Rigorous Data Analysis

PRACTICE OF RESEARCH METHOD

Volume 5

Series Editor

Wolff-Michael Roth, University of Victoria, Canada

Scope

Research methods and research methodology are at the heart of the human endeavors
that produce knowledge. Research methods and research methodology are central
aspects of the distinction between folk knowledge and the disciplined way in which
disciplinary forms of knowledge are produced. However, in the teaching of research
methods and methodology, there traditionally has been an abyss between descriptions
of how to do research, descriptions of research practices, and the actual lived research
praxis.

The purpose of this series is to encourage the publication of books that take a very
practical and pragmatic approach to research methods. For any action in research, there
are potentially many different alternative ways of how to go about enacting it.
Experienced practitioners bring to these decisions a sort of scientific feel for the game
that allows them to do what they do all the while expressing expertise. To transmit such
a feel for the game requires teaching methods that are more like those in highlevel
sports or the arts. Teaching occurs not through first principles and general precepts but
by means of practical suggestions in actual cases. The teacher of method thereby looks
more like a coach. This series aims at publishing contributions that teach methods much
in the way a coach would tell an athlete what to do next. That is, the books in this series
aim at praxis of method, that is, teaching the feel of the game of social science research.
Rigorous Data Analysis
Beyond Anything Goes

Wolff-Michael Roth
University of Victoria, Canada

SENSE PUBLISHERS
ROTTERDAM / BOSTON / TAIPEI
A C.I.P. record for this book is available from the Library of Congress.

ISBN 978-94-6209-996-8 (paperback)

ISBN 978-94-6209-997-5 (hardback)
ISBN 978-94-6209-998-2 (e-book)

Published by: Sense Publishers,

P.O. Box 21858,
3001 AW Rotterdam,
The Netherlands
https://www.sensepublishers.com/

Printed on acid-free paper

All rights reserved 2015 Sense Publishers

No part of this work may be reproduced, stored in a retrieval system, or transmitted in any form or by
any means, electronic, mechanical, photocopying, microfilming, recording or otherwise, without written
permission from the Publisher, with the exception of any material supplied specifically for the purpose
of being entered and executed on a computer system, for exclusive use by the purchaser of the work.
Contents

Preface vii
Glossary xi
Exergue 1

PART A: INTRODUCTION 3
1 Rigor in Qualitative Data Analysis 7

PART B: FIVE DATA SESSIONS 27

2 Data Session 1 (Heidi) 35
3 Data Session 2 (Vicky) 51
4 Data Session 3 (Bullrush) 65
5 Data Session 4 (Mikela) 77
6 Data Session 5 (Kiana) 95

PART C: RIGOR AND THE PRAGMATICS OF RELATIONS 105

7 Turn Sequences 109
8 Knowledge-Power and Institutional Relations 123
9 On the Shop Floor and Playing Field 141
10 The Documentary Method of Interpretation 165
11 Getting Time Back into the Analysis 189
vi CONTENTS

PART D: EPILOGUE 203

12 Socially Responsible Data Analysis 207
Appendix A: Transcriptions for Part B 225
Appendix B: Transcription Conventions 231
References 233
Index 239
Preface

In 1988, just after the completion of my PhD and while still pursuing the idea of a
second PhD in physical chemistry, I interviewed for a position as a post-doctoral
fellow with Jere Confrey, who, at the time, was at Cornel University. My own re-
search had largely been quantitative and I had done a minor in quantitative ap-
proaches in social science research so that I could teach statistical methods. As part
of the visit and without prior notification, Jere handed me a transcription and gave
me an hour to produce a written analysis. I was stunned. I looked at the transcrip-
tion and, as the allotted time was passing, I was unable to produce much of what I
thought to be useful. When I was not offered the job, I was not surprised, attrib-
uting the outcome in part to the impromptu analysis of verbal data (though insiders
later told me that there were other reasons, including possible conflicts between
quantitative and qualitative methods that the two co-investigators of the project
appeared to anticipate). I asked myself at the time, How could anyone produce an
appropriate reading of a transcript without knowing the particular context that had
led to its production? Although the data sessions in part B have precisely this di-
mension, where an expert analyst is shown in the process of reconstructing the type
of events that might have produced the particular transcriptions read, the purpose
of this book is not to show how this can be done. Instead, the purpose is to show
that qualitative data analysis can be done rigorously and how this possibility of
rigorous data analysis is enacted in particular cases. One of the important things in
rigorous data analysis is to be reflective about importing into the analysis presup-
positions and familiarity with aspects of events that the participants themselves
have not had. That is, for example, in this book I abstain from the common practice
of doing the equivalent to Monday morning quarterbacking, where games are de-
scribed and explained with 20/20 hindsight, using what was known and available
only after an event to read the event. Instead, I show how we can approach data in
a rigorous way, showing how societal relations produce social phenomena (e.g.,
power/knowledge) endogenously, from the inside, rather than the latter determin-
ing the former. It is not that teachers and students or managers and employees have
different powersomehow in the way we have money in our pocketswhich then
viii PREFACE

determines their relations, but rather, as Foucault (1975) suggests, the relations
produce (reproduce, transform) relevant and salient differences in knowledge-
power. This becomes very clear when we try to make sense of transcriptions of
which we do not know the origin. Therefore, we cannot use institutional differ-
encesteacherstudent, manageremployee, or coachathleteto explain the re-
lation of people in particular societal activities but we have to take the route Fou-
cault proposes to getting at the social: by analyzing transcriptions taken as
protocols of the relations that produce and (in part) are produced by the ongoing
talk.
For a long time, I have found courses in methods or methodology to be inher-
ently contradictory because these assume that one can first learn the abstract prin-
ciples of a practice prior to actually perform it. Whereas this may be the case when
we already know a related practice that bears family resemblance with what we
want / are supposed to learn, it is in general impossible when we have not yet in-
troduced to such a related practice. Take the learning of language as an analogy.
Once we already know a language, we may engage in learning another based on
the first language that we know. It makes especially sense to have grammar lessons
when we have already learned about the grammar of our first language. But this
first language we acquire by participating in using it; and only when we are already
familiar with the language do we learn about its grammatical features. Without
knowing a language, there is nothing that we can grammaticalize, and there would
not even be the means (tool) to grammaticalize whatsoever. Similarly, without
having done research we do not have anything that we could learn in terms of
methodsit is all words and no bodily and embodied practice of actually doing
research generally and doing data analysis more specifically.
This book is written in a way that takes its readers into the midst of rigorous data
analysis. It does so by actually featuring chapters in which an experienced instruc-
tor-analyst has taken on the task to figure out what the nature of this event is given
only the (mystery) transcriptions of some event. Rather then making up some in-
terpretation, he follows, like a detective and using the metaphor of the sleuth, what
participants in the transcription make available to each other. He listens to how the
participants themselves take up what someone else has said, investigates the kinds
of social relations that are exhibited in the back-and-forth or give-and-take of the
verbal exchanges; and, with this, he hypothesizes what the nature of the social
event might have been so that in the concrete event that had been transcribed the
particular talk was produced. He tests the emerging hypotheses, rejects some and
retains others until he states the most likely hypothesis as to the nature of the event,
which constitutes the end of the analysis. The participant in the graduate seminar
who had produced the mystery transcription at hand then judges / evaluates what
the instructor-analyst has derived, sometimes by simply describing the contents of
the video or its sourcee.g. a talk show, teacher training video, a documentary, or
daily news featureor shows the actual video clip.
These sessionsfeatured in part Bare interesting because the instructor-
analyst cannot fake competent and rigorous analysis. He cannot pretend knowing
rigorous data analysis just because he masters it symbolically, in and through the
mastery of a form of discourse. Instead, in real time and without time out, he en-
gages in the analysis knowing that someone in the audience will judge the outcome
 PREFACE ix

of the performance. We therefore can observe in the transcriptions of these data

analysis sessions the scientific modus operandi as it performs. We can observe the
feel for the game of data analysis that is expressed in what the instructor-analyst
does and how he does it. Because it is a seminar on qualitative research, he also
makes thematic what he is doing, or rather, what he has done after doing some-
thing in a particular way. To increase the potential benefits to the reader, I pro-
duced a running commentary on these performances, placed where ordinarily we
would find the footnotes and indexed to the particular place commented upon in
the transcribed analysis session by means of numbers (the text is in italics).

Victoria, BC
December 2014
Glossary

Accountable. People act in ways that they concurrently or subsequently describe

and explain when they are held to account for their actions. That is, they can pro-
vide reasons for what they have done. For example, one person may note that she
is insulted by what her counterpart said, who, in saying I was only joking pro-
vides an explanation of what he was really doing. This latter form of accounting
also falls under the practice of formulating.1
Bracketing. This term denotes the act of a social scientist who puts at bay his/her
own preconceptions that might come into play trying to understand a phenomenon.
Thus, rather than accepting power as a social construct, the brackets in the ex-
pression {power} indicate that the investigation concerns the very production of
conduct that leads social scientists to use the term power. That is, bracketing
orients the social scientist to study the work that makes power an observable
social fact.
Conversation analysis. The name for a particular analytic method that H. Sacks
(1995) originated and developed. It is not simply the analysis of conversations, as
novices in the field of social research often assume. That is, conversation analysis
always is the analysis of conversation, but not all analyses of conversations de fac-
to do conversation analysis. Conversation analysis takes the turn pair as its mini-
mum unit. This decenters the analysis from the psychological to the social level.
Documentary method of interpretation. The documentary method was intro-
duced by the sociologist K. Mannheim (Eng. 1952, Ger. 2004); it was subsequently
used, in somewhat modified form, by H. Garfinkel (1967) to explain how, based
on concrete examples from our lifeworld, we get a sense of something like a
worldview or a queue. Concrete incidences are taken to be documents of
something, like a worldview or queue, which, though general, only exists in
and through the manifold concrete experiences that we have with events that de-

1
Bold-faced and italicized words in the text of an entry refer to another entry.
xii GLOSSARY

serve these names. Neither addition nor synthesis nor abstraction can explain the
relation between the concrete document and the whole of which it is a document.
This is similar to the description Merleau-Ponty (1945) providesnow confirmed
in the neurosciencesthat we never perceive or know a cube as a whole but only
in the way it gives itself to our current position relative to it. What we do know,
because of many experiences with cubes, is what happens when we change our
position with respect to the object or turn the object in our hand. We know the cube
in and through the ensemble of concrete manifestations of the object we name cu-
be.
Dope (cultural, psychological). The term is used to denote the human being in the
way it appears in the psychological and sociological literature, who produces the
stable features of society by acting in compliance with pre-established and legiti-
mate alternatives of action that the common culture provides (Garfinkel 1967, p.
68). In sociology, school performance, for example, is explained by socio-
economic status; and in (Piagetian) psychology, schemas explain what a person
sees and does. Thus, peoples reasoned and justified/justifiable evaluations, judg-
ments, and decision-making are treated as epiphenomena when their actions are
explained by means of causal factors that determine (cause) their actions as if from
the outside.
Durkheims aphorism. The founder of sociology, . Durkheim (1919), estab-
lished what we now know as sociology in stating that the first rule, which is also
the most fundamental one, is to consider social facts as things. As things, social
facts are observable and to be treated as any thing investigated by the natural sci-
ences. Ethnomethodology is concerned with the work involved in making social
facts observable and accountable to every member to the setting rather than with
identifying, classifying, and explaining these things themselves, which it leaves to
formal analytic studies (e.g. Garfinkel 1996).
Endogenous. From the inside. The term is used in ethnomethodology to insist on
the fact that actors themselves conduct themselves in such ways that these not only
produce social facts and social conditions that in turn determine what they do but
also exhibit the accountably rational ways in which the facts are produced. The
social facts are not abstract schemas, rules, or practices. They are the results of
performances that the actors hold each other accountable for. Researcher-produced
and owned categories are exogenous: these do not matter to the societal relations,
such as, for example, the conceptions or schemas that some social researchers in-
voke to explain why people do what they do.
Ethnomethodology (EM). Literally the science of the methods everyday folk use
to produce social facts and situations in accountably rational ways. This social sci-
ence is concerned with exhibiting the ordered and orderly ways in which we pro-
duce society and societal relations as ordered and orderly phenomena. This is dis-
tinguished from formal analytic studies, which use special methods to describe
(abstract) patterns said to underlie what people do. In this, researchers constitute
forms of knowing generally invisible and inaccessible to those who actually pro-
duce the orderly ways in which society and its social relations appear. That the
methods are special can be taken from the fact that scientific articles require meth-
 GLOSSARY xiii

ods sections in which they describe what they have done and how they have done
it. If the methods merely were the common ways in which we do things in social
situations, the methods sections would be superfluous.
First-time-through. The adjective first-time-through is used to characterize a
form of analysis where the researcher takes the perspective of the social actors (or
cohort, staff of the phenomenon) studied. That is, researchers then have no way to
use future states, the outcomes of actions, to analyze earlier happenings in a teleo-
logical fashion. This changes the way in which researchers can work, for it is no
longer possible, for example, to say what a statement does, because what it will
have done is apparent only through its effects available in and from subsequent talk
and actions.
Formal analysis (FA). All forms of social research that have to specify the special
(scientific) methods of how they identified the reported social facts fall under this
category (Garfinkel 1996). Studies using formal analytic methods differ from eth-
nomethodological studies in the sense that the latter employ and demonstrate
competence of precisely the same methods that are used to produce social facts.
Formal analysis and ethnomethodology therefore are asymmetrical alternates with
respect to the study of social facts. The former focus on facts, using special re-
search methods to extract them from data, whereas the latter study the actual work
that produces what formal analysis identifies. The research methods of formal
analysis are so special that they need to be specified for each research project, in
part because the very object under study, its identification, depends on the method.
Knowing the ethnomethods always will allow us to get at the social facts, whereas
formal analytically identified social facts do not get us to the ethnomethods that
produced them.
Formulating. Speakers of natural language frequently formulate what they are
doing with language. It is one of the methods we use to bring order to and account
for order in everyday societal relations and, therefore, is something that is a form
of praxis considered by ethnomethodology. For example, a speaker might say Let
me ask you this, How much have you written today? Here, the actually intended
question is prefaced by the note that a question (rather than a statement, invitation,
or order) is forthcoming. The speaker has formulated an aspect of the ongoing
conversational work.
Glossing, glossing practices. In the ethnomethodological literature, this term is
used to refer to the fact that speakers always mean differently than they can say
in so many words. After many years of absence from Hattiesburg, MS, where my
alma mater is located, I was invited there to give a talk at the 100th anniversary of
the university. While driving through the center of town, I said to my host It cer-
tainly has changed since I was here, and my host responds, Yes, they revitalized
the entire city center to bring people back here. In this situation, the first sentence
is treated as a gloss and in the second sentence my host elaborated what I was real-
ly saying without actually having said so with the words I used.
IRE. This acronym stands for initiation, reply, evaluation, a sequence of speak-
ing turns frequently observed in classroom talk. In this sequence, teachers tend to
xiv GLOSSARY

take the first and third position (initiation, evaluation) and students tend to take the
second, middle position. In the middle position, students produce what the initia-
tion turn invited them to and what the third turn will be evaluating. It is an every-
day practice in some societal activities in which those present take particular slots
in a turn-taking routine that has three parts.
Lebenswelt (lifeworld). This term denotes the world as it appears to us and that
we inhabit (Husserl 2008). It is the familiar, self-evident, inherently shared world
given to us in our experience. It is not a world that we have to interpret and con-
struct, because construction (e.g., using language) is possible only as the result of
inhabiting and evolving in/with a Lebenswelt. Much of the initial research con-
cerning the Lebenswelt was done by the German philosopher E. Husserl during the
1920s and 1930s. A. Schtz extended this work into the field of sociology. H. Gar-
finkel, the founder of ethnomethodology, recognizes the contributions of both to
his own work.
Members (to the setting). Ethnomethodological investigations concern the en-
dogenous work by means of which social structures are observably and accounta-
bly produced. Because the phenomena are social, they are independent of the par-
ticular individuals (or cohort) that staff the phenomenon in any particular case.
Thus, although a researcher may use the queue in front of a movie theater as an
example, the phenomenon denoted by the gloss {queue} is independent of the par-
ticular cohort presently observed.
Method. The way in which something is actually done, as distinct from methodol-
ogy, which is the science or study (-logy) of methods. In research reports, we de-
scribe the methods on which our qualitative and quantitative analyses are based
rather than engaging in methodology.
Methodology. This term often is used inappropriately in the sense of method, even
though, in structural equivalence with all other sciences that include the word end-
ing -logy, methodology is the science of method. Ethnomethodology is such a sci-
ence, because it is concerned with the different methods people (ethno-) use in
everyday situations to produce the order of things. The term methodology is mis-
used in scientific journal articles, which state methods rather than engage in doing
{methodology}. Ethnomethodology truly is a methodology, because it aims at
understanding the different methods people employ to make the structures of the
social world visible to one another.
Phenomenology. Literally the science of phenomena, phenomenology focuses on
the work that makes the phenomena appear in the ways they do, that is, the work
by means of which things and events phenomenalize themselves. The term comes
from the Greek verb [phainesthai], to show itself. It has the same stem
as the Greek word [phos], light, so that the word phenomenon literally has the
sense of something that has come to light. The term phenomenology, mistakenly,
is frequently used to refer to the study of personal (psychological) experiences,
feelings, rather than to the fundamental processes of phenomenalization that lead
to this or that experience. Thus, we might see this or that cube looking at a flat
drawing (experience). Phenomenology is not interested whether we see one or the
 GLOSSARY xv

other cube or something else altogether. It leaves this to phenomenography. Ra-

ther, it is interested in the underlying processesa matter of eye movement in in-
dividuals or a matter of complex social representation practices in the case of sci-
encethat lead to one or the other experience.
Relational thinking. Much of research takes as its starting point the kinds of real-
ities that can be touched in the way we touch objects, such as social groups, social
class, and antagonisms between these. However, in some research, these objects
are recognized to be the product of societal relationsruling relations as the criti-
cal feminist sociologist D. E. Smith (1990a) calls them. Bourdieus (1992) recom-
mendation to study spaces of relation therefore resonates with the ethnomethodo-
logical concern for identifying the work that makes and exhibits the ordered and
orderly world; and it resonates with Foucaults (1975) recommendation to look at
knowledge-power relations that produce institutional distinctions rather than ac-
cepting the latter as self-same facts.
Sheffer stroke. The Sheffer stroke | is used to express that the two terms that
come to be paired through its use form an irreducible whole. For example, when
we write {question | reply}, this means that question and reply are parts, manifesta-
tions, of some whole. Like a handclap requires both hands, this unit requires both
of its part to be what it is. A statement is a question only because there will have
been a reply; and a statement is a reply only because there has been a question.
Such an approach is consistent with the call for relational thinking (Bourdieu
1992) as opposed to thinking in terms of realities that can be touched with the
finger (p. 228). The two turns specify each other. A turn on its own means noth-
ingas a famous baseball umpire aptly said, Its [a throw] nothing until I call it,
that is, until the umpire says that a throw was a ball or a strike. There is no ques-
tion on its own, or a reply on its own, in the way there is no one-handed clap. This
whole in fact turns out to be a joint, social act.
Staff. Ethnomethodological investigations concern the endogenous work that
accountably produces, for participants themselves, the surrounding social situation
and its facts. Because such phenomena are widespread, those who figure in a par-
ticular example are but the staff that in any this case brings the phenomenon to life.
Any other situation would involve different staff but the same social phenomenon.
A good everyday example is a queue. We find queues everywhere. Science les-
sons, independently of who teacher and students are, constitute another case. We
can walk into a classroom and know/see that there is a science lesson independent
of its current staff. Studies in ethnomethodology ask, What is the endogenous
work that makes a lesson recognizably a science [rather than mathematics, reading,
social studies] lesson?
Exergue

