doing “the same action”
could be better accounted for by postulating a continuum from “planned to
improvised” that includes a range of possible or likely acts within a given
repertoire of acts.11 Even though the people in the first group – many of whom
are strangers to each other – might not have been prepared for running to the
shelter (e.g., some of them might have never done it before), they still have to
behave in culturally appropriate ways, following, for example, the ethics and
aesthetics of public behavior described, among others, by Erving Goffman
(1959, 1963). Thus, in a park in the US we might expect people of different
ages to run in different ways and at different speeds. We might also expect
some old people to be helped by others, in some cases by strangers moved by
empathetic feelings, altruism, or by the politesse du coeur discussed by Henry
Bergson (2008) (see Duranti 2009a). We are also likely to expect that espe-
cially adults would be careful not to bump against anyone else or make it too
manifest that they are eager to beat to the shelter the unfortunate ones who, for
whatever reasons, are having a hard time being as fast as the rest.

10.2.3 The paradox of collective action: more planning,


less pro-social behavior
These reflections about the need to coordinate one’s action while facing an
unexpected situation suggest a reversal of the situation as imagined by Searle
given that planning how to dance in the rain is in fact a way to reduce the
range of possible collaborative actions that might be needed in achieving our
goal. A dancer who is part of a choreographed act does not have to worry
about looking selfish for running ahead of others or impolite for jumping
over the body of someone else without asking for permission. In even more
dramatic ways, in competitive games like football, players are given license
Socially distributed cognition and extended minds 219

to perform all kinds of acts, including cutting someone off or knocking him
down, which under different circumstances would be considered so aggres-
sive as to warrant restraint and even police intervention. Searle seems to take
all of this for granted by concentrating on the collaboration among members
of the same team, but the people who confront each other on the field during
a game adopt and display attitudes that violate all kinds of rules of etiquette.
The collaboration in the in-group is countered by the systematic lack of
collaboration toward the out-group, all of which is made possible by the
rules of the game.
We might then have identified a paradox in one of the three basic concepts
of social-institutional reality (Searle 2006: 16). The notion of collective inten-
tionality is designed to capture what ties people to institutions and allows them
to be involved in joint activities.12 But a closer examination of the situations in
which such “shared intentionality” is supposed to occur shows that it produces
a type of person with strong affiliation only with the goal of certain others.
What is an increase of sociality toward some members turns out to be a
reduction of sociality toward others. The “we” of the members of a team is
extentionally different from the “we” of the two teams who are playing on the
field. The members of a team use an “inclusive-we” when talking to the
members of their team but an “exclusive-we” (meaning “you are not included
in our ‘we’”) when talking with the members of the other team.
Seen in this way, in introducing the notion of collective intentionality Searle
(1990) avoids discussing a problem that was debated among phenomenologists
half a century earlier, when Sartre (1943) objected to both Husserl’s and
Heidegger’s views of intersubjectivity because he saw the we-relationship as
inherently (or essentially) confrontational and conflictual and thus something
more than taking the point of view of others (Husserl’s position) or always
already being-with or being-along others (Heidegger’s position) (see Zahavi
2001b for an assessment of these positions).

10.3 Socially distributed cognition and extended minds


A different line of research that is particularly relevant to the issue of taking
others into consideration in our mental and physical activities is the study of
socially distributed cognition, which focuses on everyday cognition (e.g.,
Rogoff and Lave 1984) and has historical antecedents in a number of classic
studies, including Vygotsky’s (1978) groundbreaking essays on the social,
interactional (or interpsychological) origins of individual (or intrapsychologi-
cal) higher mental functions such as the use of symbols (Wertsch 1985). In his
study of how people in a group coordinate their actions in order to navigate,
Hutchins (1995) found inspiration in Vygotsky’s studies as well as in Durk-
heim’s discussion of the division of labor in society.
220 A sense of the other

Because society has a different architecture and different communication properties


than the individual mind, it is possible that there are interpsychological [i.e., from one
mind to another] functions that can never be internalized by any individual. The
distribution of knowledge described [in this book] is a property of the navigation team,
and there are processes that are enabled by that distribution that can never be internal-
ized by a single individual. (Hutchins 1995: 284)

One implication of this approach is that in coordinating our actions with


others, as, for example, when we are engaged in a joint activity, we do not
have to hold everything in our mind in order to execute a given task or engage
in a given activity because others will – and in fact often do – provide us with
the necessary information and actions that are needed for the task to be
completed. Distributed cognition recognizes the fact that we are constantly
acting in the world along and with entities that not only provide us with
otherwise missing information but also guide us toward certain actions and
interactions. Such entities are other human beings or other elements of our
surroundings. This perspective is particularly relevant to envisioning a theory
of learning that builds on the study of the interaction between novices and
experts (e.g., Greenfield 1984; Rogoff and Gardner 1984; Lave and Wenger
1991).
Socially distributed cognition is a model of human behavior that extends
individual psychology into the world of human interaction to explore the role
that other people (e.g., teachers, experts, team members, classmates, siblings)
and tools (e.g., navigation equipment, computers, paper and pencil) have in
amplifying human knowledge and human abilities to accomplish all kinds of
cognitive tasks (e.g., Boesch 2007). The simplest example is the use of the
fingers to do arithmetic operations that could also be done in the head. Other
times it could be a grid of squares drawn on the ground, as in the game of
hopscotch (M. H. Goodwin 1998; C. Goodwin 2000).
Andy Clark and David Chalmers (1998) proposed a similar approach, and
they called it “active externalism” (see also Clark 2011). The idea is that
cognition can be “extended” through external, material entities, including our
fingers and feet, natural objects like rocks and trees, and cultural objects like
tools of all kinds. Any of these external objects “functions as a process which,
were it done in the head, we would have no hesitation in recognizing as part of
the cognitive process” (Clark and Chalmers 2011: 222, emphasis in the
original).13
The world is full of what Jerome Gibson (1986) called “affordances,” that
is, properties of the environment that favor certain behavior over others (e.g.,
water affords swimming, hard soil affords walking). The culturally organized
and humanly inhabited world, what Husserl (1970a) called the “lifeworld”
(Lebenswelt), does not just contain our actions, it makes them possible and
even guides them. Our daily decisions are not only based on habitual ways of
Following rules 221

doing things, but also on the possibilities made available by the presence of
people and things in the environment we inhabit. On most days I do not need
to remember the rule about driving on the right side because it comes
“naturally” to me to stay on the side of the road with the cars that go in
the same direction. I just need to concentrate on not crossing the visible line
in the middle of the road and avoid the cars coming toward me, which are
conveniently moving along the other side of the road. My “driving on the
right” is everyone else’s, as much as my, accomplishment. It is also made
possible and guided by the shape of the road, the white lines marking the
lanes, and the actions of the other motorists in and by means of their vehicles,
which are built in such a way as to allow for certain courses of actions and
reactions in time and space.
One way in which anthropology becomes particularly relevant in all of this
is that the people involved in the activities in question (and their communities
at large) may not recognize such a distribution of knowledge and action. For
example, it is not uncommon around the world that the contribution of certain
people will be ignored in certain contexts and therefore what is de facto a
collaboration (e.g., between a politician and his or her team of writers) turns
out to be seen as the work and achievement of only one individual.14 But we
do not need to go too far to find representations of activities in which
individuals do take into consideration, anticipate, or adjust to the actions of
others and in so doing participate in a collectivity and give evidence of such
participation.

10.4 Following rules


Despite Searle’s insistence that he is concerned with the “logic” and not with
the “phenomenology” of the action (Searle 2001/2b), there has to be some way
of accounting for fixing the errors that people make as they are trying to
execute a given plan, whether individual or collective. Unfortunately, Searle
seems to have interpreted “phenomenology” exclusively as “first person
description” that cannot explain rules because it is only concerned with “me
here and now in the immediate present” (2001/2b: 282). But this is both a
misunderstanding and a misrepresentation of phenomenology. It ignores the
fact that the phenomenological notion of intersubjectivity was introduced by
Husserl precisely to counter a solipsistic view of human consciousness and
human experience (e.g., Husserl 1960) and focus on the necessary anticipation
of the perception, actions, feelings, etc. of others for the individual subject to
be able to function at all. The conceptualization of collective intentions as
representing “ideal” or “idealized” behaviors – what Searle seems to be
referring to as “the logical structure” (see above) – cannot be an excuse for a
poor model of what people actually do or how they manage their daily
222 A sense of the other

interactions unless we are operating at such a general and abstract level that
details never matter. Much of the debate between Searle and Dreyfus has been
about the notion of causation (i.e., do intentions cause behavior?) and about
whether we need to have a rule in our mind (regardless of whether we think of
it) when we behave in a certain way. Do we need to have the rule “drive on the
right side” when we are in the US, France, or Italy but have a different rule,
that is, “drive on the left side,” when we are in the UK or Australia? Dreyfus,
who follows Heidegger’s description of tool-use, argues that “the rule” sur-
faces in human consciousness or talk only when our routine course of action is
interrupted or put in jeopardy. Thus, Americans would think about the rule
“drive on the left” when they go to the UK or Australia and, vice versa, people
who are used to driving in the UK or Australia would have to think about the
rule “drive on the right,” at least for the first few days.
Both Searle and Dreyfus seem to forget that rules do not exist in a vacuum
or only in the mind. When one is in London or Sydney, it is common to find
reminders on the pavement to look right when crossing the street. The fact
that there are no such reminders in Los Angeles or Rome is instructive of an
asymmetry in intersubjective stance: in some places people try to imagine
what a stranger is likely to do and show some concern about it (since when,
how, and why are all interesting questions for a potentially interesting
project).
No matter how important, the focus on rules and rule-following – a favorite
target of criticism by Wittgenstein (1958) – has in fact distracted us from at
least one other fundamental and equally important property of human thinking,
feeling, and acting in the world, namely, our flexibility and adaptability to new
situations (see the discussion of the notion of modification in Chapter 9). In the
next section, I will tackle these issues from the point of view of spontaneity
and a related dimension of all human action: improvisation.

10.5 Spontaneity and improvisation


When we reflect upon the flow of interactions that characterizes everyday life,
we realize that our mind and body are typically pulled (or encouraged to move)
toward next-actions that we did not necessarily anticipate when we were
thinking about doing something, even in cases in which we explicitly planned
our actions. This includes fairly complex new activities like when we realize
that we have to go into the left lane because there is a fallen tree that blocks the
right lane. More often we have less time to “think” and quickly adapt in an
apparently flawless way. This is the case when we lean over to grab the arm of
someone who seems to be falling, when we avoid a car that is in the next lane
and suddenly appears to be getting too close to ours, or when we reformulate
Spontaneity and improvisation 223

what we are saying in order to make it relevant to someone who is paying


attention instead of the original recipient (Goodwin 1979, 1981). What makes
these quick reactions possible?

10.5.1 Naturalizing intentions


One type of answer has been pursued by an increasing number of neuroscien-
tists who, by working on mirror neurons, found that we have an “immediate
understanding of the acts of others” (Rizzolati and Sinigaglia 2006: 153), that
is, we often can predict what others are about to do in a pre-reflective way
(Iacoboni 2008: 265). When we see someone doing something, e.g., reaching
for something or, more specifically, grasping-to-carry-the-object-to-the-mouth
(Iacoboni 2008; Iacoboni et al. 2005; Rizzolati and Sinigaglia 2006: 130), our
mirror neurons fire automatically as if we were doing the same action. This
neural reaction suggests that there is a certain level or type of understanding
that is immediate, pre-reflective, and pre-intentional (see above). A similar
type of phenomenon is our empathy for the pain of others, which has been
shown to correspond to a very rapid activation of certain parts of the brain
(e.g., Singer et al. 2004; Avenati et al. 2005; Gu and Han 2007). We respond
with empathy to certain stimuli before we can “think” about what we are
seeing and hearing. Such neural responses, however, have been shown to be
sensitive to in-group vs. out-group distinctions thereby suggesting that culture
and biology interact at the most minute level of our organism and at an
amazing speed (e.g., Sheng et al. 2013).

10.5.2 Embodied intentions: jazz improvisation


There are many situations in which, even though some of our actions would
not appear to be caused by “prior intentions” (see Chapter 2), they feel quite
“natural” to the people who perform them, as if following an unspoken plan
of action that is hidden to them until the moment they find themselves
doing it.
We know that people are able to routinely adjust to the context of an
interaction “on the spot.” Over three decades ago, for example, Charles
Goodwin (1979, 1981) showed that speakers readjust the illocutionary force
of their speech within the same utterance to fit the knowledge that their
recipient has of what is about to be said. At a more general level, most of
our speech during the day is “improvised,” that is, produced on the spot and
according to needs that we read from the situation. In having a phone conver-
sation, teaching a course, answering a question by a colleague, giving an order
to a waiter, we have access to routine ways of doing things that we do
224 A sense of the other

“spontaneously” and most of the time appropriately and flawlessly. How do we


do it?
Some insights on doing something that was not planned and yet fits with the
situation without interrupting the flow of interaction come from people whose
profession is dedicated to developing improvisational skills and using them in
front of an audience, where their actions are closely watched, heard, and
evaluated. Several of the guest musicians who participated in the course
“The Culture of Jazz Aesthetics” that I have been teaching with Kenny
Burrell at UCLA, for example, equated improvisation with doing something
that is novel and thus could not have been “planned” or “intended.” This
discovery process is invoked by the drummer Sherman Ferguson in the
following excerpt, where he describes his reactions a few minutes earlier while
listening to pianist Gerald Wiggins’ solo performance of Body and Soul in the
context of our class.
(6) (Course “The Culture of Jazz Aesthetics,” UCLA, October 22, 2002)

Ferguson; I think when [Wiggins] first started playing [. . .] it sounded like


he wasn’t exactly sure which choice he was going to make, as
far as the tune he was going to play. And then when he- when
we recognized that he had decided to play, and I’ll tell you the
s- name of the song later, that was a surprise, ah oh::!, it’s going
to “ba ba ba”! And then he played some quotes [. . .] and you
know that that didn’t have anything to do with the general song,
but it was his take on it, wha- what he felt like doing. And that’s
another one of the joys in playing music and in playing jazz in
particular is that sometimes you surprise yourself because
you’re not sure what you’re gonna do either. And to me, that’s
one of the ultimate things that happen when you surprise
yourself and that’s- and just the joy of hearing him play after all
these years, he still does that to me, even when we’re playing
[together].15

The sense of surprise described by Ferguson is common among artists embrac-


ing improvisation, regardless of the art form. For example, surprise is what
Curtis Carter sees as necessary in improvisational dance, which otherwise
“runs the risk of falling into habitual repetitive patterns that may become stale
for both performers and viewers. In improvisational dance the performer must
generate a constant flow of ideas and models, and constantly surprise himself
or herself, as well as the viewers” (Carter 2000: 182). Where does this
“constant flow of ideas” come from? When asked about it, jazz musicians –
like other artists – describe the spontaneity of their actions in terms of an
external source of inspiration, a will that comes from somewhere else. Here is
saxophonist Jeff Clayton explaining this concept to the students in Burrell’s
and my class in 2006:
Spontaneity and improvisation 225

(7) (Course “The Culture of Jazz Aesthetics,” April 25, 2006)

Clayton; [. . .] a famous jazz drummer who passed away by the name of


Billy Higgins we use to sit down and talk about . . . where jazz
comes from and he believed jazz was a vessel and that the music
was in this caldron or this vessel and that [. . .] what we did in
practice was to learn to identify the music that was there and then
we’re supposed to go to that place in this vessel and while we’re
there we’re able to identify that jazz comes through that vessel
through us and to you. He didn’t believe that we owned it or that
we had control of it we’re just merely identifying what was in the
universe. [. . .]

This understanding of the creative process locates agency outside of the


individual mind and sees it as independent of individual intentionality or will.
Here is pianist Mike Melvoin being very explicit about this aspect of the
creative process by distinguishing between the conscious, planned behavior
of practice and the spontaneous moment of performing.
(8) (Course “The Culture of Jazz Aesthetics,” April 25, 2006)

Melvoin; When I play I try to get will and intention completely out of
the picture. Will and intention is what we’re practicing. Like to
learn how you play your ax and you ab- you absorb certain
intellectual things and you work on them and apply your mind
and intend to do certain things, but the performance,
particularly of improvisation. . . . is . . . uh . . . thoughtless. And
should be.
Mihaly Csikszentmihalyi (1990) characterized such creative moments as a
“flow,” a concept that Keith Sawyer (2001) used to explain what makes
possible certain types of musical coordination among jazz musicians. The goal
of the artists, in this case musicians, is to reach a particular type of relation
between the body (e.g., the hands playing, the eyes looking at the written
music or the other musicians) and the mind (which is processing the infor-
mation and anticipating what is coming up next) that is reminiscent of a
number of religious practices including Zen Buddhism or of physical activities
related to religious practices like yoga.
As pointed out by Jean-François Dortier, Bourdieu’s notion of habitus can
help us understand the origin of such “spontaneity”:
The habitus is in the first place the product of an apprenticeship that has become
unconscious and is understood therefore as a seemingly natural attitude freely
evolving in a given context. Thus, musicians can freely improvise at the piano only
after having spent a long time practicing their scales, acquiring the rules of compo-
sition and harmony. It is only after having internalized musical codes and constraints
(the “structured structures”) that a pianist can then compose, create, invent, and
226 A sense of the other

transmit her music (the “structuring structures”). Authors, composers, artists thus
live their creations as due to a freedom to create, to pure inspiration, because they
are no longer aware of the codes and the styles that they have deeply internalized.
Such is the case for music as well as, in general, for language, writing, and thinking.
We believe them all to be free and disembodied, whereas they are the product of
deeply rooted constraints and structures. (Dortier 2002: 5, translation from Duranti
and Black 2012: 446)
Like athletes and other people who routinely train their body to do certain
things, musicians over time become more aware than most people of the fact
that their body (e.g., their fingers, mouth) typically needs to act before they can
think about what to do next. They train their body to move according to a
certain pattern in order to “feel” the rhythm (e.g., see Phillips-Silver and
Trainor 2007). Professional musicians make this knowledge explicit when
they teach younger or less experienced players. In the Latin Jazz Ensemble
class I videorecorded at UCLA, especially at the beginning of the academic
year, the instructor-bandleader Bobby Rodriguez wanted to get the students to
play a given passage with a distinct “Latin feel,” sometimes consisting of a
particular way of accenting, sustaining, or suddenly terminating certain notes
or passages and particular ways of playing “along” other players so as to create
sonically interesting contrasts. One of his teaching strategies was to ask the
students a question that identified or implied a problem in the tempo, rhythm,
or coordination among players: “Do you tap your foot when you play?” Other
times he tried to get the students to appreciate the importance of a particular
routine practice by engaging in storytelling. In (9) below, Rodriguez frames his
recommendation to get the body into the right tempo within a mini-narrative
about playing with some “hip cats” who were reading the score and tapping
their feet at the same time. The practical implication here is: “if hip guys do it,
you can do it too.”
(9) (Latin Jazz Ensemble Course, UCLA, February 12, 2004; Bobby Rodriguez
instructor)