One of the functions of a seminar such as this one is to give you an opportunity to
see how research work is actually carried out. You will not get a complete record-
ing of all the mishaps and misfirings, of all the repetitions that proved necessary to
produce the final transcript which annuls them. But the high-speed picture that will
be shown to you should allow you to acquire an idea of what goes on in the privacy
of the workshop of the artisan or of the Quattrocento painteri.e., it will include
all the false starts, the wavering, the impasses, the renunciations, and so on.
(Bourdieu 1992, p. 220)

We must learn how to translate highly abstract problems into thoroughly practical
scientific operations, which presupposes, as we will see, a very peculiar relation to
what is ordinarily called theory and research (empirie). In such an enterprise,
abstract precepts such as the ones enunciated in Le metier de sociologue (Bourdieu,
Chamboredon, and Passeron 1973; English translation 1991), if they have the vir-
tue of arousing attention and putting us on notice, are not of much help. No doubt
because there is no manner of mastering the fundamental principles of a practice
the practice of scientific research is no exception herethan by practicing it along-
side a kind of guide or coach who provides assurance and reassurance, who sets an
example and who corrects you by putting forth, in situation, precepts applied di-
rectly to the particular case at hand.
(Bourdieu 1992, p. 221)
 PART A

Introduction
Social research is something much too serious and too difficult for us to allow our-
selves to mistake scientific rigidity, which is the nemesis of intelligence and inven-
tion, for scientific rigor, and thus to deprive ourselves of this or that resource
available in the full panoply of intellectual traditions of our discipline and of the
sister disciplines of anthropology, economics, history, etc. In such matters, I would
be tempted to say that only one rule applies: it is forbidden to forbid, or, watch
out for methodological watchdogs! Needless to say, the extreme liberty I advocate
here (which seems to me to make obvious sense and which, let me hasten to add,
has nothing to do with the sort of relativistic epistemological laissez faire which
seems so much in vogue in some quarters) has its counterpart in the extreme vigi-
lance that we must apply to the conditions of use of analytical techniques and to
ensuring that they fit the question at hand. I often find myself thinking that our
methodological police (pres-la-rigueur) prove to be rather unrigorous, even lax,
in their use of the very methods of which they are zealots.
(Bourdieu 1992, p. 227)
1

Rigor in Qualitative Data Analysis

We ask what it is about natural language that permits speakers and auditors
to hear, and in other ways to witness, the objective production and objective
display of commonsense knowledge, and of practical circumstances, practical
actions, and practical sociological reasoning as well? What is it about natural
language that makes these phenomena observablereportable, i.e., account-
able phenomena? For speakers and auditors the practices of natural language
somehow exhibit these phenomena in the particulars of speaking, and that
these phenomena are exhibited is itself, and thereby, made exhibitable in fur-
ther description, remark, questions, and in other ways for the telling.
The interests of ethnomethodological research are directed to provide,
through detailed analyses, that account-able phenomena are through and
through practical accomplishments.
(Garfinkel and Sacks 1986, p. 163)

Its just your construction. This is a comment I have frequently heard among
researchers with (radical, social) constructivist bent, as if social science research
was not a serious attempt to produce results that can be reproduced by others in the
way that is the case for natural sciences. At the heart of it, constructivist ideas
about method are connected with a dictum in the philosophy of science according
to which the only principle that does not inhibit progress is: anything goes
(Feyerabend 1993, p. 14). Anything goes because human beings interpret and con-
struct, in their individual mind, a conceptual framework that can be tested, at best,
for its viability in the world. This is so both for the natural world and the social
world, where we have to engage in the construction of intersubjectivity. But any-
thing goes does not mean nimporte quoi, simply anything. Human beings do, in
accountably rational ways, whatever is required to make an endeavor successful.
Interestingly, Feyerabend points out that there is not a single rule, however plau-
sible, and firmly grounded in epistemology, that is not violated at some time or
another (p. 14). Method, in fact, requires adaptation to the situationto the object
of inquiry that has to be constructed in the process of inquiry (Bourdieu 1992).
This means that there cannot be a method that can be specified a priori for investi-
gating some phenomenon. Again, this does not imply that anything goes. Instead,
8 CHAPTER 1

the social scientist as much as the natural scientist is held to undertake research in
rigorous but not rigid ways. But rigidity is precisely what classical courses on
method generally and the adherence to one or the method specifically advocate, as
can be observed in so many instances of (junior) researchers who know only one
way of doing research. Thus, it has been noted that
[i]t is revealing that entire schools or research traditions should develop
around one technique of data collection and analysis. For example, today
some ethnomethodologists want to acknowledge nothing but conversation
analysis reduced to the exegesis of a text, completely ignoring the data on the
immediate context that may be called ethnographic . . . not to mention the da-
ta that would allow them to situate this situation within social structure. . . .
Thus, we will find monomaniacs of log-linear modeling, of discourse analy-
sis, of participant observation, of open-ended or in-depth interviewing, or of
ethnographic description. Rigid adherence to this or that one method of data
collection will define membership in a school, the symbolic interactionists
being recognizable for instance by the cult of participant observation, eth-
nomethodologists by their passion for conversation analysis, status attain-
ment researchers by their systematic use of path analysis. (Bourdieu 1992, p.
226, emphasis added)
Rigorous data analysis does not arise when one method is applied indiscrimi-
nately to all situations of interest. If researchers have but one method at their dis-
position, then they are acting like the craftsperson who has only a hammer and to
whom the whole world looks like a nail or like a place where nails are driven in.
Of course, we may always argue to be interested in what a research method allows
us to discover. But this appears to be putting the cart before the horse, because the
problems of the social-psychological sciences tend to be found or identified in and
by the various societal arenas. A case in point is that of a junior colleague, who, as
a mother-to-be, was interested in finding out about that phase in a womans life by
researching the conversations in an online forum that she would create. She wanted
to collect data as the forum was growing, and new mothers-to-be would be joining
and others would be leaving. There was a problem, however, because the only re-
search methods that she was familiar with were of quantitative nature, which, in-
herently, required that the information from the informantsmothers-to-bebe
collected under the same conditions at a particular point in time to guarantee com-
parability. That is, in the discussions with methodologists, she came to realize that
her research question could not be answered by the only method she knew, statisti-
cal comparisons on the basis of questionnaires and measures of social-
psychological characteristics. She abandoned the research project despite the tre-
mendous (vested) interest she had in the problem. Because of her self-declared
membership in a particular schooli.e., circumscribed by the use of quantitative
research methodsshe could not investigate what she was really interested in
understanding. What a pity!
It has become a truism that theoretical and methodological choices are interde-
pendent and therefore cannot be disentangled; and these define what comes to be
accepted and admissible as data. What is more important than methodological ri-
 RIGOR IN QUALITATIVE DATA ANALYSIS 9

gidity is methodological rigor, for, as found in the quotation that opens this part A
of the book,
social research is something much too serious and too difficult for us to allow
ourselves to mistake scientific rigidity, which is the nemesis of intelligence
and invention, for scientific rigor, and thus to deprive ourselves of this or that
resource available in the full panoply of intellectual traditions of our disci-
pline and of the sister disciplines of anthropology, economics, history, etc. (p.
227).
Bourdieu calls on social science researchers to watch out for the methodological
police (watchdogs) that tend to foster a climate of rigidity rather than rigor. At the
same time, we need to listen to his warning that the extreme liberty, which mir-
rors Feyerabends anything goes, should not be taken as the relativistic laissez
faire that is so much in vogue in some quarters (p. 227). In the hey-days of
(radical) constructivism of the 1980s and 1990s, its fervent advocates were often
accused of (methodological) relativism. This may not come as a surprise given
what could be observed with respect to method captured by Bourdieu as laissez
faire. The opposite of laissez faire, however, is not rigidity, methodological or the-
oretical. Instead, we need to do research in rigorous ways, which in fact requires us
to do it in the same accountably rational ways that ethnomethodologists have ob-
served ordinary people to employ in the course of accomplishing their mundane,
everyday affairs.

Criminal and Legal Affairs as Metaphor for Social Research

Constructivist epistemologies often lead their advocates to suggest that what we

make of a situation, some phenomenon, is the result of our individual and social
constructions. Explicit or implicit in such statements is the presupposition that the
constructions could be otherwise. In the more extreme versions, this leads to the
anything goes and laissez faire approaches that other researchers have come to
decry. The methodological debate is interesting in the face of criminal and legal
investigations that very much employ interpretive methods and yet are expected to
reveal the truth. When truth is not revealedwhich sometimes comes to be
known during subsequent trials or retrials where the innocence of a (wrongly) ac-
cused person comes to be establishedthen this frequently is attributed to errors of
omission or commission. Trust in the criminal and judicial systems is equivalent to
trust in the interpretive methods that are expected to lead to the truth about events,
who did what, when, where, and how. We know that the consequences of error are
serious: those guilty of a crime remain free, are acquitted because judged innocent,
and those innocent may come to be arrested and subsequently tried as perpetrators.
In the more serious instances where jurisdictions have retained the death penalty,
the guilty verdict may lead to death row or immediate execution.
Factual or fictional police and detective work, and the associated interpretive
work in courts of law, serve me as useful metaphors for orienting the ways in
which I engage in social research independently of the nature of the methodsi.e.,
10 CHAPTER 1

the formal analytic (qualitative or quantitative) approaches or ethnomethodology.

Even if none of my participants or the social groups under investigation is going to
die or going to be affected in any direct way because the research results are pub-
lished in different communities, I approach research in the spirit of criminal and
legal investigations. There are deontological consequences to what I do, because I
am aware thathowever minimalthe very discourses of society change with my
publications. There is a responsibility to society at large, the people that come to be
affected through the employment of theories and concepts developed in social re-
search, and the available discourses for describing and intelligibly accounting for
real-life phenomena. That is, the argument that it is only your construction no
longer holds. An anything-goes attitude and the methodological laissez faire that
accompanies it put researchers in ethically untenable positions. The research we do
needs to be conducted in accountably rigorous ways, aiming towards an ideal
though practically never achievable certainty about concrete social phenomena, in
the way we experience these in clearly recognizable ways, and the manner these
are described and theorized on the ideal plane.
What researchers publish enters and changes the discourses we have for describ-
ing and theorizing social/psychological phenomena. This became apparent to me
during one of my first courses on research method that I taught. In this course, a
graduate student talked about her research interest in adult children of alcoholics
driven by her experience of being a child of alcoholics. Nobody else in the class,
including me, had heard of such a phenomenon before, and we all listened to the
graduate student report on existing research findings and describe the phenome-
non.1 By the next meeting, one quarter of the students had begun to tell their own
biographies in a new way, having recognized that they, too, are examples of adult
children of alcoholics. I was able to trace the concept to Woititz (1976), whose
doctoral dissertation concerned self-esteem problems in children of alcoholics.
In creating the concept, the author thereby provided a discourse for describing and
understanding a range of manifestations that appear to be frequent among those
whose parents drank a lot. We see that the associated discourse may change not
only the ways in which people understand their past lives but also the ways in
which they orient towards the future conduct of their lives and, therefore, in the
concrete ways in which they act. Research has changed available universe of dis-
course and the actual lives of people. Today organizations have formed, such as the
Association for Adult Children of Alcoholics, with chapters in different cities and
regions in North America.
Concepts are pervasive, and those familiar with them may be able to identify in
consistent and reliable ways, other individuals falling into the same categories. A
researcher interested in doing some related project today may likely find it unprob-
lematical to name the topic and phenomenon and proceed to identify participants
with specific characteristics consistent with the phenomenon. Precisely here lies
one of the dangers, which Bourdieu (1992) describes as the pervasive presence of

1
When the researcher intends to investigate a phenomenon in which she is interested in by
definition, it becomes all the more difficult to avoid falling into the trap of the precon-
structed object (Bourdieu 1992, p. 231). This interest itself falls on the blind spot of the
investigation because the veritable principle of this interest remains unknown.
 RIGOR IN QUALITATIVE DATA ANALYSIS 11

the preconstructed. As Marx showed in his critique of political economy, terms

such as wages, commodities, capital, profit, and the likes were around before polit-
ical economists came alongsuch as Ricardo, who was one of the targets of the
critique (Marx/Engels 1962). The latter than took up these terms and imported
them into the theory to be built without further analysis. That is, the phenomenon
did not exist as such prior to its first articulation and prior to the associated dis-
course that named it and its characteristic particulars. It was the result of a scien-
tific research project. It eventually entered society, becoming a recognized social
fact to which patients and clinicians orient in active ways. Yet, qua research result,
it is the outcome of a constructive effort. When researchers use the discourse today
as if it were an indisputable fact, then they are therefore subjecting themselves to
the preconstructed. As a result, social scientists working in the field therefore liter-
ally do not know what they are doing because they operate with preconstructed
concepts: they find what members of their community have put in place. They rei-
fy rather than critically interrogate the phenomenon.
The dangers in such reification were exhibited by the feminist sociologist D. E.
Smith (1990), who not only raised a child on her own but also understood her
family in terms of the sociological concept of single-parent family that is contrast-
ed to the standard North American family. Not only was her understanding affect-
ed by the discourse but also her relations with the school when her child was in
trouble. School officials and Smith, as parent, understood the problems in terms
of the characteristics ascribed to single-parent families and their children. As a
result, the suggested actions and consequences were framed within this discourse
without questioning that the concept itself was the result of earlier sociological
research. Doing research in a rigorous manner means not only focusing on the re-
search object and the consequences of the research in its characterizations but also
critically investigating the very means (tools, instruments, methods) by which the
object and its discourse come to be produced.
Returning to the criminal and judicial metaphor, this means that researchers
need to be concerned not only in the what of their studies but also in the means
(tools, instruments, methods) of conducting them. That is, the first and most
pressing scientific priority, in all such cases, would be to take as ones object the
social work of construction of the pre-constructed object (Bourdieu 1992, p. 229,
original emphasis). It may come as little surprise to hear the analyst in part B of
this book repeatedly question why he has this or that sense or why he comes to
hear people say some rather than another thing. This (historical) investigation into
the origins of the pre-constructed object or research is where the point of genuine
rupture is situated (p. 229).
The criminal and judicial metaphor is interesting from another important aspect
in the praxis and theory of method. In many scholarly communities and the associ-
ated research journals, researchers are exhorted to provide multiple examples and
proof of the pervasiveness of some phenomenon. A single case or a small number
of cases treated analogically frequently are treated in the peer review process as
insufficient evidence. In the criminal and judicial situation, each case has to be
treated as such. Not only does a single case have to suffice in the criminal and ju-
dicial situation but also in research, where a single case, perhaps accompanied by
one or two analogical cases, may suffice to identify invariants. The analogical cas-
12 CHAPTER 1

Fig. 1.1 The logic relating the general principle, on the left, and the (eight) specific cases,
on the right. Although all eight cases are different, they share properties when considered at
different levels of abstraction.

es would serve as test cases where the system of relations established on the basis
of the original case come to be validated. This poses a challenge, which consists in
systematically interrogating the particular case by situating it as a particular in-
stance of the possible . . . to extract general or invariant properties that can be un-
covered only by such interrogation (Bourdieu 1992, p. 233). It is in precisely this
manner that the laws of psychology of art was derived on the basis of the analysis
of one fable, one drama, and one short story (Vygotskij 2005). The author empha-
sizes that it he was creating a psychology of all art, including music, painting, and
other art forms rather than a psychology of writing or a psychology of the short
story. He emphasizes that the most developed forms of art, the most difficult ex-
amples of a genre, the monsters, point to the most general principles, the essence
of art. These general principles work themselves out in very different, often con-
tradictory ways in the concrete cases considered. The relationship between any set
or all possible cases is that of a family. In the same way that parents and their chil-
dren may share few, if any, behavioral and bodily characteristics common to all
members, the analogical cases considered in research may differ as much as do the
siblings in a family. But in the way all siblings derive from the same common an-
cestors, the abstract less developed properties underlying the phenomenon of inter-
est manifest themselves differently in the concrete cases at hand. The underlying
logic is represented in Fig. 1.1. Thus, all eight cases can be led back to one under-
lying (genetic) principle, which works itself out (concretizes, specializes) different-
ly in the concrete cases. Case 1 and case 2 share more observable characteristics
than do case 1 and case 8, for example. In fact, case 1 or case 2 and case 8 may not
appear to share any characteristic with each other and yet are particulars of the
same principle. Thus, the fable, drama, and short story in the study of the psychol-
ogy of art may be case 2, case 3, and case 4, all falling under the instances of writ-
ten art forms. Case 5 and case 6 might be different European musical forms and
case 7 some African music. Case 8 might represent painting. Vygotsky derived a
psychology of art, which, therefore, would be represented by the line on the very
left of the Fig. 1.1.
 RIGOR IN QUALITATIVE DATA ANALYSIS 13