Rodriguez; I was sitting in a room the other day with some really like hip
cats . . . and every one of them that was reading was going
like this ((taps his right foot in tempo lifting front of foot four
times then starts to talk but keeps tapping with the same
tempo)) I don’t know why . . . younger people don’t think this
is like important because it IS . . . this is very important to
just get your ((as shown in Figure 10.1, the movements of
his left arm and left foot complete the utterance by
demonstrating what he means)) and this ((points with left
hand down toward the left foot that is still tapping)) has
nothing to do with ((lifts up his hand and brings trumpet up in
front of his mouth as if he was going to start playing)) what’s
Spontaneity and improvisation 227

going on up here ((pretending to play)) baduhdo ((taps one


more beat))
BUP!
[
((last tap accompanied by simultaneous lowering of both arms
holding the trumpet, in tempo))

One of Rodriguez’ teaching strategies is to get the students to acquire the sense
of timing by making them act it out. The implicit theory here is that the
embodiment of a collectively shared steady tempo will make it possible for
the mouth or fingers of each player to be part of the ensemble, even though
everyone is doing something different.
This type of demonstration provides us with a model-in-theory (as articu-
lated by the instructor) and a model-in-practice (as demonstrated by the
instructor or one of the more experienced players) of how, in participating in
a collective activity, individual players are expected to enter a particular type
of mental and embodied zone where certain tasks (or subgoals) can become
“automatic” and thus feel effortless (see the above quote from Dortier 2002).

Figure 10.1 Dr. Bobby Rodriguez, trumpet player, conductor and instructor,
demonstrates with left hand and left foot how to keep time while playing –
see example (9)
228 A sense of the other

The foot tapping is a continuous reminder of the collective activity. It


ensures that a player is doing something that others are also doing. In the case
we have seen, this happens in front of everyone else. But to become “natural,”
effortless, it must be practiced all the time, including when players are on their
own. Without a drummer and other rhythm instruments to guide them, players
might resort to the use of an external device that maintains a constant pulse. At
home or in the practice rooms the metronome plays this important role. Its
ticking is a mechanical, audible “stand-in” for the collective action. It extends
the mind (Clark 2011) and simultaneously constitutes a reminder of what it
might be like to play along with others. Through the combination of tool-use,
embodiment, and memory, we are enmeshed in a social world even when we
are alone. The collective, then, not only exists when the musicians are with the
rest of the band, but is also present when they are practicing alone, even though
at different levels of awareness and ability.

10.6 Husserl’s theory of perception as anticipation of the viewing


of an Other
Phenomenologists have taught us that the transcendence of the other, a separ-
ate being with whom I can empathize but whose mind I cannot directly
penetrate, participates in a foundational way in the very constitution of our
everyday perception and thus makes objective reality possible (Husserl 1960:
107). This is made clear in the examples that Husserl gives of apparently
simple acts of perception, e.g., looking at a table.
When we view the table, we view it from some particular side, and this side is thereby
what is genuinely seen. Yet the table has still other sides. It has a non-visible back side,
it has a non-visible interior; and these are actually indexes for a variety of sides, a
variety of complexes of possible visibility. That is a very curious situation peculiar to
the very essence of the matter at hand. For proper to the very sense of every perception
is perception’s perceived object as its objective sense, that is, this thing, the table that is
seen. But this thing is not [merely] the side genuinely seen in this moment; rather [. . .]
the thing is precisely the full-thing that has still other sides, sides that are not brought to
genuine perception in this perception, but that would be brought to genuine perception
in other perceptions. (Husserl 2001: 40)
The perception of the table, or of any other physical object in the world, must
be understood within a “perceptual field,” that is, an internal and external
horizon of possible perceptions, within a “field of things” (Husserl 1970a:
162). It is the variety of perceptions in the perceptual field that constitutes the
object as such, that is, an object endowed with meaning. Such variety, which
we experience, exists within a communal world, that is, while we are in contact
with other people. I can go beyond myself – my inner self – and thus have
Husserl’s theory of perception 229

experiences of the world because there are others out there who look and seem
to act like me but have distinct lives. They can and do perceive the same object
I perceive from different sides. And I can see the object from their point of
view if I go in their place. This possibility of “trading places” (Husserl 1989:
177) simultaneously recognizes their transcendence and the sameness of the
object through time and space:
The things posited by others are also mine: in empathy I participate in the other’s
positing. E.g., I identify the thing I have over and against me in the mode of appearance
α with the thing posited by the other in the mode of appearance β. To this belongs the
possibility of substitution by means of trading places. Each person has, at the same
place in space, “the same” appearances of the same things – if, as we might suppose, all
have the same sensibility. And on this account, even the “view” of a thing is Objecti-
fied. Each person has, from the same place in space and with the same lighting, the
same view of, for example, a landscape. But never can the other, at exactly the same
time as me (in the originary content of lived experience attributed to him) have the exact
same appearance as I have. My appearances belong to me, his to him. (Husserl 1989:
177; emphasis mine)

As shown by the last two quotes, Husserl’s understanding of perception of


external objects is quite different from Searle’s analysis of perception in his
Intentionality book, where perception is characterized by an intentional act
with a direction-of-fit from the world to the mind, that is, an act in which the
mind adapts to the outside world. In Husserl’s view, in perceiving we are also,
already, anticipating the view of an Other and thus implicitly evoking “the
sense of the other.” Despite Searle’s use of the notion of the Background, it is
not clear whether for him the “sense of the other” would be evoked when an
individual is seeing a table, a flower, or another person. This also means that it
is not clear how, where, and when the “sense of the other” could manifest itself
in order to make possible Searle’s “collective intentionality.” For Husserl, on
the other hand, the object (of perception) per se is not any one’s perception,
then, but “a unity for consciousness of the openly endless multiplicity of
changing experiences and experienced things, one’s own and those of others”
(Husserl 1970a: 164). Any act of perception is part of a field of perception in a
communal, shared world, where the body plays a major role as the zero point
of orientation (Husserl 1989: 61), which is also available for others to
interpret – this is what Alfred Schutz (1967) called the “field of expression”
of subjective experiences. This means that instead of having to be evoked or
recruited under the special circumstances of “we-intentions,” a sense of col-
lectivity is already experienced, no matter how minimal or slim, at the most
basic level of everyday experience, namely, in the routine and often uncon-
scious perception of the surrounding world, even before joint activities occur
or utterances are exchanged.
230 A sense of the other

10.7 Our everyday understanding of others


Phenomenologists long ago described and conceptualized how by living with
others we understand them without having to consciously pay attention to how
we come to such understanding:
I not only consciously experience you, but I live with you and grow old with you. I can
attend to your stream of consciousness, just as I can attend to my own, and I can,
therefore, become aware of what is going on in your mind. In the living intentionality
of this experience, I “understand” you without necessarily paying any attention to the acts
of understanding themselves. This is because, since I live in the same world as you, I live
in the acts of understanding you. You and your subjective experiences are not only
“accessible” to me, that is, open to my interpretation, but are taken for granted by me
together with your existence and personal characteristics . . . while I am directly experi-
encing you and talking with you, the whole complicated substructure of my own
interpretation of you escapes my attention. I am not interested in such matters; my living
intentionality, my attention à la vie, has other goals at the moment. (Schutz 1967: 140)

This does not mean that there is no difference between the experience of one
individual acting alone and a number of people acting together, but that the
difference is not as sharp a dichotomy as Searle suggests with the distinction
between individual and collective intentionality.
The recognition of the communal quality of life experiences is important
because, among other things, it allows for correction (Husserl 1970a: 163), that
is, for the adjustment of my perception to that of others. From the point of view
of phenomenology, when there is a problem in understanding, people can and
will move out of the “natural attitude” – what Schutz in the quote below
glosses as “my simple and direct awareness of the other person” – and
consciously engage in an act of interpretation or reinterpretation.
For instance, I may ask, “Have I understood you correctly?” “Don’t you mean
something else?” “What do you mean by such and such action?” These are typical
of the questions that I am forced to ask every day in my relations with other people.
The moment I raise such questions, I have abandoned my simple and direct aware-
ness of the other person, my immediate grasp of him in all his subjective
particularity. I have abandoned the living intentionality of our confrontation. The
light in which I am looking at him is now a different one: my attention has shifted
to those deeper layers that up to now had been unobserved and taken for granted.
I no longer experience my fellow man in the sense of sharing his life with him;
instead I “think about him.” But now I am acting like a social scientist.” (Schutz
1967: 140–141)

The possibility of such shifting – what Husserl captured with his notions of
modifications (see Chapter 9) – is an aspect of human flexibility that is
cultivated in jazz improvisation (see above) as well as other art forms. It allows
individuals to adjust to particular circumstances and learn new ways of seeing,
Conclusions 231

hearing, and speaking, albeit not without struggle and with certain limitations
(Ochs, Solomon, and Sterponi 2005). The flexibility of our intersubjective
engagement in and with the world, in turn, makes possible social adaptability,
which, however, comes with a price. The very possibility of seeing the world
from the point of view of an Other, who, in turn, constitutes us as objects of his
or her perception and evaluation, makes us vulnerable. In making the world a
shared field of perception aiming at a shared or exchangeable understanding,
intersubjectivity also makes it into a world of mutual monitoring. As pointed
out by a number of philosophers and social theorists, such monitoring has a
dark side. Being able to empathize, exchange places, and understand each
other also feeds social conformity, lack of authenticity,16 the acceptance of the
locally sanctioned prestige – what Gramsci (1971, 1975) called “hegemony” –
and, ultimately, the reproduction of the status quo.
Given all of these possible and very real effects of our intersubjective
engagement with the world of others, it should not be surprising that, as
discussed in other chapters in this book, in some places in the Pacific, includ-
ing Samoa, and possibly elsewhere in the world, some people may try to avoid
revealing too much about their own thoughts and feelings – children certainly
learned that at the dinner table (Ochs and Taylor 1995) – or speculating about
those of others (see the discussion of the “opacity of mind” in Chapter 8).

10.8 Conclusions
This chapter, which started with the appraisal of Searle’s notion of collective
intentionality (and “we-intentions”) as distinct from individual intentionality,
ended up with the discovery of the intersubjective side of any type of human
act, whether or not the individual actor (or the observer) conceptualizes what is
being thought, felt, or done as a “social” phenomenon. It should not then be
surprising that driving the car through traffic is something that we are able to
do because of a series of skills, habits, instruments, and affordances that make
it possible for our mind and body to coordinate with people and things around.
We are helped and guided by the road, the lanes, the traffic lights, the road
signs, and the ways in which automobiles are built. This is to say that our
actions are typically embedded in an artifact-filled environment that minimally
constrains and more often suggests to us ways of moving in the built or in the
natural environment. A phenomenological understanding of our involvement
not only with tools – which are made by others and are often used for others –
but with any object of perception recognizes co-presence and anticipation of
human interactions even when we are not purposely or consciously preparing
for a future “collective” activity of the type described by Searle (1990).
These observations bring out the fact that, however useful for thinking about
collaboration, Searle’s notion of collective intentionality may underestimate
232 A sense of the other

the common distribution of tasks and coordination that are not only necessary
but also constitutive of social life, as recognized by the researchers who
proposed approaches such as “distributed cognition” or “active externalism”
(see §10.3).
This discussion constitutes an interpretive “turn” with respect to the original
claim made by anthropologists that intentions are either inaccessible to certain
groups (e.g., certain Pacific communities) or not relevant to the ways in which
they interpret each other’s actions (see Chapters 2, 3, and 8). The original
anthropological critique has moved in new directions. Even when we restrict
our analysis to the misleadingly simple and obvious cases of collective inten-
tionality provided by Searle, like the “people in the park” discussed in this
chapter, the difference between “collective” and “individual” intentions can
easily be questioned or even vanish in light of what people ordinarily do when
in the presence of one another in a public place like a park. We are then faced
with the fact that some of the properties of “we-intentions,” like mutual
adaptation and mutual monitoring, seem to be occurring even among people
who are not seen as doing something “together.” Furthermore, we have found
that improvisation is a recurrent feature of human action, even when it seems
planned.
Some of the data I collected on jazz instruction and jazz performance gives
us an opportunity to hear musicians’ conceptualization of improvisation and
observe one of them while he socializes novices to be engaged in highly
coordinated collective behavior (e.g., playing in a band). Through the analysis
of what we might call “the embodiment of intentions” in playing with others,
we get a feel for how mind and body need to coordinate in the development of
a habitus that makes creativity possible within the constraints of a shared sense
of time and timing.
It is with these examples in mind that I introduced Husserl’s theory of
perception and showed that it is quite different from the individualistic theory
presented by Searle (1983). It is through perception that I introduced Husserl’s
notion of intersubjectivity, which starts from the assumption that Others are
not only always potentially “in play” but are de facto taken into account in any
kind of act, including the most basic ones like looking at a table.
We are thus reminded that, whether or not we accept Searle’s notion of “we-
intentions” as a “primitive” (logical) notion and thus distinct from an accumu-
lation of separate “I-intentions,” the Other is in fact always already there, a
point made most famously by Heidegger in Sein und Zeit ([1927]1993) but in
fact already present in Husserl, as shown by his earlier writings and lecture
notes (Zahavi 2001b). Different cultural traditions may or may not recognize
or encode in their language the wide range of our human ways of being-with-
others, but we know that they are there. Our intentions, together with our
language, are always in a world of others.
11 The intentional continuum

11.0 Introduction
In this concluding chapter I argue that, rather than discouraging any use of
“intentions,” the various cross-cultural, cross-linguistic, and historical analyses
presented in the prior chapters have shown the need to formulate a more
contextually rich and contextually sensitive notion of intentionality as a pan-
human capacity to attend to all kinds of real and imagined entities while giving
them meaning. I believe that some notion of intentionality – and thus some
notion of “intention” – is needed if we want to make sense of how people
orient themselves in the social world, anticipate the actions of others, and
follow their own instincts, intuitions, moods, needs, and desires. Rather than
just being another “language game” (Wittgenstein 1958), a discussion of
“intentions” – which includes a critical stance toward it – is important for an
anthropological understanding of human agency as constrained by existing
relations of power and yet sometimes resistant to long-established institutional
practices. One lesson of our interpretive journey has been that the reliance on
intentions and their apparent avoidance must always be contextualized. The
notion of “intentional continuum” will be here introduced as a way of making
sense of the contextual variability in intentional action that I documented in
this book.1

11.1 The meaning of intentions and their uses


As discussed in Chapter 2, three prominent philosophers of language used the
notion of “intention” to explain how people make sense of what someone
means in saying something. I called this view the “Standard Theory.” It adopts
a commonsense notion of intending to characterize a relationship of agency
and (sometimes) responsibility between persons and their acts. Within the
“Standard Theory,” there are some differences. Grice (1957) used the verb
intend to explain what makes a given utterance meaningful for both speakers
and hearers. Austin (1962, 1975) used “intention” in his “felicity conditions”
for all kinds of speech acts, including promises. Searle (1983) went beyond

233
234 The intentional continuum

speech, extending the use of intentions to provide his own definition of what
counts as “action.”
For Grice speakers are accountable for saying something that is supposed to
be true and want their audience to recognize their intention that they mean it
that way. Austin wrote that for speakers to “mean a promise,” they had to
“have a certain intention” to keep their word (Austin 1962: 11). But what does
“intend” for Grice and “intention” for Austin mean? Not necessarily the same
thing. When Grice says that a speaker “must intend to induce” a certain belief
in an audience (see §2.1), he seems to say that the speaker wants the audience
to have that belief and to recognize that she wants them to have it. Some
empirical support for the connection between meaning and wanting is found in
French and Italian where to say “to mean” one literally says “want to say,” as
in je veux dire (in French) or voglio dire (in Italian), both of which can be
translated as either ‘I mean’ or ‘I want to say.’
For Austin, an intention could have the sense of a determination, will, or
commitment to doing something. This determination may vary depending on
the specific speech act. In the case of promises, intention may cover sincerity –
a commitment to be sincere – and the “good will” to follow up with an action
that satisfies the “obligation” established by the promise. Austin admitted that
there could be varying degrees of intentionality by saying that when commit-
ting to some future action by means of the the expression “I shall . . .” we could
add adverbs like “undoubtedly” or “probably” (Austin 1975: 77). This sug-
gests that “intention” here is understood as a positive disposition to some
future action.
When Searle went beyond speech to propose his more general theory of
intentionality, which included a theory of what constitutes “action,” he tried to
clarify his use of “intention” by drawing a distinction between (a) “intention”
as the description of one among a large repertoire of possible acts that can be
described in the English language together with wish, fear, hope, etc., and (b)
“Intention” (with a capital “I”) as the more general and foundational concept
corresponding to Brentano’s and Husserl’s sense of intentionality. This is the
very broad notion of intentionality as the “aboutness” of the acts of our
consciousness in relation to real or imaginary entities. But once Searle started
to use the categories “intentional” and “unintentional action,” he seemed to fall
back on a more ordinary use of the English intention, sometimes understood as
the conscious goal of an action and other times as an unconscious goal, which
can be made explicit upon reflection. This is at least how I read the two
hierarchically related types of intentions that he called, respectively, “prior
intentions” and “intentions in action.” The former cause the latter, which are,
in turn, means through which the (higher-level) “prior intention” can be
realized. Prior intentions may have as a scope fairly complex actions like
going to the office, buying a car, telling a story, as well as more simple actions
The meaning of intentions and their uses 235

like grabbing the handle of a cup, turning on the car’s engine, putting on a
jacket. “Intentions in action” cover what Artificial Intelligence scholars call
“sub-goals” or “simple goals” (Lin 1997), which are needed in order to
accomplish the higher goal defined by the prior intention.
For Searle, an event qualifies as an “action” (that is, as an “intentional
action”) by the actor only if it is within the scope of an intention. For example,
in Searle’s description of the case of the murder of Archduke Franz Ferdinand
in Sarajevo by the anarchist Gavilo Princip, we are told that Princip’s (overall)
“prior intention” was to kill the Archduke in order to avenge Serbia, and this
“action” required a number of “intentions in action” (e.g., pulling the trigger of
the pistol). There were other events that followed the murder of the Archduke;
for example, World War I started, but such events do not qualify as Princip’s
“actions” because they were not included in his (prior) intention. World War
I would then be an “unintentional action” (Searle 1983: 98–100). The same
could be said about Oedipus marrying Jocasta. Since Oedipus did not know at
the time that Jocasta was his mother, despite the fact that “Oedipus married his
mother” is a true proposition, “marrying his mother” cannot be considered
Oedipus’ “action” because it was not in his (prior) intention, hence Searle’s
rejection of “descriptions” as ways to identify “actions” (see §2.2).