Police work requires understanding the kind of situations that might possibly
have led to the case in hand as well as the contextual particulars that make this case
different from every other case. Approaching data analysis in this manner allows
us to establish invariants as much as those particulars that make this case indistin-
guishable from every other case. We may consider the example of voices, for the
moment. Using voice analysis software, we may identify particular pitch levels,
speech intensities, pitch contours, and speech rates. However, all of these charac-
teristics do not distinguish one voice from another. For example, in one research
project I was able to show that two individuals teaching together for several
months came to speak in similar ways such that there was little difference between
how they used and articulated the word right in very different situations (Roth
and Tobin 2010). Moreover, when they were speaking, their speech rates and pitch
levels were adjusted to those of the others, so that one person always took up with
values of those parameters where the other had left off. And yet: on the telephone,
I could easily distinguish with whom I was talking. That is, there was something
very particular about the voice of each of the two individuals that still allows me to
distinguish one voice from another without trouble. The particulars that allow us to
do so are known as timbre, and timbre is precisely what cannot be captured and
extracted by voice analysis.
Criminal and judicial work require a good understanding of the invariant prop-
erties any this case has in common with other cases and the particular properties of
this case that allow the process to narrow the possibilities to the real perpetrator. In
social analysis, we are also interested to understand invariants and particulars asso-
ciated with a case. Comparative analysis, both between analogical phenomena and
within a phenomenon, is a way of articulating invariants and particulars. This may
be done in the form of tables, where we list, for example, the different cases or
different instances within a case in a vertical manner (giving rise to different rows
in our table) and each identified characteristic in a horizontal manner (yielding the
different columns of our table). Whenever we identify a new characteristic in a
case or instance, then not only do we add the column and make a check mark for
the case but also we return and consider all the other cases or instances as to the
presence or absence of this characteristic. In the simplest of cases nevertheless of
considerable value is a 2-by-2 table. Consider the example I produced during a
research project on students epistemological discourse concerning science and
religion (Roth and Alexander 1997).
In the students interview texts or in their interview talkboth understood to be
the results of specific societal activities and associated forms of language (i.e. lan-
guage gamesthere are stretches where the discourse concerns science whereas
other stretches concern religion (Fig. 1.2). These are the two domains of interest. In
each domain, knowledge was designated to be subjective or rational. Depending on
the context, in each of the four quadrants formed (Fig. 1.2), the knowledge was
declared to be a social construction or absolute. But this classification was not suf-
ficient. Thus, a student may have stated the claim that there are subjective elements
in science, making it irrational and unreliable to a certain extent, and that science
was rational. My analysis reveals that in these instances, there was something like
a truth-will-out device, which made it possible for individual scientists to be sub-
jective and science as a whole to eventually yield the truth. I also identified con-
14 CHAPTER 1

Fig. 1.2 A 2-by-2 table that allows classification of the different discursive patterns that
high school physics students used when talking about epistemological issues.

flicts between ways of talking about science and religion. Thus, some students
stated the claims that both science and religion produce truth and then found them-
selves confronted and conflicted by incompatible talk about the origin of the world
as it is today. The students often said that science and religion were incompatible.
To test the theoretical model derived in the database covering high school stu-
dents I used it to analyze texts published in the journal Zygon, a forum for bridging
science and religion. It turned out that the model required further adaptation, as
there was a feature that did not become apparent in the analysis of student dis-
course: the discourse of incommensurability (Fig. 1.2) was mobilized to account
for the differences between scientific and religious ways of articulating issues,
those in the deontological realm particularly. Had I followed the advice to study
the most developed or hardest situation (Vygotskij 2005), the discourse of scien-
tists and theologians, I might have immediately identified all the dimensions that
the model ultimately contained.
The point then is to think the case relationally, as a particular instance of the
generalized possible. Thus, when the general model for theorizing epistemological
discourse concerning science and religion is used in particular cases, then it works
itself out differently. A student, whom we may classify as a social constructivist,
considered science as a language game so that knowledge both on the individual
subjective and public-shared levels is constructed (Fig. 1.3a). Not only individuals
but also social groupsor scientific disciplinesexhibit subjectivity, which ex-
presses itself, for example, in the forms of cultural, subcultural, or national biases.
Similarly, there is a rational side to religious experience in the shared public realm,
but the personal experience of revelation and the sense of spirituality is absolute.
As a result, the epistemological discourse of the person concerning science and as
a whole is represented in Fig. 1.3a. In contrast, a student talking about the subjec-
tive elements of doing science and about the objective nature of science makes
contradictory statements, which were resolved by means of a discursive device
called truth-will-out device (Fig. 1.3b). Because in the particular discourse config-
 RIGOR IN QUALITATIVE DATA ANALYSIS 15

Fig. 1.3 The general model (Fig. 1.2) works itself out differently in each concrete case. a.
A student espousing a constructivist epistemology. b. A student, though recognizing indi-
vidual fallibility, espousing a realist epistemology and experiencing a conflict between sci-
entific and religious discourse about controversial issues.

uration, the student talked about science and religion as representing absolute
forms of knowledge, he found himself conflicted between the discourse of science
and the discourse of his church with respect to issues such as the origin of the hu-
man species, abortion, cloning, and other social issues with different ethico-moral
discourses of science writ large and religion.
It is precisely the familiarity with or the working out of the invariants (invariant
laws) that prevent analysts to drown in the particulars of the case, a situation that
frequently arises when novice researchers delve into the specifics of the data they
have or are in the process of collecting data. The questions that I often hear is
something like Where do I start? and How do I make sense of this all? Gener-
alization, used here, does not arise from extraneous and artificial application of
formal and empty conceptual constructions, but through this particular manner of
thinking the particular case which consists of actually thinking it as such (Bour-
dieu 1992, p. 233).

Towards an Ethnographically Adequate Account

There is a close similarity in the primary problem facing ethnographers and

other persons engaged in everyday life. This problem, common to both, is the
necessity of achieving a working consensus with others about what is going
on in any scene available to their senses. (McDermott et al. 1978, p. 246)
In the preceding section I suggest taking the metaphor of the detective trying to
find the criminal for thinking about doing data analysis. Using this metaphor ori-
ents us to rigorous ways of going about our work as researchers. But much of re-
16 CHAPTER 1

search is not about finding out who done it but in describing some aspect of the
social world. One way of thinking about rigor in data analysis then is in terms of
the production of an ethnographically adequate account of events. In fact, the two
ways of thinking about doing data analysis go together because the ethnographer,
as the criminologist, is concerned with the adequacy of an account. In the introduc-
tory quotation to this section, the authors point out that the ethnographers problem
is not unlike that of the participants in the situations observed: what is going on in
the scene that presents itself to the senses. In other words, the ethnographers task
is that of the person having stepped out of the house and onto the street rather than
observing street life from above and through a window, separate from the pulsing
stream of life.
In part B of this book I present what an expert analyst does when presented with
transcriptions of which he does not know the origin of. He therefore is forced into
something of a double role, the detective and the ethnographer. The analyses, pro-
vided under the real-time constraints of an ongoing graduate class, are those of a
person who has stepped into some situation. In that situation, people are in the
midst of the things they do. The analyst is trying to figure out what is going on
without asking more background informationwithout being able to askand is
therefore working with nothing other than what the people themselves make avail-
able. How is this possible? Certainly not because what we do is just a construction.
It is by following what people do. It is possible because people manage concerted
activity only by constantly informing and conforming each other to whatever it is
that has to happen next (McDermott et al. 1978, p. 246). It is by closely attending
to what people do and how they do it that the rigorously working expert analyst
recovers the nature of an event from a piece of transcript even without having any-
thing else at the disposal. The purpose of the analysis, therefore, lies in using the
ways members have of making clear to each other and to themselves what is going
on to locate to our own satisfaction an account of what it is that they are doing
with each other (p. 247, original emphasis). There are four aspects, often of the
same kind of behavior that people display to each other for the purpose of organiz-
ing the setting:
Members usually reference or in some way formulate some of the contexts for
their behavior. (p. 247)
Members usually organize their postures to form a configuration or positioning,
which signals the contexts for behavior. (p. 248)
Members behaviorally orient to the order in their concerted behavior and ac-
cordingly constitute and signal their contexts for each other. (p. 249)
Members usually hold each other accountable for proceeding in ways consistent
with the context for their concerted activities. (p. 250)
The adequacy of an ethnographic account derives precisely from describing how
members do what they do to be able to do what they do. Here, it is not anything,
any just-so construction on the part of researchers, for if they were to behave in the
setting in the way they describe, then others would find their behavior out of place.
For me one of the telltale signs of an inadequate account is that often provided by
individuals doing conceptual change or constructivist research. They describe what
their participants do in ways that we do not think and act. For example, we do not
 RIGOR IN QUALITATIVE DATA ANALYSIS 17

construct meaning or conceptual frameworks when we participate in a conversa-

tion. We just talk. Even though there may be fleeting thoughts in our private expe-
rience, there generally does not tend to be sufficient time to stop and hang on to an
idea, for if we do, we are out of synchrony with the other participants and no long-
er participate in the conversation. Thus, ethnographically adequate accounts also
require the researcher to focus on data in a different way. We cannot just make
something up about how people go about doing what they do and make up just-so
narratives about how they do it. A most striking example for me has always been
that when individuals are telling the interviewer during research interviews that
they had never thought about something and yet talk about the topic without much
hesitation. In the conceptual change and constructivist paradigm, there should be
conceptual frameworks that underlie what a person says and which are externalized
in and through the saying. But never having thought about some topic means that
the research participants could not have constructed a relevant conceptual frame-
work. This, therefore, is a good example of a research accountgenerally about
some science-related phenomenonthat is not adequate.
Ethnographically adequate description, therefore, is objective in the sense that it
can stand in for the original for those who cannot make the observations them-
selves (Smith 1981, p. 314). This account is objective not in the sense that it says
everything, from a single gods-eye perspective, but that it says what the ethnog-
rapher needs as description to understand the situation of interest. Thus, the ethno-
graphically adequate description must describe the observed rather than the ob-
server, in other words, it does not distort the original in ways which are products
of the observers particular perspectives or interests (p. 314, emphasis added).
As a way of entering the problematic issues raised throughout this book, consid-
er the following transcription that the experienced instructor-analyst featured in
part B analyzes for his students in real time.2 The transcription begins with a
statement about landforms, together with the indexical adverb here, which is a
reference to where the speakers are (turn 01). That is, wherever they are at that
instant, it is remarkable for its landforms. These landforms are remarkable because
they really look like other things. Three speakers reply using the affirmative ad-
verb yea (turns 0204), which a closer hearing and transcription may be able to
identify as an acknowledgment of the statement or an indication that they are fol-
lowing the speaker.

Fragment 1.1 (From Transcript 1, Heidi, David Suzuki)

01 Heidi: Its amazing how the landforms here really look like other things other
than rocks. Do you notice that sometimes?
02 Amanda: Yeah
03 Ashley: Yeah
04 Michael: Yeah
05 Heidi: Its funny
06 David: As long as youve got a good imagination

2
The transcripts analyzed in part B can be found in their entirety in appendix A. The gradu-
ate students attending a seminar course in qualitative data analysis produced these tran-
scripts. These are reproduced here, for purposes stated in in introduction to part B; and only
the turn numbers have been added.
18 CHAPTER 1

07 Heidi: Well, weve noticed that maybe, yeah, maybe lets go down this way.
We noticed that the longer youre out in the badlands, and the hotter it
is, the more things look like things

At this point and without any further access to the videotape that was tran-
scribed, we do not know the postures. But in turn 07, the phrase maybe lets go
down this way may in fact be an invitation produced to take one path rather than
another while walking about in nature. Watching the video, we might be able to
see additional aspect of the behavioral orientations of the speakers that rests upon
their postural orientation and the configurations that position them towards the
direction to be taken next in this walk-about. In turn 07, the speaker Heidi also uses
the plural pronoun we together with the verb notice. Because that same speak-
er is talking about things that the other participants apparently are not familiar
with, the we does not in fact include them. We then are introduced what might
be the name or type of the place: badlands. This configuration of people are out in
the badlands, and it is hot, or a place where it is often hot so that one can see
things, perhaps the landforms, as (other) things.
In the pair {turn 05 | turn 06}, we find a statement coupled with an evaluation or
qualification: {It is funny | As long as youve got a good imagination}. In fact,
we can hear this turn pair as a form of holding others accountable. It is funny is
an account of something, which we do not quite know at this time, which may be
what is articulated next on the part of the same speaker, that is, the statement that
the longer one is out in the badlands, the more things look like things. That is, the
it is funny formulates what is to come in a particular way, and the speaker is held
accountable for this way of framing in the evaluative turn 06: as long as one has a
good imagination.
Readers observe that the analysis works without having recourse to the back-
ground of the physical and societal situationthe history that has brought these
individuals together here in this place and for some reason. The analysis can do so
because of the self-explicating nature of talk outlined and specified in the four cri-
teria for ethnographically adequate description for concerted activities, which al-
lows us to recover a lot of what is going on. We further notice that Heidi appears to
be the person familiar with the setting, she and those that are included in the we
are sufficiently often out in this part of the world to know that when it is hot, the
more things look like things. What is said is articulated not merely for itself. In-
stead, it is said to the others and for their benefit in ways that are presupposed to be
intelligible. This means that these people are not normally out there in the
badlands: perhaps even for the first time. They may in fact be visitors to that place
referred to as the badlands.
In this brief introductory analysis, I have nothing confabulated; I have not spec-
ulated about something not immediately given to all the participants, such as meta-
physical meanings, concepts, or ideas in their minds. What has been stated
about the talk in this concerted activity was not just my construction. I did not
need any special methods to state what I have stated. Instead, by rigorously going
about what the members to that setting make available to each other for doing
whatever they do, I can recover what kind of situation it may be and where the
event may be occurring even without having access to what an ethnographer nor-
 RIGOR IN QUALITATIVE DATA ANALYSIS 19

mally has access to. Rigor here is another way of saying that we attend to what
speakers and their recipients make available to each other and we reject specula-
tions about what is in the minds of participants when this mental content is not
made explicitly part of the exchange. Rigor does not arise from a description of
whatever special method we describe but in the way we attend to what is relevant
to the participants in the situation at hand. What is relevant is not just some myste-
rious thing but instead is highlighted by the members for the benefit of each other.
The kind of rigor is different from the one that the descriptions of special methods
aim at, for example, in the following quotation.
All interviews with students were transcribed, and a coding schema was de-
veloped and refined in an iterative process, in which categories emerged from
the data (Lincoln & Guba, 1985). An extensive coding schema was utilised,
although the current discussion focuses primarily on findings corresponding
to the cognitive dimension of students responses. Cognitive codes ranged
from statements about how the exhibits functioned, to descriptions of the
phenomena that were observed, to their understanding of the underlying sci-
entific principles that the exhibits demonstrated. Further sets of codes were
developed in order to capture the affective dimensions of students interac-
tions with exhibits and responses to the visit overall. Affective codes were
generally used to categorise students reasons for considering exhibits as en-
joyable, fun, or interesting, such as the fact that an exhibit was challenging or
allowed for hands-on interaction. A list of all code definitions is available
from the authors. In addition, transcripts were also coded to indicate whether
students were referring to a photo or a video at any given point, so that any
differences in their responses based on the type of stimulus could be noted.
(DeWitt and Osborne 2010, pp. 13701371)
In this description, the authors no longer exhibit a concern for the ways in which
the children they interviewed oriented towards the interview task and the inter-
viewers. This would have been important because the children, in their replies,
respond to that context as a whole, doing and saying what they do for the purpose
of their participation in a concerted setting, which apparently was a stimulated re-
call interview session. Because of the nature as a concerted activity, not just any
response will be appropriate, and members to the setting orient to and hold each
other accountable for the underlying concerted activity that they produce in and
through their mutual orientation. Instead, the authors write about using and produc-
ing a particular coding schema, that is, they describe doing something that we do
not do in everyday life when interacting with the others. That is, although the par-
ticipants in these situations mutually oriented toward each other, talked for the
others benefit, using a language that is mutually intelligible, special methods are
described here to suggest that something was extracted from the transcriptions that
could not otherwise be extracted. There is therefore a problem of the relation be-
tween the researchers descriptive language and the original situation of which it is
intended to be a description (Smith 1981). It has been suggested that in its sim-
plest form [this] is the problem of what enters into the work of coding or categoriz-
ing other than the properties of the event and is either a property of the observa-
tional process itself or the conditions of the setting in which the description is
20 CHAPTER 1

done (p. 315, emphasis added). We might ask ourselves: so what can a special
method extract from transcriptions that the interview participants do not already
understand in some intimate way?
Is the description of method intended to convey that only these special methods
will be extracting whatever the cognitive refers to, or all those other codes said
to capture the affective dimensions of students interactions with exhibits? If
such dimensions can be extracted from the transcriptions do these not require com-
petencies of the same kind that interviewerinterviewee already exhibit to each
other? Does the categorization of students reasons for considering exhibits as
enjoyable, fun, or interesting not require the same competencies as the display of
such reasons? Does presenting museum visits as enjoyable, fun, or interesting not
require competencies that are also required of the analyst for recognizing the de-
piction of visits as having been enjoyable, fun, or interesting?
In considerations of the relationship between how researchers describe some
social phenomenon and the reality of this social phenomenon, we might consider a
pragmatic characterization of the difference between the philosopher (theorist) and
what people do in their everyday lives:
When philosophers use a wordknowledge, being, object, I, prop-
osition, nameand try to grasp the essence of the thing, one must always
ask oneself: is the word every actually used in this way in the language-game
which is its original home?
What we do is to bring words back from their metaphysical to their every-
day use. (Wittgenstein 1997, p. 48 [116])
In this quotation, the we refers to pragmatists, who locate the words in the
contexts of their everyday use. This clearly is the methodological move that we
observe in a field denoting itself as discursive psychology (Edwards and Potter
1992), where there is no longer an interest in such things as knowledge when
applied by a psychologist-researcher. Instead, the interest lies in how people use
some psychological concept to get their work done. Take, for example, the concept
situation awareness, which is frequently used by researchers to theorize perfor-
mance from a human factors perspective. When such researchers use the term, then
it is within a particular scholarly context, linked to other theoretical terms. Discur-
sive psychologya theoretical orientation and research methodshows little in-
terest in the theories of situation awareness and how it relates to other human fac-
tors concepts, including management, decision-making, and factual knowledge.
Instead, it is interested in how people in the field make use of such terms to ac-
complish their work (Mavin and Roth 2014). Thus, an examiner of pilots might
say,
Situation awareness for me probably is the most important thing. Some-
bodys got to be able to show me that theyre pretty much on, got a good feel
for whats going on in the environment around them at all times and that they
dont lose it. So we go and watch those things, we got a form with some as-
sessment markers that . . . you can refer to, which help, and ultimately youll
just make the assessment. The end result that comes is an element of gut feel
in it, which is difficult.
 RIGOR IN QUALITATIVE DATA ANALYSIS 21

Here, the term is both used and described in terms of what the examiner does at
work, which he is in the process of explaining to a researcher who also is an expe-
rienced pilot and examiner. In essence, then, research concerned with such con-
cepts constitutes an ethnography of the language in play as part of a larger lan-
guage-game, a form of ordered, orderly, and order producing activity.
In sum, therefore, the rigorous classification and categorization of talk itself
requires competencies that remain unstated and unarticulated. Moreover, there is a
problem in that analytic method, which attributes what is said during interviews to
the children when in fact that talk is talk between participants. As one social psy-
chologist noted, a word is impossible for one but a possibility for two (Vygotskij
2005). In the verbal exchange, language that has come from the other is produced
for the other and, in this way, returning to the other. Reducing the talk from an
exchange to the individual speaker fails to account for the fact that the same words
simultaneously ring in the ears of the recipient (Roth 2014c). The ethnographical
adequacy requirement aimed at in the movement towards rigorous forms of data
analysis does orient researcher to the concerted nature of the effort of making so-
cial events, which is inconsistent with the reduction of talk to the individual.