11.1.1 Matching theory with some ethnographic and linguistic data


If we apply these criteria to the case of the Samoan orator Loa discussed in
Chapter 3, we would say that Loa’s (prior) intention to communicate that the
district MP was going to bring some presents to the members of the village
council caused Loa to call a meeting of the council (fono). The speech act
(some version of an utterance like “Let us gather because the MP [said that he]
will come next week with gifts for everyone”) would then be an “intention in
action” (with its own conditions of satisfaction) caused by the prior intention to
let the council members know. As we saw in §3.3, according to the accusatory
speech delivered by the senior orator Iuli, Loa’s summons (see line 3481 in
example [5]) caused an unpleasant situation for the members of the village
council, who felt humiliated for having waited in vain for the MP and the gifts.
In Searle’s model, the “intentional action” of Loa’s announcement could not
include the effects that were listed by Iuli in his accusatory speech – in
Austin’s terms, they would be considered to be perlocutionary effects. In
Searle’s (1983) terms, they would all be unintentional actions. The empirical
(and analytical) problem presented by this case is that the distinction between
what Loa did and did not intend was never brought up in the discussion by Iuli,
Loa himself, or any of the others present at the meeting, including those who
did not seem aligned with Iuli’s position to fine Loa. For example, one reason
that was mentioned for dismissing the case was a procedural detail, namely,
236 The intentional continuum

that Iuli’s charge against Loa had not been announced at the beginning of the
meeting during the first speech of the day when the ‘topics of the meeting’
(matā`upu o le fono) or ‘topics of the day’ (matā`upu o le aso) are announced
(see §5.5.1). We might then conclude that, from the point of view of the
council members, Loa’s “action” included the embarrassment his reported
speech had caused, even though from the point of view of speech act theory,
such embarrassment was a “perlocutionary effect” and from Searle’s point of
view it was an “unintended action” and as such did not qualify as his action.
How can we make sense of this case for an anthropological model of
human action? One answer to this question would be to say that Samoans
subscribe to the so-called “absolute liability” doctrine, according to which
people are judged on the basis of the consequences of their actions, regard-
less of their intentions. This is what Goldman (1993) critically imputes to
what Ochs and I wrote in the 1980s (see §3.4). But as I have shown in
§3.4.1, this hypothesis does not hold. Samoans do have terms for causes and
reasons, which the members of the village council use in discussing whether
certain individuals should be held responsible for certain actions. It is just
that they do not go into much detail in such cases, tending to stick to causes,
e.g., being drunk, that avoid getting into a person’s specific in-order-to
motives (Schutz 1967).
A second possible answer is to say that Loa’s actions are not evaluated in
terms of his intentions but in terms of Loa’s responsibility, as suggested by
Iuli’s speech where he complains that Loa did not do anything to remedy the
situation once he realized that the MP was not coming. This would be a
departure from Searle’s model because he thought that it is a “mistake to
suppose there is some connection, perhaps even identity, between intention
and responsibility” (Searle 1983: 103). However, a connection is recognized
by those legal systems in the West for which there are actions that one might
not have intended but might have foreseen, thereby implying that a person
might have known or thought that what he was doing might have caused harm
to others.
And what about the intentions of Loa’s accuser, Iuli? As we saw in
Chapter 3, during the discussion of Loa’s case, no one brought up possible
reasons for Iuli to be so adamant about the need to punish Loa or possible
goals of his accusation. Given Loa’s family connection with the MP, it is
possible that Loa had campaigned on the MP’s behalf and thus against the
three candidates from his own village, including Iuli. Whatever information
Iuli had access to and might have motivated his attack on Loa, the issue of
personal revenge was not mentioned during the meeting. In providing an
ethnographic analysis of the events, what Geertz (1973) called “thick descrip-
tion,” such details would matter, but the same reluctance to talk about inten-
tions that makes this case interesting from a comparative perspective would
The meaning of intentions and their uses 237

make it difficult for us to find textual evidence about Iuli’s ulterior motives,
especially in a public arena like the Samoan fono.
The fact that Loa could be accused at all, together with the fact that his
“intentions” in Searle’s sense of the term – that is, the limited goal of reporting
the news of the MP coming – were not taken into consideration suggests that,
in engaging in this type of cross-cultural analysis, we might find differences in
the very scope of what counts in evaluating a person’s actions and the
responsibility for those actions. In Loa’s case, Iuli and some of the other
members of the village council might have thought of Loa’s “action” as
something more than just “announcing” or “reporting” what the MP told
him. It could have been that the very fact of making the announcement about
the MP’s visit might have been seen as one move within a wider sequence of
actions, which might have had as a goal the accumulation of cultural capital
and personal prestige. For example, Loa might have been seen as using the
announcement as a strategy to claim a preferred relationship with the MP. It is
also possible that Iuli and others interpreted Loa’s action as an attempt to act or
appear as a political broker, someone who could assume the important role of
peacemaker after the tension created by the incumbent’s victory over Iuli and
the other two matai from Loa’s village (see §3.3). In this case, instead of being
irrelevant and inscrutable, as suggested by the doctrine of the “opacity of
mind” (see Robbins and Rumsey 2008 and my discussion in Chapter 8), Loa’s
“intentions,” understood as his individual goal in reporting the MP’s message,
might have been in everyone’s mind. In fact, the other members of the village
council might have been wondering why the MP told Loa about the visit
instead of communicating directly with one of the senior orators (Iuli or
Moe`ono). Claiming or implying a privileged relationship with a higher-status
person like the MP can be dangerous in such a status- and rank-conscious
society as Samoa. All of this might have become aggravating circumstances
once the MP and the gifts did not arrive.
What can we conclude from all of this? One lesson is that if we interpret
intentions as goals of actions, as Searle seems to be doing in his examples (see
above), we must allow for the fact that those goals are subject to local
understanding and therefore that the scope of the conditions of satisfaction
for an action will vary across cultural (including institutional) contexts. We
have thus at least two possibilities. We can say that certain social actors are
accountable for intentions that they might not have had, but others attribute to
them anyhow on the basis of their institutional role, as commonly recognized
in western legal systems, where an employer can be blamed and be financially
responsible for negligence in the actions of the employees, and in everyday
life, like when parents are blamed for their children’s actions. The second
possibility is to agree with those who believe that “the intention” of a given
action is more likely to be a post hoc process of justification based on the
238 The intentional continuum

supposed typical state of mind in which a person in general or a person in


particular might have been. As we have seen in Chapter 10, where I discussed
approaches to human action that are relevant to or require the use of tools and
the collaboration with people, there is a considerable amount of knowledge,
e.g., of what to do next, which Searle might see as represented in intentions,
that relies on rapid and non-conscious neural processes, affordances in the built
and natural environment, and a habitus acquired and embodied through
repeated interactions, all of which are processes that seem difficult to capture
with mental “representations.”

11.2 The “intentional continuum”


As shown in Chapter 2 and Chapter 9, for Husserl intentions cover a wide
range of relationships between human consciousness and the phenomenal
world. Intentionality is sometimes synonymous with attention or attentiveness,
a “turning toward” (Zuwendung) (Husserl 1973: 55) or “being directed at”
such entities (see §2.7 on the meanings of the Latin verb intendere and the
nominal intentio, used by the Scholastics and, later, by Brentano). This
intentionality corresponds to the awakening of an interest in something, which
may become colored by positive or negative affect or be further subjected to
aesthetic or ethical evaluations. Other times intentionality is a layered, com-
plex ensemble of sensations, expectations, evaluations, activations of memor-
ies, and projections of what others will do next. This is, for example, part of
what happens when members of a band make music together or when we have
a conversation. As shown in Chapter 9, Husserl’s various notions of “modifi-
cation” help us explain how the experience of the same object can be trans-
formed into something else, thereby giving rise to new types of phenomena
and attitudes, all of which participate in producing more complex intentional
contents.
As I discussed in Chapter 10, a phenomenologically informed theory of
intentionality also includes an understanding of the role of Others in our
perception, thoughts, feelings, evaluations, and actions. This means that inten-
tionality is intrinsically intertwined with intersubjectivity, a property of our
being in the world that covers, albeit in different ways and different degrees,
the most mundane and apparently isolated acts, including our dealings with our
own body, and the most complex forms of social collaboration (Zahavi 2001b;
Duranti 2010). Rather than being the end of subjectivity as a focus of inquiry,
the emphasis on the intersubjective dimensions of our thinking, feeling, and
doing simply recognizes that any individual act exists over against an inter-
subjective horizon. Ontogenetically, by the time language makes its entrance
(as speech or as embodied signs), the Self of the child has been ready for the
great cognitive and social development that conventional symbols make
The “intentional continuum” 239

possible (Trevarthen 1979, 1980). Some of these processes are reproduced at


the interactional level, where the use of language is typically built on a
preexistent sociality, which it then further transforms.
Building on these observations and reflections, it would seem more appro-
priate for a contextually sensitive notion of intention to avoid any kind of
dichotomy, whether it is between “intention” and “Intention” (see §11.1) or “I-
intentions” and “we-intentions” (see Chapter 10). It then makes sense to posit
an “intentional continuum,” that is, a range of graded ways of being disposed
or mentally (and sensorially) connected with some entity in the world, from a
basic relationship between our consciousness and some entity that attracts our
attention (through sound, vision, smell, taste, or touch or any combination of
them) all the way to our partly explicit anticipation or planning of complex
activities, resulting from several hierarchically organized dispositions, like
telling a story, having a dinner party, or preparing for a business meeting
and then running it. In any of such activities, individuals in fact go in and out
of “intentions,” in the sense that they alternate between different states of
mind, embodied attitudes, and projected courses of action, in a temporal
unfolding where certain goals and plans are (often a posteriori) recognized
and evaluated as “meant” and as such qualify as “intentions.”
As we saw in prior chapters, languages vary in the encoding of such states of
mind and projections of actions. In English, legal scholars have debated and
judges have provided glosses to juries for the meaning of intention, intent, and
intend, but there is no general agreement (Garner 1995: 458–459). Outside of
legal arenas, the verb “to intend” can be synonymous with “to plan,” as in
questions like “What do you intend to do about X?” But there is something
marked or non-ordinary in this use of intentions and intending, which is made
apparent in statements such as “I intend to eat my lunch,” which has a
connotation of self-affirmation or defiance. All of this suggests that the pre-
sumed “neutrality” of these terms for the purpose of philosophical argumenta-
tion is not reflected in other contexts and that, in fact, the use of the English
intention or intend to frame (future) actions carries indexical values that need
further empirical investigation (see D’Andrade 1987 for a “folk model” of the
mind that includes an analysis of “intention”).
Cross-linguistic analysis further exposes the difficulty of finding any
“neutral” terminology to use for comparative purposes, especially if we keep
in mind the range of interpretations of intentionality also implied in Husserl’s
discussion (see above). As we saw in §2.7.1, for example, Samoan does not
have a term that corresponds to the meanings typically attributed by philos-
ophers to the English intention or intending. But if we reexamine the Samoan
lexical choices in terms of an intentional continuum, we realize that Samoan
has encoded specific aspects of the continuum that are missing or back-
grounded in the English lexical choices. Thus, when combined with other
240 The intentional continuum

words, the Samoan word loto covers a wide range of meanings that make
evident the affective element of human action and the ways in which cognitive,
emotional, and embodied dispositions participate in the constitution of the
moral character of a person (see also Mageo 1991, 2010). As always in cross-
cultural and cross-linguistic analysis, we may be tempted to imagine what it
would have been like to propose a theory of meaning (and action) that starts
from something like loto instead of intention. Alternatively, one could work
with English and other European languages (e.g., French, German) and ask
what kind of theory of human action could be built on notions that share a
family resemblance with “intention” but come with a different set of connota-
tions, like, for example, “will” and “volition,” two closely related concepts
whose history in the social sciences and humanities has been recently exam-
ined by Murphy and Throop (2010).

11.3 Hypo-cognition of intentions and hyper-cognition of


perlocutionary effects
The concept of the intentional continuum captures the variation that we find
across individuals and groups in recognizing or discussing intentions, and
therefore it could be combined with Robert Levy’s (1984) notions of hypo-
and hyper-cognition. We could then say that intentions are hypo-cognized in
those communities where ethnographers have identified reluctance to speak
about intentions (and/or motivations) and to explicitly engage in reading
other minds or even revealing the reasons for refusing to do something (see
Chapters 3 and 8). Conversely, intentions might be attributed to a wide range
of entities, from supernatural forces to people to animals and even to the earth,
as seems to be case among the Hopi (Whiteley 1998: 42; Richland 2010).
The evidence and sources discussed in the preceding chapters suggest that,
in addition to what we could call the “taxonomy of intentional states and
stances,” speech communities also vary in terms of the practice of talking
about or evoking intentions as explanations for people’s actions. What had
been claimed by ethnographers like Mead (1928) and Shore (1982) was
confirmed by the study of Samoan language socialization (e.g., Ochs 1982,
1984, 1988) and my own analysis of Samoan political-juridical discourse (see
Chapters 3 and 5). There is no question that both ethnographic observations
and textual analysis support the claim that Samoans are more likely to focus on
the consequences or effects of actions than on their origins, causes, or possible
intentions and motivations of the people involved. The data presented in
Chapter 3 constitute strong empirical evidence for this generalization.
Rephrased in Austin’s (1975) terminology, we could say that Samoans are
just as interested in the perlocutionary effects of people’s actions, utterances
included, as they are in the illocutionary force of a communicative act. To
The performance of interdependence 241

extend Levy’s (1984) language, we could say that in Samoa perlocutionary


acts are hyper-cognized. In Husserl’s terminology we could say that the
attention and imagination of Samoan social actors is drawn to perlocutionary
effects more than to possible original mental representations, motivations, or
intentional dispositions.

11.4 The performance of interdependence


The just-described emphasis on the end results of an action (or a series of
actions) should, however, not just be conceptualized as a cognitive preference,
something analyzable in terms of individual psychology. In a society like the
Samoan, where kinship relations continue to play such an important role in
providing both rights and obligations, individuals are constantly reminded by
complex rituals as much as by language-mediated routine interactions that they
are members of a network of relations, which include other individuals,
groups, or institutions that have authority over them and expectations that
need to be met with or without a warning. This is true in many if not all
societies, but in Samoa this performance of interdependence – as opposed to
the performance of independence that other societies might favor – is enacted
in countless and continuous requests for help, services, and material goods that
are testing and reproducing political alliances and the reliability of family
relations while providing evidence of alofa, a term that covers love and
compassion (Pratt 1893: 72) and is included in the word for ‘gift,’ mea-alofa
(lit. ‘thing-love’), that is, the things you give to show that you care.
Requests for such performances can only be realized through the collabor-
ation of others, who may act as messengers, intermediaries, supporters
(tāpua`i), lenders, workers, advisers, etc. and therefore by definition any
“action” that involves others is also likely to participate in the production
and reproduction of social obligations. The individual, in all of this, plays an
important part, but in order to let the social role be effective, he or she must
also forget himself or herself as a ‘person’ (tagata), as entailed by the negative
attribute fia-tagata, ‘having the urge to be a person,’ the traditional expression
referring to someone who is a ‘show off’ (fia-siō), too proud or even irreverent.
In such a system of continuous production and redistribution, which highly
values collective projects (building or rebuilding a church, preparing for a
wedding, funeral, installation of a chiefly title, visit by a traveling party, etc.),
individual commitments occur but they have to be taken with a grain of salt
because the “reasons of the family, subvillage, village, district, etc.” – the local
equivalent of the “reasons of the state” in western nations – tend to have
priority over individuals’ plans. It should not then come as too much of a
surprise that the notion and practice of “promise” was introduced by foreign-
ers, as I suggested in Chapter 4, and even today is not expressed by a
242 The intentional continuum

universally accepted, unambiguous term. Oaths, of which we do have evidence


in the historical record (see §4.7), might have been more fitting for a society
whose members value the public display of obligations, are vigilant about lies,
and do not trust private deals or commitments, which are easy to forget or
deny.

11.5 Individuals, social coordination, and social control


The comparative discussion of intentional discourse and intentional language
has also brought out the issue of social control.
Searle’s notion of collective intentionality is meant to make explicit the
ways in which we guess what an Other is about to do and for what purposes, in
order to achieve smooth interactions and a high level of social coordination. At
the same time, the very ability to imagine or predict what the Other perceives,
wishes, or fears creates the conditions for the Other’s social and emotional
vulnerability, just as the transparency of our own intentions, motives, and
wishes creates the conditions for our own vulnerability. In a society like Samoa
(as shown in Chapter 3), one is allowed to shut down communication and thus
avoid any inquiry about what one is thinking or feeling by adopting the posture
called musu (see §3.1 and §8.1), The avoidance of finding oneself in such
vulnerable conditions may provide a social motivation behind the doctrine of
the “opacity of other minds” proposed by Robbins and Rumsey (2008). But, as
I argued in Chapter 8, it is hard to find a system that is fully consistent in such a
stance. In the examples provided in the ethnographic literature, we find cases
where even in those places where people seem to explicitly reject engaging in
mind-reading, there are contexts in which they do. Similarly, when we control
for the domain of inquiry and we look at discourse that is politically charged,
we find that people from quite different societies may show similarities in how
they try to exert control over the interpretation of what someone has just said.
Thus, when, in Chapter 6, we looked at political discourse in the US, we saw
that an audience of supporters might read meanness into words that the
candidate claims to come from a neutral or non-judgmental stance. In this
context, we saw an individual speaker in the US having a hard time regaining
control of the interpretation of his “action” just as we saw, in Chapter 3, an
orator in a Samoan village council meeting twenty years earlier having a hard
time avoiding being blamed for reporting someone else’s “promise.” For both
of them, the control of the meaning of what they had said is a “serious game”
(Ortner 2006). For us, the study of such situations is a serious interpretive
enterprise.
Appendix A Transcription conventions
for English examples