Unit (of) Analysis

Over the years of doing research, I have come to realize that the unit of analysis is
perhaps the most underrated and least-attended-to aspect of research generally and
data analysis more specifically. What we choose as the unit of analysis determines
what we find and how we understand it. Thus, for example, if the individual mind
is the unit required for understanding social processes, then the totality of some
event is reduced to the intention of individuals and the contents of their minds. The
phenomenon, such as a meeting, is then re-constituted as the sum total or inter-
action total of the parts thought to be identities existing in and for themselves. This
position, in other words, affirms the decomposability of the world into individual
facts:
1 The world is everything that is the case.
1.1 The world is the totality of facts, not of things.
1.11 The world is determined by the facts, and by these being all the
facts.
1.12 For the totality of facts determines what is the case and also all that
is not the case.
1.13 The facts in logical space are the world
1.2 The world divides into facts.
1.21 Any one can be the case or not be the case and everything else re-
mains the same.
2 What is the case, the fact, is the existence of atomic facts.
(Wittgenstein 1978, p. 11)
Here, Wittgenstein articulates the essence of the way in which the world is
thought and researched. Embodied in this approach is traditional logic, as shown in
22 CHAPTER 1

point 1.21 with its statement about identity, non-identity, and the impossible of a
third position. A dialectical conception, instead, recognizes that the world ultimate-
ly does not exist of two different substances, thought and (material) extension, but
two attributes, two different manifestations of one and the same substance
(Ilenkov 1977). Whenever we do research, we therefore have to ask the question,
what is the smallest unit that exhibits all the characteristics of the phenomenon.
Even though such a unit exhibits parts, manifests itself in different ways, these
parts, these manifestations, are not elements into which the whole can be decom-
posed and from which it can be (subsequently) re/constituted. Instead, there is a
whole-part relation that cannot be understood by investigating and abstracting from
properties. Take the following example of five members of the same family A
through E with five properties (predicates) a through e.

A B C D E
bcde acde abde abce abcd

A quick look shows that there is no one common attribute that all members
share. It would be impossible, therefore, to derive some common ancestor based on
the properties, for there is nothing in common to all. We could also think about this
example in terms of the classical ways thinking and thinking about concepts (Kant
1968). The five entities could not be classified into the same category because
there is no one property on the basis of which we could combine them by defining
a class that contains elements each of which has the same property or set of proper-
ties in common. In fact, one of those five individuals just might be a parent to the
others, or there might be three generations of individuals from the same family
included in that set. The commonality is that of origin, but it may manifest itself in
very different ways.
Following such a line of reasoning, social psychologists including L. S. Vygot-
sky and A. N. Leontev suggested, taking up in this the dialectical logic articulated
by K. Marx, that humans are fundamentally societal creatures. What is different in
humans from other animals is the predominance of society and culture over nature.
The minimum unit for understanding psychology has to retain the societal charac-
ter of being human. This has been done, for example, by using societal relations as
the minimal unit, the origin and life of specifically human characteristics, including
higher psychological functions and personality (Vygotskij 2005); it also has been
done by using productive activityfarming, manufacturing, food-producingas
the unit independent of which the individual subject cannot be understood (Le-
ontev 1983).3 The individual is but one of the different manifestations of that unit,
others including the means of production, the activity-focusing object/motives, and
the reigning divisions of labor. Pragmatists make a similar move by understanding
language together with the activity in which it is an integral part: I shall also call
the whole a language-game: language and the activities with which it is interwo-
3
Following Marx, Leontev also understands consumption to be an activity, for all produc-
tion ultimately is oriented towards consumption, itself oriented towards production (of life).
We see it increasingly around us: leisure has become itself a productive activity. Thus, how-
ever useless some sport is in terms of helping others, it is part of a fabric that provides op-
portunities to productively participate and make a living.
 RIGOR IN QUALITATIVE DATA ANALYSIS 23

ven (Wittgenstein 1997, p. 5). The author uses an inflection of the verb weaving,
thereby evoking the analogy of something woven. The language-game is some-
thing woven from language and material activity. Like any woven object, if we
pull on a string, the whole thing will come undone and we no longer have the
thing. There is no fabric. At a minimum, to understand something woven, I need to
have a piece of fabric as a fundamental category. For Leontev, this piece of fabric
that still retains all the characteristics of human society is productive activity.
Thus, to understand a teacher or a student, we must not try finding out by looking
into their minds, not even by investigating their interactions or the events in the
classroom. Instead, we have to look at schooling, which is a societal activity, the
purpose of which is to serve a fundamental collective, generalized need: reproduc-
tion of society.

Radical Doubt and Critique of Ideology

A scientific practice that fails to question itself does not, properly speaking,
know what it does. Embedded in, or taken by, the object that it takes as its
object, it reveals something of the object, but something which is not really
objectivized since it consists of the very principles of apprehension of the ob-
ject. (Bourdieu 1992, p. 236)
Bourdieu notes that the preconstructed can be found everywhere and constitutes
and is constituted by the very ways in which we experience and make sense of our
social and material lives. This system of ideas not only is pervasive but also invisi-
ble, as is the proverbial water to the fish that swims therein. The phenomenon is so
insidious that even critical (feminist) sociologists may find themselves trapped in it
without knowing or being aware of it. Thus, in an example evoked above, D. E.
Smith describes how her own thinking and research design were structured by a
mothering discourse that (initially) prevented them to collect the kind of that that
with hindsight they realized should have been collected. This mothering dis-
coursewhich historically had evolved from psychological studies and was dis-
seminated by womens magazineis part of, and ordered by, a discourse of and
about the standard North American family. The sociologist writes that when she
and her graduate student embarked on our study, we failed to register the extent to
which our thinking and research design were organized by the mothering discourse
and by conceptions of the Standard North American Family (Smith 1999, p. 162).
Literally then, Smithand presumably at some stage all social researchersdid
not know what she was doing until the point where she began to discover (register)
how the existing discourse had framed what she attempted to do without her being
aware thereof. It is not that everyone decides to actively disregard critical engage-
ment with his/her own methods, concepts, problems, and instruments, as this may
appear in the following critique:
It would be easy to show that this half-scholarly science borrows its prob-
lems, its concepts, and its instruments of knowledge from the social world,
and that it often records as a datum, as an empirical given independent of the
24 CHAPTER 1

act of knowledge and of science which performs it, facts, presentations or in-
stitutions which are the product of a prior stage of science. In short, it rec-
ords itself without recognizing itself. (Bourdieu 1992, p. 236)
Even critical, Marxist sociologists such as Smith are subject to the pervasive
effect of discourse, so that it sometimes takes considerable time and experience
before a researcher recognizes having been subject to ideology. It was through her
own participation in the mothering discourse, and through her experience of being
subject and subjected to it when her son was in school trouble, that the structuring
of the discourse became visible to Smith rather late in her research on womens
work as mothers and in an accidental manner. The functioning of the mothering
discourse was particularly covered up by the fact that the women interviewed and
the interviewers (Smith, her graduate student) were subject to the same structuring
by the mothering discourse. The interviewees oriented to the very interpretive
schemain trying to figure out what the researchers are after, this may have well
involved the documentary method of interpretationthat also had structured the
questionnaires and interview questions.
The purpose of rigorous data analysis is to read what people actually do in the
more or less contentious concerted social practices in which they engage, which
they not only agentially produce but also patiently undergo. Rather than importing
concepts such as power to explain social relations, rigorous data analysis focuses
on analyzing relations as people participate in making and undergoing them. Any
difference in knowledge, power, or institutional positionshown particularly in
chapter 8is the result of societal relations rather than the determinant causes
thereof. Rigorous data analysis focuses on the ordered and orderly making and
expositing of the (material and social) world as ordered and orderly phenomenon.
Because it goes together with the bracketing of received and often accepted con-
cepts, it may therefore serve as a tool for emancipation and critique of ideology
(see chapter 12). Rigorous data analysis is thereby aligned with the possibilities
that critical psychologists (Holzkamp 1983) and critical sociologists (Smith 1990a,
1990b) alike have outlined in the past when subjectivity as directly immersed in
contentious social practices was the focus and the starting point for a process that
would seek to reconstitute subjects as they reconstructed and transformed the cul-
tural categories and conditions that shaped their lives (Langemeyer and Nissen
2011, p. 190).
In this book, I focus on how people make the social world, exhibit to each other
the social facts as facts that matter in the situation, and account for the order that
they make and that surrounds them as an integral part of this making. There is gen-
eral agreement, however, that this work can only be part of a more general, rigor-
ous because (self-) critical approach to data analysis. When people use and exhibit
to each other specific things and categories, then they tend to submit to these and
accept the normative powers of the factual without investigating the very produc-
tion of the categories and (social) facts. Instead, rigorous data analysis brackets the
categories used by insiders and witnesses alike, takes them out of their circulation
as explanatory resources, and investigates the history of the categories usede.g.
by tracing the origin of the psyche to the single-cellular origin of life (e.g. Le-
ontjew 1964)and not unlike genealogy and archeology (e.g. Foucault 1969,
1975). It is therefore never sufficient for social analysts to conduct themselves like
 RIGOR IN QUALITATIVE DATA ANALYSIS 25

Sherlock Holmes or Miss Marple in the attempt to find who done it and to bring
him/her/them to court. Instead, we also need to understand the very historical con-
ditions and categories that allow us to understand the very production of the social
facts and situations that have led to the crime in the first place. It is not sufficient to
simply nominate the individual subject as the author of the crime but, to be rigor-
ous, we have to analyze the very conditions that constitute not only the contexts
within which the event has happened but also the very subject and subjectivities in
play. This is so because accountability, the fact that we can provide reasons to oth-
ers and to ourselves, can be found as well in those cases where the legally en-
shrined bourgeois order of society comes to be questioned the most: the legal and
criminal codes that not only provide orientations to societal subjects what to do
and how to do it, but also as guides for deciding retroactively what past action lies
within the law and what lies outside.
 PART B

Five Data Sessions

 Non-experts often know more than experts and should therefore be consulted
and . . . prophets of truth . . . more often than not are carried along by a vision
that clashes with the very events the vision is supposed to be exploring.
(Feyerabend 1993, p. xiii)
In this part B of the book I present transcriptions of five data analysis sessions col-
lected according to a think aloud protocol in an introductory graduate course on
qualitative research method. Over the course of several semesters, the instructor-
analyst introduced his students to thinking about data analysis as a rigorous activi-
ty. To highlight the point of rigor, he had invited the students to prepare transcrip-
tions of videotapes that somehow related to their research interests but to leave out
as much of the specifying detail as possible. The text was projected onto a screen
visible to all participants in the course and then analyzed, line-by-line, while the
instructor-analyst worked his way through it (Fig. B.1). He stated as the purpose of
the self-posed task the production of an analysis that would reconstruct the original
situation to the extent possible with any given transcription. Throughout each of
the courses from which the following sessions issue, he pointed out that rigor was
required. To underscore this point, he suggested that fictional or actual detectives
such as Sherlock Holmes, Ms. Marple, or Arsne Dupin get their stories wrong, an
innocent person could end up on the gallows or on the electric chair. Thus, the
events, as reconstructed from the transcriptions, must not be the results of just
some constructions, individual or shared. The person who had prepared the tran-
scription would judge the accuracy of the event description and would also be in-
vited to share the original video clip. That is, the instructor worked against the fre-
quently perpetuated idea that the reading of data is a subjective, and perhaps even
solipsistic exercise, where the only underlying criterion is the within-person fit
with the reality (here the data) s/he perceives. (This would be the solipsistic posi-
tion that I often hear and see being taking by colleagues who also denote them-
selves as constructivists.) Instead, because the analysis is publicly available and its
results would be checked by the audience, it is inherently intelligible and therefore
30 PART B

Fig. B.1 Typical arrangement for a data analysis session. The transcription is projected
onto a screen visible to all participants in the course on introductory qualitative research
methods. The instructor-analyst highlights and points to the text by means of the cursor or
gets up and directly gestures to the text talked about.

shared within the small community. As a result, it is not anything that goeshere
reversing the aphorism of the philosopher of science P. Feyerabend (1993)but
instead, the inherently social and shared viability of a situation description derived
through a close reading of a transcription was at stake in each and every session.
The purpose of these presentations is to work towards a response to the question
raised by Garfinkel and Sacks in the opening quotation of chapter 1: What is it
about natural language that permits speakers and auditors to hear, and in other
ways to witness, the objective production and objective display of commonsense
knowledge, and of practical circumstances, practical actions, and practical socio-
logical reasoning as well? The analyst displays, in his (discursive) actions, ways
in which researchers can recover precisely that which is displayed so that we can,
in accountably rational ways, produce and perceive commonsense knowledge,
practical actions, and sociological reasons.
As readers can observe in all five of the brief chapters of this part, the procedure
of the analyst relies on consulting the non-experts that appear in the transcrip-
tions and the non-expert graduate students who have produced them from vide-
otapes found on the Internet. Although they are non-experts, they do in fact make
available to others the accountably rational structures of practical actions and soci-
etal relations. The non-experts, therefore, provide all the clues that the analyst
works withas if he were overhearing a conversation and figuring out the where,
what, when, how, and who of the societal situation that has given rise to, and was
produced by, the conversation. That is, in a very literal way, the analyst acts ac-
cording to the rule presented in the open quotation namely that non-experts should
 INTRODUCTION 31

be consulted . . . precisely because from the situation that unfolds before our eyes
while reading the transcription has arisen from their visions and di-visions.
The problem of the instructor-analyst was posed in a particular way and created
particular conditions that led to the analyses as observed. The task was that of find-
ing out the type of situation thatparticularizing itself in the case recorded
produced the transcription. It is therefore not surprising to see the analyst drawing
on ethnomethodology and conversation analysis, suited for the present purposes
because specifically developed to uncover the ways in which everyday folk pro-
duce the social world in accountably rational ways and make activities the struc-
tured and more-or-less predictable lifeworld that it is. These examples are not
meant to suggest that ethnomethodology and conversation analysis are the only
research methods valued. But, as shown in part C, these methods constitute an im-
portant aspect of the researchers toolbox because they allow approaching certain
research questions much better than others.