The excerpts presented in this book are transcribed according to a modified


version of the conventions originally established by Gail Jefferson for the
analysis of conversation (Sacks, Schegloff, and Jefferson 1974: 731–734).
W. Capps; (abbreviated) names (or initials) of speakers are separated
from the rest by a semicolon (;) and one or more spaces. The
convention of using semicolon instead of colon (:) was
introduced by John B. Haviland when he developed a com-
puter program for transcription work. It is meant to free the
colon to mark only the lengthening of sound.
?; question mark instead of the name of speaker indicates that
the identity of the speaker cannot be established.
no::: colon (:) stands for lengthening of sound.
anybody underlining represents emphasis or contrastive stress.
NO!! capital letters indicate higher volume than the volume of the
surrounding talk.
job¼I mean equal sign (¼) stands for “latching,” i.e., no hearable interval
between two turns or between two utterances by the same
speakers.
go //next point in a party’s turn where overlap by other speaker(s)
starts. The point of overlap is also marked sometimes by a
square bracket that connects the utterance by one speaker to
the utterance of the other.
((laughter)) double parentheses frame contextual information about the
following talk.
(yesterday) words between parentheses indicate reasonable guesses at
what a speaker said.
(? ?) question marks between parentheses signal that a reasonable
guess of what someone said was not possible.
... three dots signal an untimed pause.
(2.5) a number between parentheses is used for timed pauses, in
seconds and, sometimes, tenths of seconds.

243
244 Appendix A

[. . .] a portion of the transcript was left out.


. a period stand for a falling intonation.
the women, a comma represents a slightly raising intonation.
I have also used boldface to highlight parts of transcripts or quotes that are
being discussed in the immediately prior or following text.
Appendix B Transcription conventions
and abbreviations used in the
Samoan examples

In transcribing Samoan speech, I have tried to use in a consistent way the


spelling conventions adopted in the nineteenth century by the members of the
London Missionary Society who introduced literacy in the process of translat-
ing the Bible into Samoan and making its text available in print (see Chapter 7).
Some of the conventions that the missionaries chose at times create some
confusion in foreigners and native speakers, partly because they are not used
consistently. I have followed G. B. Milner (1966) in trying to be as consistent
as possible in my spelling, but I have deviated from his use and from the
choice of most students of Samoan, including anthropologists (e.g., Mead,
Shore, Freeman), in one thing: I have not altered the sounds of what people
said in order to conform to the standards of written Samoan or of the language
that is considered to be proper by the general population. In most of the
interactions that I recorded (and participated in), including the informal con-
versations among peers and the highly formal meetings of the village council
(fono) discussed in Chapters 3 and 5, Samoans use a phonological register
where the opposition between /t/ and /k/ and between /n/ and /ŋ/ are neutralized
so that the alveolar segments (/t/ and /n/) disappear (see Mosel and Hovdhau-
gen 1992: 22–24). This means that a word like talitonuga ‘belief’ will often be
pronounced [kalikoŋuŋa] instead of [talitonuŋa] and the word for ‘village’
nu`u is pronounced [ŋuʔu] instead of [nuʔu]. Samoans call this pronunciation
‘bad speech’ (tautala leaga) and contrast it with the ‘good speech’ (tautala
lelei) that is written, read, and spoken in prayers, songs, and in classroom
interactions (see Duranti and Ochs 1986). After years of reflection on the
dilemma of honoring both native sensibilities and my own professional stand-
ards of accurate linguistic documentation, I have decided to adopt the
following conventions: I use the ‘proper’ pronunciation (‘good speech’ or
tautala lelei) when mentioning expressions out of context and when citing
written Samoan texts in their published form (e.g., stories recorded by Krämer
or Stuebel, sentences from the Samoan Constitution). But I use ‘bad speech’
(tautala leaga) when it was used in the verbal interactions I recorded. When
I lift expressions in ‘bad speech’ out of a transcript in order to discuss them,
I use obliques (i.e., phonemic slashes) before and after them to alert readers

245
246 Appendix B

who are familiar with Samoan orthography that I am citing from a recording of
what people actually said. For example, I might write /e fogo ma pokopoko/ to
make it clear that the phrase that people would normally see written e fono ma
potopoto, meaning ‘to meet and gather,’ is taken from a transcript of a
recording where it was pronounced [e foŋo ma pokopoko].

Samoan vowels
Samoan has five vowels, which can be short or long, roughly corresponding to
the quality of the five vowels of a language like Spanish (for a more precise
phonetic analysis, see Mosel and Hovdhaugen 1992: 25–28):
a, e, i, o, u

ā, ē, ī, ō, ū
The length is, however, phonologically distinctive, that is, in some cases it is
the determining factor distinguishing between two different words, e.g., sal-
amo [salamo] for (the English borrowing) ‘psalm’ vs. salamō [salamoː], with
the stress on the last syllable, for the verb ‘repent.’ Unfortuately, the macron on
long vowels is only occasionally used in the Bible and was abolished a long
time ago by the (then Western) Samoan Department of Education. This
sometimes creates confusion not only for foreigners trying to learn Samoan,
but, more importantly, among Samoan children reading texts like the Samoan
Bible that contain previously unheard long and complex words. The same
could be said about the use of the inverted apostrophe for the glottal stop (/ʔ/),
which is often omitted in writing in both word-initial and word-internal
position, e.g., the word for ‘boat’ is typically spelled vaa instead of va`a –
the glottal stop here corresponding to Proto-Polynesian *k, which is still
present in other Polynesian languages, e.g., in Maori where the word for ‘boat’
is waka.

Samoan consonants
Samoan has a total of thirteen consonants (Mosel and Hovdhaugen 1992: 20).
In written and ‘proper’ Samoan (see above), three of the consonants (k, h, and
r) are only found in borrowings from other languages, especially, but not only,
for the many proper names found in the Bible, e.g., Herota ‘Herod,’ Esera
‘Ezra,’ and Mareko ‘Mark.’ The sound /k/, however, is also found (in place of
/t/) in the ‘bad speech’ style (see above). Here is the entire set of the conson-
ants as represented in standard Samoan orthography:
f, g (for /ŋ/), h, k, l, m, n, p, r, s, t, v, ` (for /ʔ/)
Transcription conventions for Samoan 247

In transforming an oral language into a written code, the missionaries made a


number of decisions based on earlier experience in other Polynesian lan-
guages, partly adapted to the specific features of Samoan. One of these
conventions was the rational (but confusing) choice of transcribing the velar
nasal sound [ŋ] with one letter, g, instead of the sequence ng that had been
adopted for languages like Tahitian and Tongan. The logic of this choice was
that since there is no native sound /g/ in Samoan (and the English sound /g/
becomes /k/ in borrowings, e.g., English guitar becomes Samoan kitala), the
letter ‘g’ could be used instead of two (ng) to recognize the fact that this is
indeed a single segment in Samoan. Unfortunately, this “rational” decision has
created much confusion among most non-Samoan readers who interpret the
Samoan “g” as a “g” and not as “ng” and pronounce Pagopago, the name of
the capital of American Samoa, [pagopago] instead of [paŋopaŋo].

Abbreviations used in interlinear glosses

2nd second person (pronoun or adjective)


3rd 3rd person (pronoun or adjective)
Art article
Comp complementizer
Du dual
Dx deictic particle providing information about the source and goal of
an action
Emp emphatic post-verbal particle
Erg ergative
Fut future
Imp imperative
In inclusive
Neg negation
Nom nominalization
Perf perfect (often translated into English present or recent past)
Pl plural
Pred predicate particle, same as topic or focus marker
Prep preposition
Pres present tense
Pro pronoun
Pst past tense
Qu question
Sg singular
TA pre-verbal tense-aspect marker
Top topic or focus marker
Voc post-nominal vocative marker
Notes

1 Rethinking anti-intentionalism
1 On the notion of “preference” in conversation, see Schegloff, Jefferson and Sacks
(1977). For the notion of “local preference,” see Ochs (1996: 429–30), and for
“cultural preference,” see Chapter 3 and especially §3.4.
2 By mentioning “extended period” – a vague and relative concept – I mean to stress
the fact that ethnographers tend to be with the people they study for a period of time
that is much longer (sometimes weeks, other times months, and in many cases a year
or more) and more involving from a personal point of view than the occasional
encounter where a “fieldworker” may conduct an interview, administer a question-
naire, or record an event and then leave the site never to return or without finding out
much about the lives of the people he or she recorded.

2 Intentions in speaking and acting: the Standard Theory and its foes
1 It is important to acknowledge that Austin himself did not seem comfortable with the
label “analytic philosophy” for his kind of inquiry and considered the possibility of
calling it “linguistic phenomenology,” as he wrote in his “A Plea for Excuses”:
In view of the prevalence of the slogan “ordinary language,” and of such names as
“linguistic” or “analytic” philosophy or “the analysis of language,” one thing needs
specially emphasizing to counter misunderstandings. When we examine what we
should say when, what words we should use in what situations, we are looking again
not merely at words (or “meaning,” whatever they may be) but also at the realities we
use the words to talk about: we are using a sharpened awareness of words to sharpen
our perception of, though not as the final arbiter of, the phenomena. For this reason
I think it might be better to use, for this way of doing philosophy, some less
misleading name than those given above – for instance, “linguistic phenomenology,”
but that is rather a mouthful. (Austin 1970: 182)
2 On the history of the central role of the individual in society, see Sennett (1974) and
the references given by Rosen (1985).
3 A similar use of intention is found in the second volume of Husserl’s Logische
Untersuchungen:
The articulate sound-complex . . . first becomes a spoken word or communicative bit
of speech when a speaker produces it with the purpose [in der Absicht] of “expressing
himself about something” through its means; he must endow it with a sense [Sinn] in

248
Notes to pages 12–20 249

certain acts of mind, a sense he desires to share with his listeners. Such sharing
becomes a possibility if the listener also understands the speaker’s intention [Inten-
tion]. He does this inasmuch as he takes the speaker to be a person, who is not
uttering mere sounds but speaking to him. The person accompanies those sounds
with certain sense-giving acts, which the sounds reveal, and whose sense he wants
to communicate. (Husserl 1901: 32–33; English translation by J. N. Findlay in
Husserl 1970b: 276–277, here reproduced with some minor changes)
4 Additional references of this nature can be found in Bach (1990).
5 I am using “action” here in Searle’s (1983) sense of the word, that is, as synonym-
ous with “intentional action,” see below.
6 Searle clarifies his position on causality early on in the Intentionality book:
“Intentional states stand in causal relations to the neurophysiological (as well as,
of course standing in causal relations to other Intentional states), and . . . Intentional
states are realized in the neurophysiology of the brain” (Searle 1983: 15). A few
sentences later, Searle gives his view of the “mind–body problem.” “[T]here is no
such problem. The ‘mind–body problem’ is no more a real problem than the
‘stomach–digestion problem’” (Searle 1983: 15).
7 Searle does not address the issue of the mental representation or regulatory mech-
anisms by which variation is endemic at the level of intentions in action and thus not
predictable from the “prior intention.”
8 Given that they are supposed to be unconscious, one might be tempted to consider
such intentions in action as part of what Searle (1983) calls “the Background” (see
§2.1.4), but they are not, because the Background is meant to cover “preintentional”
stances and “preintentional skills” (1983: 144–145) that are not “represented” in the
mind, that is, they do not have corresponding “representations.”
9 This conjecture seems reasonable given that Searle and Dreyfus co-taught courses
in the Department of Philosophy at Berkeley. But the debate between the two
philosophers over the last two decades suggests that they might have been mis-
communicating about their respective goals, approaches, and methods (e.g., see
Dreyfus 1991, 1993, 2001/2; Searle 2000, 2001/2b).
10 The phrase does not seem to be an actual quote from Heidegger’s Being and Time
even though the terms given between quotes by Searle are found in Heidegger
(1962).
11 Cerbone (2000) argued that there are similarities between Searle and Heidegger, but
only discussed the two philosophers’ attitude with respect to the old issue of the
existence of the external world.
12 It was for this reason that I did not include intentionality in my definition of agency
(Duranti 2004).
13 Here is Searle’s attempt to explain his own way of using “representation”:
There is probably no more abused a term in the history of philosophy than
“representation,” and my use of this term differs both from its use in traditional
philosophy and from its use in contemporary cognitive psychology and artificial
intelligence. When I say, for example, that a belief is a representation I am most
emphatically not saying that a belief is a kind of picture, nor am I endorsing the
Tractatus account of meaning, nor am I saying that a belief re-presents something
that has been presented before, nor am I saying that a belief has a meaning, nor am
250 Notes to pages 21–29

I saying that it is a kind of thing from which one reads off its conditions of
satisfaction by scrutinizing it. The sense of “representation” in question is meant
to be entirely exhausted by the analogy with speech acts: the sense of “represent” in
which a belief represents its conditions of satisfaction is the same sense in which a
statement represents its conditions of satisfaction. To say that a belief is a repre-
sentation is simply to say that it has a propositional content and a psychological
mode, that its propositional content determines a set of conditions of satisfaction
under certain aspects, that its psychological mode determines a direction of fit or its
propositional content, in a way that all of these notions . . . are explained by the
theory of speech acts . . . [representation] is just a shorthand for this constellation of
logical notions borrowed from the theory of speech acts. (Searle 1983: 11–12)
14 For an overt reading of Bourdieu’s theory in terms of Heidegger’s criticism of
Husserl’s theory of intentionality and the implicit focus on the Subject–Object
relation, see Dreyfus (1991: 204–205). For a more Husserlian interpretation of
Bourdieu’s writings, see Throop and Murphy (2002).
15 The idea of a “psychological turn” in Searle’s adoption of Austin’s perspective on speech
acts was common in the early 1980s among the anthropologists who were familiar with
speech act theory. One of them was the influential Clifford Geertz, to whom I had sent the
paper that I had presented at the 1983 Meetings of the American Anthropological
Association, together with an invitation to participate in a panel at the 1984 Annual
Meetings. To my surprise, Geertz answered me right away, generously praising my paper
and suggesting the phrase “socio-centric” as a possible name for the type of approach that
I was promoting, which he saw as closer to what Austin was trying to do than “Searle’s
psychologization of him.” I am grateful to Professor Karen Blu, Geertz’s widow, for
granting permission to reproduce the letter in its entirety (reproduced on p. 251).
16 For some textual and biographical details on the professional, intellectual, and
personal relationship between Husserl and Heidegger, see, among others, Hopkins
(1993), Husserl (1997), Moran (2000a, 2000b: 226–233), Heidegger (2002), and
Kisiel and Sheehan (2007).
17 See Moran (2005: 17). Brentano had studied Scholastic theology as a seminarian
(Runggaldier 1989) and had absorbed Thomas Aquinas’ notion of “intentio” as some-
thing immanent – hence his use of the expression “intentional inexistence” – in the form
of a doubling or correlate of the external object in consciousness (Spiegelberg 1976).
18 Husserl (1913) used the phrase “die natürliche Einstellung” which was translated as
“the natural standpoint” by W. R. B. Gibson (Husserl 1931) and changed to “the
natural attitude” by later translators and interpreters (see, e.g., Husserl 1982).
Unfortunately, the psychological flavor of the English attitude made it unappealing
to anthropologists like Clifford Geertz (1973: 110n).
19 On the success of Heidegger’s philosophy in post-WWII France and his relationship
with French intellectuals and academics, see Janicaud (2001, 2002) and Kleinberg
(2005). For a discussion of Levinas’ role in introducing Husserl in France and the
reception of Heidegger’s critique of Husserl’s notions of intention and
intersubjectivity, see Moyn (2005: 21–62 and passim).
20 The Latin habitus is the translation of Aristotle’s hexis, a connection that is lost by
the earlier English translation of ‘habitus’ with ‘habit,’ e.g., in Thomas Aquinas’
([1225?–1274] 1947) discussion of human virtue as ‘habit’ (Latin habitus) (see
Summa Theologica, “Treatise on Habits”).
Notes to page 29 251
252 Notes to pages 29–35

21 For the list of the German edition of the Husserliana volumes, see http://hiw.
kuleuven.be/hua/editionspublications/husserliana-gesammeltewerke
22 Since the 1950s a growing number of authors have carefully examined Husserl’s
books and lecture notes to offer accounts of continuity and change in his interests and
theory, from the earlier analyses of numbers to the lectures on intersubjectivity.
Among the most recent contributions in English that cover Husserl’s entire career,
see Moran (2005), Mohanty (2008, 2011), and the essays in Smith and Smith (1995).
23 Earlier attempts to reclaim the relevance of Wittgenstein’s ideas for the field of
linguistic anthropology might have been too limited (Duranti 1985) or too late to
have any impact on the new generations (see Duranti 1997: 236–244 and passim;
Das 1998).
24 For a phenomenological interpretation of Wittgenstein’s use of “intention,” see Gier
(1981: 135–153).
25 For a critical appraisal of Apel’s position and premises, see, among others, Leilich
(1993) and De Mulder (1993).
26 This is a citation found in Nettleship (1889: 493), who quotes from Cicero’s Orator,
but the English translation is mine because Nettleship did not provide one.
27 This is what Bratman later called “the planning theory of intention”: “The main idea
is to see intentions as elements of stable, partial plans of action concerning present
and future conduct” (Bratman 1999: 2). This is a theory of intentions in which
commitment plays a key role. For this reason, Bratman’s (1992) notion of “shared
co-operative activity” differs from Searle’s (1990) description of “collective inten-
tionality,” which I will discuss in Chapter 10.
28 This section is an expansion of an earlier discussion (Duranti 2006a: 33–35).
29 From transcript “The inspection” (1978). Here is the word-by-word gloss:
magaku `oe e makagā
think 2ndSg TA ugly
‘you think it’s ugly?’ or ‘you think it looks bad?’

30 From transcript “The watch” (1979). Here is the word-by-word gloss:


lo`u ā magaku
my Emp think
‘my very thought’ or ‘what I did think’

31 Here is the Samoan text