On Reading the Chapters in Part B

In the following chapters, all of which report actual think-aloud data analysis ses-
sions, there are three levels of text. At the lowest level are the transcriptions that
the graduate students had prepared and made available to the instructor-analyst
who saw them at that point for the first time. These transcriptions are made availa-
ble, in their entirety, in Appendix A. As most of these originally did not contain
numbered turns, I added these for the convenience of reading and referring to
them. At a second level, readers find the transcribed think-aloud session. This is
following some cleaning up, removal of the interjections (e.g., uh, uh, and
um) and half-finished words, and grammatical corrections for the purpose of
representation as written textwhat the instructor-analyst actually has said. This
text therefore constitutes a first go at an analysis without the analyst having had
any background information or pointers other than what was available in the tran-
scription itself. The text at this level therefore shows what was salient to the ana-
lyst as he moved through the text, line by line, turn by turn, to produce a situation
description that possibly generated the concrete talk in the transcription at hand.
This is where readers actually can develop a feel for what experienced analysts do,
what is becoming salient to them, what they focus on, and how they understand
their task. At a third level, I provide further pointers and explications to the work
of the analyst. This third level is of the type that readers may normally find in a
book on qualitative data analysis. It is at this level that I also refer to and cite rele-
vant research.
A considerable time after I had written a first draft of this book generally and
about the three levels specifically, including the three-level presentation that marks
this part of the book, I found a conceptualization of what I had done in the early
work of D. E. Smith (1981). She describes level-one talk to be the talk of the peo-
ple at work, doing whatever it might be to get the days work done. This talk is part
of the language-game that constitutes the ethnographers interest. Level-two talk is
talk about the original situation, which may be used by the informants themselves
32 PART B

when they talk about what they are doing in a particular situation, such as when
scientists speak about what they have done as part of their workday. What may
confuse is that this language-game overlaps with the original language-game by
making use of the same words. What is often not visible is the different use to
which words are subjected on that second level. Essentially, however, level-two
talk, talk about level one, presupposes familiarity with the ways in which level-one
talk works. Level-three talk is part of the social scientific discourse, for example,
the sociological, psychological, or social-psychological discourse. It serves to cat-
egorize and subsume the various ways of talking at levels one and two. Level-two
talk mediates between the talk that participants employ as part of doing their eve-
ryday workdoing shopping, doing caring for children, or doing teachingand
the theoretical discourse. Importantly, there are sets of social relations put into re-
lation by level-two description: that relevant to the social actors in their situation
and that relevant to theorist theorizing society (Smith 1981). Using the sociological
term assignment for a study of newsrooms as example, the author suggests that
the solution to how it is possible to speak of assigning and assignments
and to point to pieces of paper to describe courses of action, and perform as-
signing, is not to be found in the context of which these usages make their
current sense. If we see that the terms of the setting must be taken up as ex-
pressing its social relations rather than as categorization procedures, and
hence that a work of inquiry must investigate the social relations in which
such usages as these and others are both possible and sensible in that context,
we will be able to move to a non-ideological method of sociological descrip-
tion. (p. 334)
To easily distinguish the three levels and also to distinguish the different con-
cerns that they exhibit, I chose the following convention. The first level text, that
is, parts of the transcription that the analyst currently is focusing on, is printed in
block quote indented twice and printed in a smaller font (9 points). Level-two text
appears in block quote indented once, normal font (10 points). I use double paren-
theses and text in italicse.g., ((turn 06))to provide ethnographic descriptions,
for example, where the analyst is pointing to on the projected transcription. I use
square bracketse.g., [the analysis]to add words to make the sentences com-
plete. Quotation marks are used to enclose quotations that the analyst indicated by
means of voice inflections or air quotes. Finally, I use underline when the analyst
prosodically emphasizes textual particulars. The third-level text appears below the
footnote line in italics. Each part of this text is directly keyed to the phrase/s or
paragraph that it is about. The intent of this presentation and structure is to provide
readers with ways of making connections between actual data of the type they
might be collecting and the descriptions that might be found in a traditional meth-
ods text. These two levels are held together in and through the practices of the in-
structor-analyst, who, with over 25 years of experience at the time, had gathered
considerable experience in and familiarity with data analysis.
In each of the following sessions, what is happening is to be understood as a
beginning rather than as a full analysis. Readers need to keep in mind that the in-
structor-analyst was taking only about 30 minutes with each transcription. He did
so for a first time, and upon extended analysis, he might have picked out further
 INTRODUCTION 33

Fig. B.2 The structure of the work exhibited by the instructor-analyst.

features, deepened the analysis, found more documentary evidence for the hypoth-
eses that already captured the essence of the actual situations that have given rise
to the recorded situation, a version of which was subsequently transcribed for the
purpose of the data sessions. (These are described at the end of each analysis.)
We may think of the sessions as exhibiting a particular structure (Fig. B.2).
Confronted with the transcription (Fig. B.2, right), he works out the (kind of) rela-
tional work that manifests itself in and through the give-and-take on the part of the
participants in the verbal exchange. Once articulated, an inference is generated
about possible societal situations (Fig. B.2, left) that could have, in the concrete
realization of a specific instance, has produced the talk that is rendered again in the
transcription. In other words, if we take a type of recognizable situation, TV talk
shows, then it might be realized in a variety of ways, for example, in a particular
episode of the Oprah Winfrey show, Late Night with David Letterman, or Conan
with Conan OBrien (Fig. B.2, center). The transcription represents the talk by
means of which the participants in the specific societal situation make apparent to
each other, in ordered, ordering, and orderly ways, the structured nature of the so-
cial world. This happens independent of the specific individuals, who staff a socie-
tal phenomenon that they not only produce but also to which they are also subject
and subjected.
2

Data Session 1 (Heidi)

(Erin, David Suzuki)

And then you are usually tempted to import concepts. For example, when peo-
ple say, Oh there is a power relation, I say, Put your finger on it. Show it to
me. So I dont like people explaining things with power. What I want to see is:
I want to actually see it, how people both produce and reproduce institutional
relations.1 For example here we know David Suzuki. So, we probably can ex-
pect, and this is the danger, that there are maybe certain things going on. Work-
ing with a transcript where you dont know who the people are, then you dont
have this kind of resources. The danger might be that because we know David
Suzuki, we might think, Oh, people will show respect. Even if we dont sort
of explicitly think it. It could certainly tweak our analysis. And then we see
people giving respect to him when in fact we dont have any evidence for it. So,
analyzing a transcript where we dont know much about the people forces us to
look at whats going on. Now if I dont have a good starting point, what I usual-
ly do is this: I describe what happens.2 I begin with a description, in my words,
what is happening here. And in my words it could be in varied ways. If I look at

1
The instructor-analyst begins with a warning concerning the dangers that an
analyst faces when drawing on common concepts. Power, or knowledge, is not
something of the same nature as a desk or chair. It is a form of relation that partic-
ipants produce and reproduce (e.g., Foucault 1975). The analyst therefore recom-
mends focusing on the relation, as available from the back-and-forth of the talk,
rather than using the position David Suzuki takes in societyhe is a well-known
Canadian scientist, broadcaster, and environmentalistto explain what is happen-
ing in the conversation specifically and in the society-specific relation more gener-
ally.
2
The analyst describes how he begins an analysis, especially in this case where he
does not know anything about the background or about the actual videotape that
has yielded the transcribed talk under consideration.
36 CHAPTER 2

this, if I look just at this first page of the transcription globally, then I see that
Heidi talks a lot. The other person that talks on this page is David Suzuki; and
Amanda, Michael, and Ashley hardly talk at all.3 And so, initially, the first thing
we see Heidi . . . and if I dont have a starting point for my analysis, if I dont
see patterns right the way, then I might do what I am doing just now.

01 Heidi: Its amazing how the landforms here really look like other things other
than rocks. Do you notice that sometimes?
02 Amanda: Yeah
03 Ashley: Yeah
04 Michael: Yeah
05 Heidi: Its funny
06 David: As long as youve got a good imagination

Heidi begins to speak in this episode. She describes the landforms as amaz-
ing; and then a the three individuals Amanda, Michael, Ashley, they all say yes,
yeah. And then, are these funny? Then David Suzuki comes in. So I would
begin with describing this, because I dont have a handle yet on the situation.
And as I start describing that I hope stuff sort of comes up. And those who have
observed me write in real time, analyzing in real time, they know that in trying
to articulate what is going on, stuff comes up.4
So we see Heidi, Its amazing how the landforms here really look like other
things other than rocks. Do you notice that sometimes. Yeah. Yeah. Its funny.
And David, As long as youve got a good imagination. Now we can understand
what is being said here in response to something that happened or was said be-

3
Gazing over the page of transcription provides an overall impression of the dis-
tribution of talk. This is equivalent to other ways of making the data look strange,
for example, when we run the video at a much slower speed, in slow motion, or at
a much faster speed than normal. I do almost all my video analytic work with
QuickTime Pro because of its ease of operation, features to work with videoe.g.,
overlaying transcriptions, producing picture-in-picture or side-by-side video, and
information panelsand general simplicity.
4
Over the course of the past 30 years doing analysis, I have encountered many
colleagues, especially beginning scholars but also more seasoned ones, who find it
hard to get started. In part, the difficulties arise because the individuals intend to
write an article or a paper for a conference. In view of intending to produce a fin-
ished product, they experience themselves stifled. I therefore tend to recommend
opening a text file with the name Notes, to which I often append an identifier
number and a tracking number (e.g., Notes_1_100.doc). The writing in these
notes is just for myself. This allows me to write anything that comes to mind with-
out worrying about the political correctness of the content. Prior to including it in
an article or paper I write, I will edit it or make it consistent with whatever ideas I
might espouse in the text to be published. The version number is updated from time
to time (Notes_100, Notes_101, . . .), which not only gives me a historical record of
the notes produced but also prevents me from losing materials in case the word
processor crashes.
 DATA SESSION 1 (HEIDI) 37

fore.5 So here, what has been said beforewell, it was very little by Amanda,
Michael and . . . Heidiwhat we actually have is some statement: It looks like
other things other than rocks. The next one isnt what I read as an unfinished
question and is possibly not what David reacts to ((turn 06)). So they look like
other things, is the likely statement that has this one ((turns 0204)) as the re-
sponse or the next turn. So, the next thing then is, Heidi comes in, Well weve
noticed that maybe yeah, maybe lets go down this way.6

07 Heidi: Well, weve noticed that maybe, yeah, maybe lets go down this way.
We noticed that the longer youre out in the badlands, and the hotter it
is, the more things look like things
08 David: Yeah-ha-ha-ha. Your imagination gets looser, huh?

We notice that she may have actually articulated something salient in the sit-
uation: if it is on videotape, you may actually see that they walk somewhere and
she says, We noticed that the longer you are out in the bad lands and the hotter
it is the more things look like. The more things look like things.7 And David,
Yeah, yeah, youre imagination gets loser? If I now look at Davids lines ((turn
06, 08)), I already see the appearance of this idea of imagination.8 He appeared

5
The analyst will be seen to look for relations and turns, each line or turn in re-
sponse to something else. Thus, even the first turn (turn 01) is in response to some-
thing else, which might have been something said before or the situation itself. Not
a single analysis we may want to conduct is intelligible outside of its historical
contextthat of the event to be analyzed as much as that of the person analyzing it.
6
If taken as responses, each of the turns on the part of Heidis interlocutors has a
counterpart. But which aspect of Heidis talk in turn 01 do the other turns respond
to? The identification of pairs is important for reconstructing the inner dynamic of
the event of which this transcription is a concrete document: the talk is an integral
part of the situation, making and being made by it. The talk, here transcribed, also
constitutes a trace that the living conversation will have left behind.
7
Here, the analysis points to the statement as a possible response to something to
be seen. We observe a hypothesis about the context: A nature walk involving the
five individuals, who are talking about what is visibly available to them. That is,
the analyst is taking the text read so far as a concrete document of a situation as a
whole. It is a documentary method of interpretation, where some whole phenome-
non is reconstructed on the basis of quite varied and even contradictory manifesta-
tions it leaves behind (Mannheim 2004). Although the analyst does not articulate it
here, he takes the transcription as a protocol of a situation in which the partici-
pants make available everything they need to produce the situation as an account-
ably rational one. They talk about and in ways required by the situation and do not
talk about anything that goes without saying where and in the context they are,
and when they are in that place. Yet what goes without saying, the literal, the ordi-
nary, or the obvious can be uncovered in close readings of a written text or tran-
scribed talk (Garfinkel 2007).
8
There is a pattern, something invariant across speaking turns. The topic of imag-
ination appears repeatedly in the talk of David. Invariants are precisely what the
analyst is searching because these provide clues to the type of situation that may
38 CHAPTER 2

to have picked up on something articulated by Heidi, like things that look like
other things. And now, repeatedly, he at first says If you have a good imagina-
tion and then, and then, in response to what Heidi says, which maybe We no-
ticed that the longer you are in the badlands and the hotter it is. Then we can
hear David Your imagination gets looser. So one may hear what David says
here as almost ironical. And in the end, its not in the conversation.9 It doesnt
contribute to reproduce irony because Heidi doesnt play with it.10 Now here,
for Heidi it actually does. So some people claimed that it might be kind of hallu-
cination. And what we are seeing here is a build up of denying.

07 Heidi: Well, weve noticed that maybe, yeah, maybe lets go down this way.
We noticed that the longer youre out in the badlands, and the hotter it
is, the more things look like things
08 David: Yeah-ha-ha-ha. Your imagination gets looser, huh?
09 Heidi: Well, it does. Some people claim it might be a kind of hallucination
now watch out for the cactus. When we get around here, you want to
take a look around and see if theres any landforms that look like some-
thing that would be familiar to you, not just like a rock. So what do you
think that landform over there is? Does this one look, look like anything
to you?

Heidi, she has said the longer you are out there, the more you see different
things. David is saying Your imagination gets looser; and it may be a probable
implication that its almost like as if you were stoned. I mean, you can think of,
my mind goes to The Doors going out into the desert and then hallucinate.11

have produced what can be seen in and as irrefutable, concrete evidence: the video
and the literal transcription of what participants can be heard to have said.
9
The analyst is warning us against attributing intentions to a speaker when there
is no evidence of knowing what is in his or her head. Moreover, it is a warning
against making attributions when these in fact do not matter to the conversation,
where participants only have available whatever others in the situation provide
them with. Hidden intentions, contents of the mind, feelings, or beliefs are not gen-
erally available to others. However, if others were to hear or see something that
matters, this would be made explicit as part of the work of maintaining the relation
between the speakers.
10
It is only if Heidi plays with [the irony], that is, only if she takes up and puts
irony into the play again that it will have mattered to the situation. As a result, a
statement is to be analyzed as a joke only when participants treat the statement as
such (e.g., Roth et al. 2011).
11
This is a reference to the rock group The Doors, the members of which had gone
into the desert tripping on peyote. We can understand what is happening here as
the elaboration of a particular (documentary) sense that is evoked. Analysts there-
by explicitly take a line on, and indeed start with, a certain politically or socially
charged description of, the speakers or the subject of the talk being analyzed (An-
taki et al. 2008). However, it is in and through the rigorous analysis of the con-
crete documents that the relevance of this general sense or take on the situation
has to be borne out.
 DATA SESSION 1 (HEIDI) 39

Heidi was talking about hallucination, When we get around here you want to
take. . . .
See what I am trying is to? I dont have a starting point. I mean: I am cold. I
am starting cold, knowing nothing about the situation. In order to get into [the
analysis], I try to elaborate and describe what I see going on, but also explain
and link it to other things.12 So, When we get around here you want to take a
look around and see if there is any landform that looks like something that
would be familiar to you, we already hear that as picking up a theme that al-
ready appeared in the statement and later on about looking differently. Its all
about perception and how things look and how they might look like other things
that they were familiar with landforms that look like something that would be
familiar to you. Not just like rock, but like something else.

09 Heidi: Well, it does. Some people claim it might be a kind of hallucination

now watch out for the cactus. When we get around here, you want to
take a look around and see if theres any landforms that look like some-
thing that would be familiar to you, not just like a rock. So what do you
think that landform over there is? Does this one look, look like anything
to you?

So what do you think that landform over there is like? You probably all have
played this: kids looking at the sky going, Oh, a sheep, oh something else.
Thats what this generates for me, the idea.13 And if I am blank with the analy-
sis, these are the kind of things that I build from. I describe to get myself started.
So, What do you think that the landform over there is? Does this one look like
anything to you? I stop reading now, again, and I think about what kind of rela-
tions there are. Well the three ((points to Amanda, Ashley, and Michael)), they
havent talked yet at all. I havent seen the video. But the image I have is this:
maybe they are people unfamiliar with the wilderness and maybe they are
younger people.14 There is David; and there is Heidi, someone who functions in
a situation where the kinds of questions seem to presuppose that the person al-

12
In these first few statements of the paragraph, the analyst is saying what he is
doing with his other talk, that is, he is formulating. In the demonstration, talk
about what he is doing, his method of praxis, has to follow or precede the talk
where he is doing what he is doing, that is, the praxis of analysis.
13
The analyst in fact describes a kind of situation that could have led to the state-
ment, So what do you think that landform over there is? We may hear this as a
hypothesis as to a social situation that could have led to the statement, which can
in fact be heard as a question. In this case, the statement would be a concrete doc-
ument of one kind of social situation. That is, we see an aspect of the documentary
method of interpretation at work.
14
Here again, the analyst articulates the image he has, of a situation type; the
analysis now has to work out the structure and content of the concretely available
documentary evidence to show the invariants that are typical for the situation hy-
pothesized and those aspects of the transcription that are incidental and particular
to the case. Heidis proper name is incidental, her occupation is not, as it belongs
to that kind of social situation, which could have been staffed by someone else.
40 CHAPTER 2

ready knows the answer. So what do you think? What do you think that that
landform over there is? Does this one look like anything to you? There is some-
thing in this question that makes me think that the person already knows the an-
swer, or knows an answer.15 But it is not asked like in a situation where, where
a person says, Oh that looks like a sheep to me and to you? or like the ques-
tion What time is it? And well, you respond.16 Whereas in teacherly discourse
you will have: What time is it? And it is asked in a way where the asking per-
son already has the right answer. And this question seems to be of that kind of
questions:17 the person, the relation of the person ((Heidi)) to the others. So you
see how even without having seen the video how Im attempting to provide a
description of the situation, of whats happening here.

09 Heidi: Well, it does. Some people claim it might be a kind of hallucination

now watch out for the cactus. When we get around here, you want to
take a look around and see if theres any landforms that look like some-
thing that would be familiar to you, not just like a rock. So what do you
think that landform over there is? Does this one look, look like anything
to you?
10 Amanda: Hmm. Oh! That rock right there looks like a camel
11 Heidi: Oh . . . no . . . thats it! We actually have a name for this guy. We call
him Fred the camel . . . see the hump . . . see the big droopy lips point-
ing to the left. And if you look off in the back can you see anything
else?

15
This analysis focuses on the structure of the statement, here heard as a question,
which indicates that there already is a prefigured reply against which whatever the
participantsAmanda, Ashley, or Michaelsay will be judged. Here it is the in-
terrogative What do you think . . . and the definitive article the landform that
together lead to the clue. The three are asked about their thoughts concerning the
specific landform that exhibits some characteristic that is to be disclosed. We can
hear the opening as implying that the questioner already has something in mind
and now asks the three individuals what they think the form looks like. The analyst
notes: There is something. It is now up to the analysis to give a concrete form of
this something, the objective sense associated with it, so that we can understand it
as documentary evidence for the kind of situation generally and this society-
specific relation particularly.
16
This analytic move is to be understood as a variation. That is, the analyst just
has provided a description of a hearing. Now he provides alternative descriptions
of what the heard statement does not sound like. Variation here provides alterna-
tive descriptions and, in so doing, alternative hypotheses about how to hear a par-
ticular statement.
17
This would be the statement of the hypothesis about the particular statement
heard as a question on the part of the analyst. He says, It seems to be of that kind
of question, which makes the tentative nature of the analytic statement apparent.
This production and maintenance of multiple hypotheses, each of which is to be
tested in the transcribed turns to come or, backwards, to what has already been
read.
 DATA SESSION 1 (HEIDI) 41

Hmm, Oh! that rock there looks like a camel. Oh no, thats it. I havent been
there, but the person saying Thats it confirms that the answer was the one thats
prefigured, preconceived in the question.18 So, What do you think that landform
over there is? There is a childrens game: Do you know what I think? Or you
look at the some cloud, I see. What do I see? And then she goes, We actually
have a name for this guy, I call him Fred the camel see the hump? So the big
droopy lips pointing to the left, and if you look off in the back, can you see
something else? So again, Can you see something else? The question is of the
the kind that signals that the person already knows the answer. And you are
tested whether you can actually produce it. This is the kind of question that in
the literature has come to be called by some as preformatted answer, where
the person who asks the question already knows the answer.19 The question is
simply out there to test the person. Whereas genuine questions are the kind of
questions people genuinely ask each other: What time is it? You know, when
people dont have their watches and you know as a competent member of the
culture that the question is genuine. People dont . . . Imagine yourself in the
street and a person asked you what time is it, and you go, A quarter after sev-
en. And the person says, No, it is actually sixteen minutes after seven. You
go like, What a weird-do.20 So we are accountable for the kind of questions
that you we are asking; and in this particular situation, the questions seem to be
already prefigured, there is a particular kind of answer. And those who answer
do so knowing that there is some such preformatted answer. So and the same, If
you look off in the back can you see anything else?

18
Now there is a confirmation of the hypothesis and, simultaneously, a disconfir-
mation of the alternative hypothesis or hypotheses. It may not be confirmation and
disconfirmation to certainty (i.e., probability p = 1) but an updating of probability
of the hypothesis under consideration based on the data. This is expressed, in a
Bayesian approach, as the posterior probability of hypothesis Hx given the data,
or, in formal terms, p(Hx|data).
19
The phenomenon is referred to differently in different communities of practice.
In education, we frequently find reference to a turn-taking sequence denoted as
IRE, which is short for initiationreply-evaluation (e.g., MacBeth 2003; Roth
2009). In schools, teachers tend to take the first and third position in this sequence,
students the second position. It is not only a way of constituting whether a student
does or does not know, but also a way of producing science as a heroic effort (van
Eijck and Roth 2011) or of providing feedback to students who, in this way, can
find in their own actions and talk those aspects that are relevant to the production
and reproduction of science or mathematics or any other subject (Roth 2013).
20
Here we find a variation produced. It stands both as alternative, and therefore
as a second hypothesis, and also as a ground against which the actual situation
becomes figure. It is highlighted and articulated specifically for the benefit of the
graduate student audience, but may operate in a tacit manner if the analyst was
working on his own. Variation assist us in identifying other possibilities and, in
this way, get closer to why this statement has been made in this situation rather
than some other statement.
42 CHAPTER 2

Now, to me, what Ive said just now, what is it? Why would I, who has not
read the transcript before, who has not seen the video, who has not heard the in-
tonation, why might I have detected that here this is a preformatted question?21
What appears later on here shows that she indeed already has had the answer,
this answer in mind. Why would Ive picked that out? What is my competence
that allows me to detect this just in the question? That could be a research ques-
tion in its own right. What is it that allows us to look at or hear a question for a
first timemaybe its a question in contentand make a suggestion that this
person already knows the answer to the question? And the question is only a
fake one to test someone else. What is it? What is it about that question? What
is it in the question? And what is it in my cultural competence? Because once
you understand better what the kinds of question are, for example, that teachers
ask, you become attuned. One might find this kind of relation between parents
and their children or between teachers and their students, this particular kind of
question.22
Someone might think that this has to do with the idea of habitus. Habitus is a
theoretical term that the sociologist Pierre Bourdieu generated.23 Its actually

21
This is a crucial instant in this analysis, because the focus no longer is on the
construction of a social reality on the part of the participants in the situation. Ra-
ther, the analyst turns attention to his own instruments of construction. This is im-
portant because the worst (critical) analysts can do is remain blind to their own
instruments of construction all the while believing they are critical because they
take such a stance with respect to the object (e.g. Smith 1990a). This demands a
double-edged criticism, both with respect to the research object under construction
and with respect to the instruments of this construction. Why, he is asking, does he
hear the question as one that has a preformatted answer? What is it in his hearing,
standing in for hearing in general, that makes the statement a question with a pre-
formatted answer rather than a genuine question? This paragraph then continues
to raise questions about the specific hearing and where it might originate, and
what kinds of cultural competencies are expressed in such a hearing.
22
In this final statement there are two important ideas. First, the concrete ques-
tion, which is treated as a document, may be functioning as such for different kinds
of societal situations (relations). Second, the statement can be heard and read as a
formulation of two hypotheses (H1, H2) about possible phenomena that manifest
themselves in this statement-question (i.e., Can you see something else?). Without
any additional data, the two hypotheses would be equally likely, i.e., p(H1) = p(H2)
= .5. As soon as additional data pertinent to these hypotheses are available, the
probabilities will change and will be updated to the posterior hypotheses
p(H1|data) and p(H2|data).
23
Readers may be interested following up on this concept, articulated by Bourdieu
(1980) in a book subsequently translated and somewhat inaccurately entitled in
English as The Logic of Practice. (Inaccurate because Bourdieu uses the term
practical sense, which is different from the term logic of practice that evokes
the specters of the rationalism that the book critiques. However, and in this the
translators had some justification, Bourdieu also talks about the logic [logique] of
practice in the text itself.
 DATA SESSION 1 (HEIDI) 43

not a very easy concept. The translation some people use is disposition. Habi-
tus then is a set of dispositions that makes us act and see in the way we do. So
its not actually describing what we do and see, but names our predispositions to
see and act in that particular ways. So when I say its my habitus, then we go
very quickly to an explanation, to a theory. And so we basically subsume my
actions to this theory, whereas I would encourage you to work with the data and
sort of come up with an ethnographically adequate description, from the data,
and to understand what is going in there.24

11 Heidi: Oh . . . no . . . thats it! We actually have a name for this guy. We call
him Fred the camel . . . see the hump . . . see the big droopy lips point-
ing to the left. And if you look off in the back can you see anything
else?
12 Amanda: ?
13 Michael: No.
14 David: Ill give you a clue. Where are camels found?

So she asks the question, And if you look off in the back can you see some-
thing else? Theres something said, probably the question mark means that you
didnt hear Amanda.25 Then David . . . now David is in cahoots, because he says
Ill give you a clue. So he also knows the answer. So those two, Heidi and Da-
vid, they seem to be in cahoots. David also knows what is been asked. The two
are in cahoots about whats been asked. And something particular is to be

24
At the beginning of the paragraph, the analyst points out that the audience might
explain his own competencies in terms of the habitus concept. However, just as he
warns them of the dangers in applying high-level theoretical concepts in the analy-
sis of the data at hand, he warns them of the danger to think about what it makes a
researcher to read data in a specific way by drawing on high-level concepts. In-
stead, he encourages descriptive and data-driven analysis. Such analysis works
from the data upward in the attempt to make intelligible what is going onin the
data as much as in the analysis.
25
Readers may ask themselves what is it that allows the analyst, without even
stopping for an instance, to take the question mark in turn 12 as an indication that
the transcriber did not hear what Amanda is saying, whereas he takes the question
mark in other places as the transcribers hearing of a question. Although this
might be taken as a self-evident situationnot so self-evident because in this way,
the placing of the question mark does not follow standard conversation analytic
practice that makes use prosody markersit provides a nice example of some of
the impossible-to-teach aspects of data analysis. Why is something salient in a
particular way? For example, why may the analyst focus on the use of definite and
indefinite articles but not do so in another transcription? This question will be
taken up in the chapter 10 on the documentary method of interpretation. In brief,
however, my answer will focus on the relation between the documente.g., the use
of definite and indefinite articlesand the original situation to be reconstructed,
which, at the moment, is only an object*- or situation*-in-the-making (e.g. Roth
2013b). (The asterisk denotes the unfinished nature of the phenomenon it marks.)
44 CHAPTER 2

seen.26 When some people know and ask for things that are there to be seen and
others they dont see it, then you maybe in some kind of didactic situation
where people actually learn to see certain things. This is what happens when
parents read with their children. Read images. You begin and you point to
thingsbanana, the elephant, or the birdand later on you question: Do you
see the bird? Or, Do you see, whatever, Peter who is hidden? Where is Pe-
ter? So you have this whole cultural routine of learning to identify things.27
And here we have the kids, and David asks Ill give you a clue. Where are cam-
els found? Eh, Egypt. Very good. And the other person, A pyramid, a camel and
then a pyramid. Very good.

14 David: Ill give you a clue. Where are camels found?

15 Ashley: Egypt.
16 David: Very good!
17 Amanda: A pyramid. A camel and then a pyramid
18 David: Very good! Has Fred changed that much while youve been here Heidi?
19 Heidi: Not that much although he did get a bit of a facelift, hes lost his double
chin. But, uh, were really concerned that the cap rock on the hump of
Fred may fall off. That ironstone. And if that falls off the hump could
erode away very quickly.

Has Fred changed that much while youve been . . . Not that much, although
he did get a bit of a facelift, hes lost his double chin, but very really concerned
that cap rock on the hump of Fred may fall off and then come back and I have
an idea. So David asks Heidi, Has Fred changed that that much while youve
been here Heidi? David appears to know that Heidi is here or has been here
more than once.28 The question also shows that Heidi has been here more than

26
We clearly see the analysts orientation to the actors point of view. These artic-
ulate something for each other; and this something is salient and important to the
event in which they are both subjects and patients. What they say is a verbal proto-
col, a way of exposing and accounting for the rationality of what will have been
the outcome of their irreducibly joint work. The members to the setting not only
talk to have some conversation but also make the situation be of a particular kind;
and this making of the situation requires exhibiting the situational structures and
the members work.
27
Earlier the analyst evokes school-like situations, where formatted-answer ques-
tions are asked in the IRE routine. Here another didactic situation is articulated
where the same type of questioning and turn taking might occur, but where the
purpose differs: Allowing the learner or newcomer to see what others already can
and do see. The question is part of the way in which whatever can be seen is ac-
counted for. It is grounded in the insight that language has apophantic function,
that is, allowing a phenomenon, which the learner does not know and therefore
cannot intend to see, to show itself from itself (Heidegger 1977).
28
The analyst attends to the grammatical structure of the locution. The conjunction
while together with the present perfect tense have been articulate a continu-
ing presence in that place (here) over a period of time. While functions as a
conjunction to joint Heidis being there to the change in the camel-shaped land-
 DATA SESSION 1 (HEIDI) 45

once. David knows; and David may actually be less frequently there because he
says while youve been here. Now if he says while youve been here it could be
that Heidi has been working in that area. If I bring knowledge in that David is
concerned with the environment and so on, I could, for example, hazard a guess
that Heidi is something like a ranger or a naturalist working in a particular ar-
ea.29 There are three visitors to that park, and David is there. David works with
Heidi for some time; and they are in cahoots. This ((transcription)) is my data.30
They know the questions and the answers that they want. But David also gives
away clues that Heidi has been there for a while. She actually has been there
more frequently than David; and he asks the question, has he changed. Its pos-
sibly not the same kind of question . . . well there are two possibilities. The
question is not the question that David truly knows the answer to, and then it
would be a genuine question; or it could also be a didactic one.31 He might want
Heidi to explain something that she didnt think of in the situation, but he wants
her to bring it up in the end. Rather than him telling what it is, she could ((i.e.,
following the question)).
These are possibilities. So when I analyze data, I often to test whether I un-
derstand whats going on I generate hypotheses simply for the pleasure of gen-
erating hypotheses, which I can then test afterwards.32 Im almost conducting an
experiment with my own understanding and then I can weed out. Because if the
hypothesis is confirmedit is not a very good research to merely confirm hy-

form called Fred. Moreover, the continuing presence may be taken as a document
for a work-related presence in that placewhich we encounter below as an inte-
gral feature to the analysts description of Heidi. The same locution also articu-
lates something about David, who, in stating what he does, also communicates
knowledge of Heidis presence over an extended period of time.
29
This analytic statement articulates what already is prefigured six lines earlier:
Heidis continuing rather than occasional presence, the latter eventually attributed
to David.
30
This is a direct pointer to what is hard fact, the words that participants actu-
ally have used, and which are not only the outcome of this specific situation (the
case) but also the concrete document of situations like this. The accountable ra-
tionality of the situation (event) is made available in these words, which, therefore,
constitute a protocol for this kind of situation what ever it will have been when the
analysis has ended.
31
Here again, alternative hypotheses are articulated and, thereby, prepared for
testing within the datawhich might come in subsequent parts of the transcription
or, if the analyst were to return backwards, from previous parts of it. The analyst is
quite explicit about the unfinished nature of the hypothesis, which may but also
may not be confirmed. Equivalently we may say it may or may not be supported
and become more probable given additional data that the analyst finds as he reads
along.
32
In this part of the think aloud protocol, the analyst comments reflexively about
the role and purpose of generating alternative hypotheses, which are maintained,
eliminated, or changed in their likelihood by means of evaluation given further
data.
46 CHAPTER 2

potheses, its better to reject them. But if it is confirmed it shows me that that I
have somewhat of an understanding for what is going on.33 So theres a question
and then she says, not that much, although he did get a bit of a facelift, hes lost
his double chin. There was a question. I cannot hear or read whether it is a set
up or genuine question. But it is a question that allows Heidi to articulate that
there is a change. The students may not have thought of the question that there
might be change in the hump, in the camel, in Fred.
Fred is the camel, because it has been named. This is another interesting
thing. So they are looking at nature and identifying a hump. They say, Ok its a
camel, and then Heidi introduces it: Oh, we call him Fred. This is another indi-
cation that shes been there more often. She also says we. And it may not imply,
entail, or include David. The We may not include David at all. It certainly does
not include the three ((Amanda, Ashley, Michael)), because they seem to be the
target of this didactical situation. But David, although he has been familiar, he
may actually not be part of the club that is included in the We. So that could
point me towards another hypothesis: Heidi is familiar with the area, she works
in the area, whereas David, he comes regularly, hes familiar with them, hes
familiar with the area. But he is not there as often as she is.34
So then theyre really concerned that the cap rock on the hump of Fred may
fall off. That ironstone. And if that falls off the hump, it erodes away very quick-
ly.

19 Heidi: Not that much although he did get a bit of a facelift, hes lost his double-
chin. But, uh, were really concerned that the cap rock on the hump of
Fred may fall off. That ironstone. And if that falls off the hump could
erode away very quickly.
20 David: But erosion is natural, its going to fade away over time.
21 Heidi: Fred is naturally going to erode away. But if he ever lost his hump, all
wed do is change his name to Humphrey the camel.
22 David: Awwwwwwooooooo
23 Heidi: Ha ha Ha Ha HA Ha
24 Amanda: [Ha ha
25 Michael: [Ha ha

33
The logic of the investigation is not one of refutation, where (null) hypotheses
are stated and rejected, and where confirmation is viewed as a form of bias (Pop-
per 1962). Instead, the logic is better understood in terms of abductive reasoning
and the Bayesian approach. New data are used to update the likelihood of all al-
ternative hypotheses, which, for some data, may not lead to any change.
34
The analyst makes reference to one of the hypotheses articulated earlier, which,
in and through the use of the we, a concrete piece of evidence and therefore
document of, is updated. Thus, the data is constituted by the use of we, which is
heard in the context of the differential knowledge about the situation and what
participants (David) state about knowing it. This data is used to update the proba-
bility of the hypothesis H1 = {Heidi is a naturalist working in that place from
which the camel-shaped mountain called Fred can be seen}: p(H1|data); it is also
used to update the equivalent probability p(H2|data) concerning hypothesis H2
about David as being familiar with the situation but as an occasional visitor.
 DATA SESSION 1 (HEIDI) 47

26 Ashley: [Huh ha
27 Heidi: Sometimes it takes a while. Hmmm Hmmm Hmmm

Heidi articulates a lot of the kind of things that may go on in this environ-
ment. This shows that she has a concern that it may fall off, that she is probably
more familiar with that particular situation and that what kind, whatever is go-
ing on, possibly erodes that sufficiently quickly, that whatever she says is going
happen.35 And David then comes in, but erosion is natural, its going to fade
away over time. He now makes reference to it, but erosion is natural. So what
he does is . . . what does he pick up on?36 This comment ((i.e., but erosion is
natural)) is made to something that has been said. And here is somehow its
modified. But erosion is natural. When we look at the interaction, what this
seems to be the next, the follow up of seems to be the concern articulated previ-
ously. A concern that the cap rock may now fall off, but its natural. What are
the other things David might react to?37 He doesnt react to all the statements.
So what we get here is a statement of concern and the next, the follow up is, but
its a natural process, if these things erode. Dont forget that David says, very
good ((turn 16)). Very good! You have an exclamation mark, which points me

35
Here the concerns are used as data to further update the hypothesis about Heidi
and the role she plays in the concrete situation the type of which the analyst at-
tempts to reconstruct based on the documentary evidence left behind.
36
In the analysts question about what David picks up on, we have (documentary)
evidence for a transactional orientation. What David says is not taken as an indi-
vidual piece of talk, an idea somehow externalized by means of talk. Instead, David
picks up and acts on something said before, and, in so doing in the form of talk,
speaks for others, to whom the talk thereby returns. That is, any locution is under-
stood in terms of language that comes from the other, is produced for the other,
and returns to the other (e.g. Derrida 1996). Every locution, therefore, is an aspect
of the societal-cultural situation rather than an emanation and product of the
speakers subjectivity. For the conversation to be an intelligible social event, the
intelligibility of anything speakers can say already has to be presupposed. It is not
that hearers have to interpret and construct what the speaker means to say; in-
stead, the intelligibility of the Said is presupposed in any situation where talk oc-
curs. When the intelligibility is in question, then the participants in the situation
tend to make this a topic of talk (unless that is more difficult, such as in a lecture
situation where the members of the audience may not want to interrupt even
though they do not comprehend the talk).
37
The referent of what some previous speaker has said and that is taken up in the
current speakers locution is not taken as given. The analyst exhibits a concern
with finding those parts of the transcript that together constitute a pair, where the
second member picks up on whatever is stated by the first member. Without ad-
ditional information, for example, the bodily orientation of the speakers, the text
alone may not provide sufficient resources for a more definite determination of the
first member of a pair taken up by the second member currently in the analytic
focus.
48 CHAPTER 2

to the fact that this is sort of made as an assessment.38 This [assessment] points
me to another issue. Its giving me sort of a clue that David and Heidi are in ca-
hoots. And David knows the kind of answers that are expected and he assess it:
Very good! Very good!
So, do you see how I reconstructed the situation? I try to reconstruct the situ-
ation from the clues that I have. Think of yourself as Sherlock Holmes and Dr.
Watson or Kathy Reichs, who is a female anthropologist. Think of yourself as
trying to piece together a story, but you dont really have the truth. So what you
are trying to piece together, what you are trying to document and explain in-
stead is the most plausible story, a plausible interpretation. You always go back
and you check: what is my evidence?39 What is my evidence here? I dont have
a need to know where the data comes from. What I try to do is piece the best
story together that I can.
One question I would have is how much David Suzukis name weighed in on
my analysis? If it hadnt been David Suzuki who was named, I just may have
played more of all these other things.40 Their nature is giving away the situation.
There are so many things giving away the situation that I did not need David
Suzukis name in there. So think of yourself as the anthropologist that Kathy
Reichs writes novels about or as Sherlock Holmes. You are taking this video or
transcript and you are trying to make sense and piece together what is going on.
The detective may be a good image to bring to your work. And you can start

38
The assessment was not articulated by the members. Instead, whatever preceded
the exclamation mark was heard by the transcriber as an assessment, a hearing
that was then marked and articulated as such by means of the exclamation mark.
All punctuation used in transcription should be taken as pointers to the compe-
tence of the transcriber and, therefore, constitute as an index to ways in which the
situation can be seen and heard.
39
This can be heard as an encouragement to do data-driven analysis, where every-
thing said in and as part of the analysis is directly linked to the available data.
This means that mere speculation is not allowed. Mere speculation consists in
statements that by their very nature cannot ever be tested. For example, statements
about what someone thinks, how someone feels, or what someone intends all be-
long to the category of the inadmissible, for it cannot be connected to the data at
handunless we are in a situation where the members to the setting formulate
what they are doing for the purpose of accounting for their and others actions.
But even in this case, the formulation needs to be taken for what it is, a part of an
inherently social action produced for the accountable rationality of the situation
rather than as a true statement about the inherently inaccessible.
40
Here again a critically reflexive articulation of the possible role that being fa-
miliar with one of the participants may have played in the analysis. Analysts al-
ways have to ask themselves about those aspects of our analysis derive from the
prefigured constructions (Bourdieu 1992), unless they allow themselves to practice
a science that does not know what it is doing. If we do not know how our precon-
ceptions do or may affect our analysis, we literally do not know what we are doing
and where our analytic results derive from. We simply reify the common sense un-
derlying our own preconceptions.
 DATA SESSION 1 (HEIDI) 49

your analysis. Try to piece together, in your analysis, a plausible narrative, a

plausible explanation. The only things that matter are the questions Where is
the evidence and how does it connect? Put together a plausible story.41 Ask
yourself, Where is the evidence? I will ask, Can you show me the evidence
for this or that? Or I may say, But you are invoking a concept thats really
from far away. Lets come back to the data and work upwards. So from the da-
ta upwards, try to come up with explanations and explications only after we
have described what is going on.42

41
Readers also need to hear, though not stated explicitly, that the analyst is ac-
countable to others for the analysis in the same way that the fictive or actual detec-
tive is accountable to victims, their families, perpetrators, and society to point to
the actual perpetrator rather than an innocent person.
42
The transcription was produced from a documentary presented by and featuring
David Suzuki, a Canadian naturalist, broadcaster, and environmentalist. The doc-
umentary in question features the Canadian Badlands (Fig. 2.1). David was ac-
companied by three children and a park ranger (Heidi). The situation, therefore,
was precisely of the nature that the instructor-analyst had produced a description
of.

Fig. 2.1 The five participants in the situation, David to the far left, Heidi to the far right, are
looking wards the camel-shaped configuration in the not-too-far distance; pyramid-shaped moun-
tains are mountains are visible further back, just as Heidi describes them in the conversation.
3

Data Session 2 (Vicky)

01 Vicky: So you may be on red you may be on amber, like some of those ques-
tions you can answer . . . you may be on green. Lets just look at first
impression. Would everybody just show me . . .

Im looking for clues right now. Vicky: So you may be on red you may be on
amber like some of those questions you can answer. You may be on green. Lets
just look at first impressions. Okay, so we only have one speaker here. So we
could scan for the moment I see [turn] one, Vicky, [turn] two, Vicky, [turn]
three, Vicky.1 So far I see that there appears to be only one speaker.2 We dont
see another person. So we might be looking at a lecture.3 So you may be on red
you may be on amber like some of those questions you can answer. You may be
on green. So she is, Vicky is talking about something. She is, we seem to be in
the middle of something, not at the beginning because there is something about,
like some of those questions you can answer.4 So it is about some questions that

1
As in chapter 2, the instructor-analyst has quickly glanced downward on the
page, without actually reading the transcription, who is speaking. He notes that
the first three turns all pertain to Vicky.
2
Note that the analyst says, there appears to be . . . rather than saying defini-
tively that there is only one speaker.
3
This is the first hypothesis, the first conjecture about a possible type of situation a
specific case of which that might have led to the production of this concrete tran-
scription. The incontrovertible (objective) data for this is the fact that there are
three (longer) consecutive turns, all of which were produced by the same speaker.
A lecture would be one possible case of a situation type that leads to one speaker
having all the turns.
4
The analysis highlights that the opening part of the transcription likely is from
the middle of the video. What might it be in the transcription that allows an analyst
to hear it as being from the middle of something rather than of the beginning? It is
not good enough to say that the analyst is an expert or has a lot of experience.
52 CHAPTER 3

they can answer; and about the different kind of states. They may be acting out
an analogy, or a metaphor is being used of which state you are in prior to an-
swering a question.5 It could be red or amber or green. Would everybody just
show me, would everybody, no, Lets just look at first impression. Would every-
body just show me. Okay, so they are in a situation, theres an audience. Theres
everybody so they are a bunch of people, lets just look at first impression would
everybody just show me.

02 Vicky: If you show green traffic light are you on green means yes I understand
it all, I can do the question, Im confident, Im happy. Um an amber
traffic light means that you . . . have grasped it a little . . . you you think
. . . you think you can sort of do it . . . but youre not really shhh, youre
not absolutely sure and a red traffic light means Ive got no idea, I dont
get it.

If you show green traffic light you are on green, means yes I understand it
all. I can do the question. Im confident. Im happy. Um an amber traffic light
means that you have. So we are in a kind of situation where they are giving an-
swers and where the audience would everybody just show me.6 It could actually
be that if we listen to this lets just look at first impression, we can hear it as
the person articulating something that just looks [like a first impression], she is

There is something about the talk that is self-explicating, so that beginnings sound
like beginnings. Here there is not such beginning a meeting or conversation part.
The talk provides clues about the setting of which it is a part. That is, the talk not
only has some content. It is a form of text that also refers us to the context. No text
ever stands on its own. If talk is tied in this way to the context of talk, then knowing
a language is equivalent to knowing ones way around the world more generally
(e.g. Rorty 1989). We may therefore not attribute it to an emanation from Vickys
mind, completely shaped by Vickys subjectivity. Once we have accepted this, we
have made a radical departure from the normal ways in which language is used: a
tool for ex-scribing thoughts into a public arena.
5
Even without knowing anything else than what Vicky has said so far, the analyst
takes the talk about the three color in the statement You can be color as the
description of a situation in which the colors are used to signal the state of the
person. This is much like the braking light of a car signals that the car is in the
state of braking, or like the turn light signals that the car is in the state of turning.
For humans, a red face often is a signal that a person is embarrassed (i.e., in a
state of embarrassment). How does an analyst come up with such hypotheses? It
may well be because of the experience of analyzing, which involves creating pass-
ing hypotheses or theories. It has been suggested that any such passing theory is
like a theory at least in this, that it is derived by wit, luck, and wisdom from a pri-
vate vocabulary and grammar, knowledge of the ways people get their point
across, and rules of thumb for figuring out what deviations from the dictionary are
most likely (Davidson 1986, p. 446).
6
Here Would . . . just show me is heard as an invitation to an audience, for the
benefit of which the speaker invites a larger group (everybody) to show some-
thing.
 DATA SESSION 2 (VICKY) 53

actually telling a mental narrative.7 What we dont have here are quotation
marks.8 It may actually be in a real situation, in a novel, if we have two levels
of text. One, what is, a person telling something and then the other and that per-
son also telling about another situation. So you may be on red and you may be
on amber, and it is about question and answer. Lets just look at first impres-
sion. So it may be that the speaker Vicky sets us up for a situation where were
teachers, we listen to Vicky and she is trying to tell us how to teach and maybe
that those who respond, just look at first impression.9 From your class, the stu-
dents, they might give indications with a flag red, the amber, or green, how they
feel about responding to the question. Hypothesis. Its a possible hypothesis of
who and what the story might be.
If you show green traffic light youre on green. So if she is talking about the
situation, about how to teach, and she is telling me she is a teacher. Shes asking
students to show the state they are in then a student raising a green flag would
indicate that she or he knows the answer to the teacher question. So it could be
that we are in a situation where this teacher explains how she is teaching in a
particular way and how she gets so quick answer to the question of how stu-
dents feel about, a quick indication of how students feel about their answer.
Again I can do the question, Im confident, Im happy. An amber traffic light
means that you have grasped it a little. Or it could also be that Vicky is the ac-

7
The analyst picks out a change in voice brought about by a change in the intend-
ed recipient. One form of talk is directed towards the students (turn 02), whereas
the other form of talk is like a narrative (turn 01) where the audience is invited to
look at a first impression. The analyst does suggest a shift to a different audience,
a shift in voice, even without hearing the original soundtrack, where the intonation
might give away such a different orientation.
8
These quotation marks would have been further indicators to the fact that a nar-
ration about the other event is occurring. The analyst then develops this idea of
two levels of talk, one directed towards others in some first situation and the other
one towards a recipient with whom the speaker is talking about the first situation.
9
Here the analyst clearly identifies the two levels. Turn 01 would be talk about the
events to be see in turn 02, a fact that the analyst expresses by saying that Vicky is
setting us up, the viewers, to watch out for something that she will subsequently
show or talk about. Previously he has noted two levels of talk. Here there is a fur-
ther specification as a prospective formulation, that is, as a formulation of what is
to come, which is then happening in turn 02.
In both ways of looking at the situation, knowing what will happen next appears
to be implicit in the talk. If the talk is prospective, being certain about what is to
come tends to be observed in situations where a presenter anticipates what the
audience is to see or hear. If the talk is retrospective, it describes what the particu-
lar audience has seen or heard. In both instances, the speaker (and audience)
know the referent of the talk. The analysts tasks lies, in part, in teasing out which
of the two possibilities was de facto enacted to have led to the transcription as it is
available.
54 CHAPTER 3

tual teacher and she is in the class and she introduces her method.10 She says,
she might have said, if it was a class situation, Okay students, I have here, I
have three traffic lights and from now on when I ask questions, I dont just want
to have . . . I dont want you to respond right away. I first I want you to think
and then I ask you to raise a flag and you indicate to me how confident you are.
If you say green then you feel confident.11
So at the moment I dont have enough evidence from the text [for updating
the hypotheses]. So it could be one situation where she actually talks to the stu-
dents about what is going to happen and how they have to respond, and how
they indicate that they are ready for a response, or she might explain it to an au-
dience of how shes teaching. See now I create, I have two hypotheses. I dont
know yet, I have no further evidence. From this text that much I can see.12

03 Vicky: Weve got a whole range of answers here . . . look at this see part one . .
. k . . . Right . . . 25.73 to correct to one decimal placccce . . . you look at
the next number and you say; is it five or bigger? Does anyone know the
relevance of that question? Hands down.

An amber traffic light means that you have grasped it a little. But you are not
really sure, you are not absolutely sure. And the traffic light means Ive got no
idea, I dont get it. Weve got a whole range of answers here, look at this, see
part one twenty five point seven. So now we seem, okay its three, there may
have been something. I dont know what it says on the video before. It could be
that they are in a natural classroom, she says Raise the flag, and then she says
Weve got a whole range of answers here.13 Look at the, see part one okay right

10
Here an alternative hypothesis is stated. The analyst is not yet certain whether
turn 02 is an actual teaching situation where the teacher Vicky introduces her
method to the students or whether Vicky is commenting on her teaching. In fact,
Vickys talk in the classroom also would be talk about the classroom, where she
describes to the student of what theyteacher and studentsare projected to do
when questions are asked and answers to be provided.
11
This is the generation of an alternative hearing as a form of producing variation,
which serves as a hypothesis the probability of which is to be tested with subse-
quent data.
12
In this paragraph, we observe meta-talk about how to conduct the analyses. The
instructor-analyst explains the generation of hypotheses, which, at this state in the
analysisi.e., after two turns of talkcannot be further specified with respect to
their probabilities. That is, the probabilities of hypothesis H1 and hypothesis H2
are equally likely based on the data analyzed so far, that is, p(H1|data) =
p(H2|data) = .5.
13
He previously suggested the different voices, or the voice directed towards a
different audience. Here the analysis picks out that the talk appears to be describ-
ing the situation, that is, the talk is formulating what is happening. This formulat-
ing is happening in situation to make salient something. That is, it is a way of the
accountably rational aspect of the situation: it can be talked about and seen by the
audience. In this, the description also functions as a prescription, an instruction to
 DATA SESSION 2 (VICKY) 55

twenty-five point seven three, to correct to one decimal place. Oh, to correct,
now we are to correct to one decimal place. Now we are in a situation, one of
the things I would pick up here is Why would the person say if shes a teacher,
correct to one decimal place? It could be during a math class. And it might be
because, a history teacher, geography teacher might not care because one deci-
mal place, its [not important]. It could be a science class, physics. In different
science classes, my hypothesis would be more like a physics teacher or a chem-
istry teacher than a biology teacher. Okay? Hypothesis!14
So its five. Is it five or bigger? Okay so here, what all of a sudden what is
happening, okay, twenty-five point seven three to correct to one decimal place,
you look at the next number and you say at the next or last number, whatever
its seven three, or seven four. So what happens here in turn three? There seems
to be a teaching going on: on how you round to one decimal.15 Seven three . . . is
it five or bigger? Does anyone know the relevance of that question? Hands
down.

04 Newscaster: Vickys using traffic lights together with the controversial no hands
up, policy.

Newscaster. Vickys using traffic lights together with the controversial no

hands up policy. Okay now we seem to have a newscaster, perhaps a voice-

see something. If it is in the classroom, it would then be an instruction for the stu-
dents to find in their own actions or situation a state that here is intended to be
present. The Raise the flag can be heard as an instruction, and the We have a
whole range of answers here is a description of a situation where the students
have indeed raised the required flags. It is also an instruction for seeingto the
studentsthat this state, the raised flags, is what ought to occur after a question
has been posed.
14
The analyst draws on some unstated familiarity with the fact that rounding
would be less common among history and geography teacher than for a mathemat-
ics and science teacher. Among the science teachers, it is more likely a physics or
chemistry than a biology teacher. The analyst thereby indicates being considerably
certain about Vicky being teacherthe probability of this being near certainty,
that is, p(H) 1. He assigns different degrees of probability to the different subject
matters, which we might capture in the following relation: {p(HGEO|data),
p(HGEO|data)} < p(HBIO|data) < {p(HMAT|data), p(HPHY|data), p(HCHEM|data)}, where
data = {teacher uses rounding of second digit following decimal}.
15
Whatever was said before the statement there seems to be teaching going on
is further data. This data supports the hypothesis that this turn 03 is a teaching
situation rather than the alternative hypothesis that this is a situation in which
Vicky talked about here teaching. We can represent this has p(HABOUT-TEACH|data)
< {p(HTEACH|data), where the data = {instruction You look at . . . and say Is it
five or bigger?}. That is, the analyst hears this as an instruction of how to go
about deciding whether to round up, as opposed to the unstated rounding down.
56 CHAPTER 3

over.16 There seems to be an explanation of what is going on, if this is the case.
We are in a real classroom, and there is a commentator explaining to us what is
happening. We also see that there is a no-hands-up policy. You have to indicate
with flags, Vickys using traffic lights together with the controversial no-hands-
up policy. Now we have evidence, it is a disconfirmation that it is Vicky ex-
plaining in turn 03 to an audience, it is Vicky in a class we have a second com-
mentator explaining to us what is actually happening here.17 It is a policy, the
traffic lights, where students have to indicate their response, levels of response.
Not only indication that they want to speak but in fact the levels of confidence
that they have for their answer. And theres also a policy in this classroom of no
hands up.

04 Newscaster: Vickys using traffic lights together with the controversial no hands
up, policy.
05 Vicky: (whispers) hands down, hands down

Vicky, hands down hands down. Oh, the newscaster now has actually
perhaps it was not even visible on the any hands up or down. But the newscaster
could see, O-four ((turn 04)), O-five ((turn 05)) being teaching, actually the
newscaster or the commentator commenting on this teaching lesson might antic-
ipate, see, tell the audience, Now you need to watch, its a hands down policy
and the next instance is the teacher implementing no hands up policy.18

16
Notably, the analyst does not simply assume that a newscaster might be shown
and heard but also generates the hypothesis that the text was heard in the form of a
voice-over. Here, the graduate student, in denoting the speaker as newscaster,
actually gives away information. But this information is not required to hear turn
04 as a comment about Vickys teaching. As stated, without the preceding or suc-
ceeding line, the statement may be heard as a backward-looking description of
what has just happened; but it also may be a prospective description of what is to
come. In fact, it is more than that: It is also an instruction for intentionally looking
at what is coming as an instance of Vicky using traffic lightsi.e., red, amber, and
greenas a teaching strategy. In other words, the projective description functions
like a caption to an image, which both describes to the reader what there is to be
seen and, in the same move, instructs readers what to look for or how to orient
their gaze. As a voice-over, it instructs the viewer of the video to see what is cur-
rently occurring as Vickys traffic-light-cum-controversial-no-hands-up policy in
action or as an instruction to find in what is happening evidence for the traffic-
light-cum-controversial-no-hands-up policy in progress.
17
The analyst suggests that besides Vickys own commentary, there is a second
commentary, a second voice talking about her teaching. What Vicky is showing or
exemplifying is the traffic-light-cum-controversial-no-hands-up policy; and in the
classroom, this may be staged for the sake of the video or actually occurring, as
stated in a previous hypothesis.
18
The analyst has noted a contradiction between the newscasters comment that
there is a no-hands-up policy and Vickys whispering Hands down, hands down.
The hypothesis stated here is that the newscaster invites the audience to see the
 DATA SESSION 2 (VICKY) 57

06 Vicky: Theres some very enthusiastic students there who were dying to speak
out and . . . they find it boring if they cant . . . um so they . . . I suspect
they do get frustrated (break to class) . . . as a teacher you sometimes re-
sort to picking them because you want the lesson to move forward . . .
and . . . it often can be that five or six students will have a dialogue with
the . . . teacher all the way through the lesson and that the other 25 will
sit . . . dormant.

And then Vicky is actually commenting. Vickys next turn is Theres some
very enthusiastic students there who are dying to speak out and. So we are
probably in the classroom where there is a no-hands-up [policy] and the signals
are a group policy. Vicky makes available to us, she says Hands down hands
down and then she comments upon or elaborates how she does it, this one ((turn
05)) here. Hands down is actually a description. She says what she wants to
happen. Whatever the students do, we do not get it here. But then we have here
((turn 06)) a reading. The teacher reads what she sees or how she understands
all of this, it may have been asked, some very enthusiastic students. And they
forgot ((in turn 05)) about the rule that she implements: namely hands down
hands down, no hands up.19
And so we can read here what might have happened. We can make up a hy-
pothesis of what might have happened before even though in this case I dont
have what has happened before. I do not have the beginning of the story or what
that video was being used for. The guys are speaking; they find it boring. So she
provides a commentary ((in turn 06)), she articulates what she sees in all these
hands up, how she understands all these hands up. But that she wants, whisper.
Why whispering? And it is Hands down hands down and not I told you, didnt
I tell you that we have this, its called hands-down policy. This may be an indi-
cation that this is a demonstration lesson and the students did something that
they were not supposed to do.20 And now that they are so eager, they really want
to participate even though this was this demonstration lesson for her method.

policy enacted even though the students apparently have their hands raisedfor
Vicky did not have to say Hands down! if the students already had their hands
down rather than up.
19
In this paragraph, the instructor-analyst distinguishes two types of descriptions.
The first one occurs in turn 05, where she articulates what is to happen because
this is not currently the case. In other words, she provides an instruction, which
takes the form of a description of what the future state ought to look like: Hands
down. The second description occurs when Vicky reads what she sees and, in
this, what she makes available to the audience to see: The hands are up despite her
hands-down policy because theres some enthusiastic students there who were
dying to speak out.
20
The analyst works with a variation. Why the description that Hands down,
hands down was whispered? Why would the teacher, if she were teaching her
class under normal circumstances, whisper rather than doing as the analyst de-
scribes. This description is an alternative hypothesis. The analysis then uses this
observation as documentary evidence for the type of situation: this is a demonstra-
tion lesson for this method rather than a regular lesson. The students are to exhibit
58 CHAPTER 3

Now, I do this analysis not knowing the video. But if you already know the
video, then pretend to go in as if it was the first time. When I write in my arti-
cles doing first-time-through analysis, this means that nothing that has hap-
pened after, nothing that I knew only afterwards was used in my analysis. Some
may find it hard [implementing this policy].21
And it is not something that comes overnight. The other issue is this: It is ac-
tually very helpful that we can do that ((see something as an instance of a cate-
gory, such as force)). Once we are familiar, we recognize these situations eve-
rywhere. So it is actually an evolutionary product: Knowledge-wise it is of an
advantage that we can recognize a situation very quickly. But there is also a
danger. We recognize the situation, we dont even worry about it and we may
actually be wrong.22

compliance with the hands-down policy, which they are not, so that Vicky is re-
minding them of the policy. She is not only reminding them of the policy, she is in
fact pointing out that they have not acted according to the instruction, No hands
up. Using this as data, the probabilities of two alternative hypothesesH1 =
normal lesson, H2 = demonstration lessonare updated to p(H1|data) > p(H2|data),
in other words, the probability of the first hypothesis is higher than that of the se-
cond hypothesis.
21
One common mistake in qualitative data analysis is to take the final outcome of
an event and then go back and read the beginning in view of the outcome. One of
my favorite examples for showing how fallacious such an approach can be is when
one partner in a couple says, Honey did you do the dishes? Most readers may
be tempted to say that a question has occurred. However, if the recipient were to
say, Why do you always have to nag?, then the event now is at a point that the
first speaker has to defend him-/herself, because s/he has been heard to nag. That
is, from the perspective of the event, the participants now have to deal with the
nagging issue rather than with the response to a query about the dishes. If we are
interested in how events evolve, and how they sometimes get out of hand (e.g.,
were this couple succumb to a spiral of back-and-forth statements until they end in
a major argument). Suggesting that there was a question at the origin misses the
entire development that unfolds from the way the statement was heard and affected
the situation in its unfolding.
22
Wherever we look, we see familiar things, or rather, instances for which we have
concepts. We look outside and we say, Look at the pear tree. What we see is in
terms of the concept pear tree. We look at certain interactions and we might say,
This is a pushy person even without having to analyze the situation. This is so
because the pre-constructed is everywhere. The relationship between the social
world, the ways in which bureaucratic (institutional) and professional discourse
produce accounts (records), and the description of professional scientists (sociolo-
gists, psychologists) are displayed in Fig. 3.1.
In some sense, this is an evolutionary advantage, as the analyst points out, be-
cause it allows us to capture very quickly the essence of a situation and act without
having to think very much. But, as the analyst points out consistent with critical
sociologists (e.g. Smith 1990a) and critical psychologists alike (e.g. Holzkamp
1983), there are also serious disadvantages. One of the most important ones is that
 DATA SESSION 2 (VICKY) 59

The method that I am trying to articulate [here for you] is this: working
blank. Starting blank, I am trying to use only what I have available. I am gener-
ating hypotheses: This could go on, this could happen, oh no this cannot. And
then later on I say, Oh no, this hypothesis described the little girl, its not girl
one thats the house, its the other one.23 Because all the evidence thats after-
ward appears to indicate, to the moment that we got that girl-one is the little girl
and girl-three is the house. The same is the case here. So we tried to do that.
Now your analysis might slightly differ because you want to show something
different.24 Or you might say: In this situation, in my paper I want to show how
it is. Or I want to use a first-time-through method in order to show whats go-
ing on. What these people actually make available to one another. And so I on-
ly go with what I have. Initially I have no clue what is going on. But as you see
here, there are many things that I can use as clues in making something out of it.
To begin, what I do is this: I just write down what I kind of think and I ask
myself okay, On what basis do I say that? See you have to almost train your-
self not to draw on stuff, not to infer. We looked at the video and we read the
situation. We were talking about power and we were talking about what some-

we merely contribute to the reproduction of the rules of relations, institutional and

institutionalized differences that play themselves out in struggles of pow-
er/knowledge. The only way to undercut the automatic reproduction of the differ-
ences and struggles is to critically interrogate ones own instruments of construc-
tion. The forms of analysis proposed here, which begin with the relations prior to
attributing explanatory concepts, are ways of deconstructing what is actually hap-
pening and how power/knowledge phenomena are actually produced.

Fig. 3.1 The relations between different forms of discourses that let researchers find in
field settings what they have previously found there and theorized. The three levels of talk
are clearly articulated in the distinction of the events on the left, in the center, and on the
right.
23
This is a reference to another analysis that was conducted in this class, which
turned out to involve little girls in the course of enacting the story of one of them. It
functions here as another example for the approach presented: the continued up-
dating of the probabilities of hypotheses with additional data.
24
Not all analyses have to work as that conducted here by the analyst. But if re-
searchers are interested in understanding the unfolding of events, then they have to
place themselves within it rather than using the final outcomes in the description
and theorization of its earlier stages. This is so because from the perspective of the
participants, who act upon and react to what has happened so far, the endpoint is
not available. Even the event itself is an event*-in-the-making, where the very na-
ture of the event is at stake (Roth 2014a).
60 CHAPTER 3

one might think.25 But, see, I dont know the video. Nobody here ((graduate
student audience)) knows what was before and what is after the instant de-
scribed. In this situation we pretend. We come in there, into the conversation,
and all we have is what we hear. This already gives us many clues about what is
going on, what is presupposed, what people know or possibly know and what
they possibly dont know.
The person, the interviewer, we dont know if it is a he or a she. Even that is
not known. But there might be evidence in the text that allows us to eliminate
this question. It could be a male or a female. Here we already have the answer.
For Sherlock Holmes a smoking cigarette is a give-away, is an indicator.26 For
me, Vicky is just another such indicator. But it does not tell us the whole story.
We have to piece together the story by eliminating some alternatives and leav-
ing others open. It could be that [the transcriber] sets us up, using a pseudo-
nym.27 [The transcriber] could have said, Im getting this to you [instructor].
And he is saying, This is a woman and really it is a man. But at the moment we
do not know. Some names, such as Taylor, we would not know if it was a boy
or girl, a male or a female teacher. Vicky is female teacher. But we do not
know. [The transcriber] might have said, Oh I want to get [the instructor].
Right? So at the moment I take it in this way: Okay, it could be in here. But
nowhere in my information have I brought in that it is a female teacher.28 [The

25
The instructor-analyst makes reference to a video he had shown, involving chil-
dren in the construction of model buildings from straw, and where the graduate
students had used concepts such as power to explain the relation that they were
seeing. They suggested that everything in the group they watched was controlled
by one of the children. Unbeknownst to the nave analysts, the children themselves
had described and experienced their relations of power very differently. This,
therefore, served as an example for the problematic ways in which concepts may
functionmuch like horoscopes that predispose their perusers to find confirmatory
evidence in virtually every step of their lives.
26
The cigarette, the indicator, is the hard facts, incontrovertible results of the
specific events that produced them, and documentary evidence of the type of phe-
nomenon that is to be described. These facts are the data that are used in the modi-
fication of the hypotheses about the nature of the phenomenon of interest.
27
Some students in these classes do indeed attempt to make the instructor-
analysts task more difficult by choosing difficult situationssuch as where the
same person speaks to different audiences, where some individuals talk very little,
or where translation between languages may be one of the characteristics of the
situation.
28
This is a reminder to not explain relations based on a-priori characterizations,
such as the fact that Vicky (likely) is a female, but to analyze relations first and to
work out anything that is itself material or instrument in the production of gender
differences. If we were to begin with gender, then it is easy to identify differences
between any two relations one involving a boy the other a girl and to suggest.
However, we do not know whether these differences are coincidental, that is, par-
ticular to the relation or whether these are invariants that hold for all girls and all
boys.
 DATA SESSION 2 (VICKY) 61

transcriber] knows [the gender], because he looked at it. I would refrain from
saying something definitively. My most likely hypothesis is that it is a female.
The same earlier, when you said Little girl. When we already put together
three pieces of information that we may or may not have found out whether the-
se are three little girls. And we have a teacher but the teacher and the interview-
er, we probably would have figured out very quickly the gender. If we could
from just by looking at these transcripts without knowing the gender of these
people, figure out the gender, that tells us that there is a gender issue, that some-
thing in this discourse is different for different genders. And that is actually a
much stronger for me much stronger evidence about what our socialization
does. By putting boys and girls on very different trajectories then all the dis-
course about gender that people have.29 But you do that only if your research is
based on a gender, based on evaluating the differences in gender.

29
I much admire V. Walkerdines work on gender and mathematics (e.g. Walk-
erdine 1998). But I do find in her analyses frequentlythough much less frequent-
ly than in the work of much of mathematics (or science) education research on
genderreadings in which the differences between gender are already accepted
and theoretical concepts come to be mobilized for the purpose of doing so.
Charlotte constantly asks questions which her mother answers patiently and
explicitly. Throughout the afternoon, she engages Charlotte in essentially
domestic taskssome commonplace, like helping mummy put the shopping
away; others involving making things. Charlotte helps her mother to make
muesli, and this becomes a site for number work. Unlike Pennys mother, she
does not have to resist her daughters demands because she immediately and
consistently sets up what she must get done alongside strategies for amusing
her. This form of mixing domestic work and play forms a particular mode of
regulation, a way of disciplining the child. (Walkerdine 1998, p. 48)
Here, regulation, disciplining (Foucault 1975), and domestic work become ex-
planatory terms rather than being the end result of deconstructive readings. We do
not see that disciplining has a double function, that it not only means imposition of
an exterior will but also the development of the practices of a discipline. Moreover,
we do not see in parallel the work involving boys. In particular, the differential
mode of production of mathematics is already assumed rather than shown before
showing that these modes exhibit differences that are associated with gender dif-
ferences. That is, I personally would analyze relations of people and categorize
these. If it turned out that one type of relation only involved girls whereas another
type of relations only involved boys, then there would be strong evidence for the
production of gender differences. The relations themselves are the source of these
differences. This approach also would be more consistent with the theoretical ap-
proach of L. S. Vygotsky, who suggested that every higher psychological function
specifically and personality more generally is the result of the ensemble of societal
relations that a person has participated in (Vygotskij 2005). If the societal rela-
tions are the sites of the production of higher psychological functions and of per-
sonality, then we need to analyze relations and show the work accomplished and
the results achieved. We must not use the results, gender differences, and then
62 CHAPTER 3

If I wanted to see that there is something about gender, then I would love for
someone, for a research person, to provide me [a transcription] so that I do not
know whether it is a boy or a girl. And if I can figure out from the discourse that
there are differences, that would be much stronger evidence about gender dif-
ferences, if I can identify it without knowing beforehand.30 It is very easy to say,
okay here there are a teacher and a girl, and here are a teacher and a boy. There
are differences; and these differences are because the participants are of differ-
ent gender. That for me is a weak argument.31 If you give me a transcript and if
I can, just by seeing differences in discourse, pick out all the boys and all the
girls in this class, then there would be strong evidence that there is a bias, that
there is a difference in which the different genders are addressed. Because if I
can identify gender without even seeing these people, that is strong evidence
that there is difference in this conversation, in the way these people relate. And
if this is such, then that is much stronger evidence for making a hypothesis
about how discourse, how interactions, not just discourse but interaction, forms
of interaction put people, boys and girls, on very different trajectories.32

point to any relation and say that it shows the difference in gender. In the philoso-
phy of science, such a move of going from gender to differences in relation is
called using the explanandum (that which is to be explained) as the explanans (that
which is supposed to do the explaining). The relations produce gender and gender
differences not the other way around.
30
Readers interested in gender issues might find it an interesting and rewarding
exercise to analyze transcriptions without knowing beforehand the gender of the
speakers or even of the nature of the speaker (e.g. student or teacher). They would
then be focusing on the relational work that comes to be accomplished, which may
or may not be associated with sex differences. With a little more experience, the
analyst may then conduct such impartial analyses even when confronted with the
actual videotape. They may then be attuned to the fact that not only institutionally
designated students are learning while talking to institutionally designated teach-
ers. Instead, more symmetrical analyses become possible in this approach, where
we can see how teachers learn while teaching and where students teach while
learning (Roth and Radford 2010). Vygotskys zone of proximal development may
then be shown to be the worksite for the learning of all participants. Far too many,
if not all studies of mathematics or science classrooms are blind to teacher learn-
ing that occurs while the individuals engage with their students.
31
It constitutes in fact a form of confirmation bias, which not only characterizes
many everyday practices of laypeople but also scientific research (Couzin-Frankel
2013; Nickerson 1998).
32
The transcription analyzed here was produced from a video on SchoolsWorld.tv.
The video is part of a series on secondary assessmentassessment for learning,
here addressing questions and answers. There are three levels of talk: Vicki teach-
ing mathematics, Vicki talking about how she is teaching, and a narrator providing
voice-over comments. The description of the video reads:
In this programme we visit Valentines High School in Ilford where the
teaching and learning group are spearheading a whole-school strategy for
Assessment for Learning.
 DATA SESSION 2 (VICKY) 63

At an evening meeting we see teachers brainstorming ideas and offering

feedback on experimental work undertaken in their own departments.
Science teacher Richard Griffin uses Key Questions with Year 9, but also
grapples with more complex ways of questioning pupils to deepen their un-
derstanding.
Maths teacher Vicky Inman combines Traffic Lights and No Hands
Up techniques with her Year 9 class. She comments on their value as tools
for inclusion in a subject that traditionally alienates many pupils.
In addition, Vicky discusses the excitement of getting positive responses
from students rather than the usual sea of blank faces.
Back at the meeting, we hear views on the deeper meaning of AFL that lies
beneath the surface ticklist of strategies.
(http://www.schoolsworld.tv/node/908?terms=719)
Just as the analyst hypothesizes in the opening lines of this chapter, the gradu-
ate student had transcribed a fragment from the middle of the video rather than
providing the transcription of the beginning (at about 7:15 of the video provided at
the website indicated). Such a beginning, because it is a beginning, would have
provided even more clues to allow the audience to make sense of what the video is
about and how to hear what is coming.
In turn 01, Vicky is teaching students (Fig. 3.2a); in turn 02, she is talking about
her teaching, initially hearable as a voice-over to her class and then full face-on to
the camera in an interview situation (Fig. 3.2b). In turns 03 and 05, we are back in
the classroom, whereas in turn 04 we can hear a female speaker in voice-over
mode while Vickys teaching is continuing. Turn 06 again begins in voice-over
mode while viewers can see her classroom and then turns to the interview mode.

Fig. 3.2 a. Vicky talking about her teaching. b. Vicki teaching mathematics in her class-
room. (Source: http://www.schoolsworld.tv/node/908?terms=719)
4

Data Session 3 (Bullrush)

01 I: And the girls are putting their hand up too.

02 P: Were tough out west. [girl tackling boy in background]
03 I : And those girls, more than hold their own.

So, it is in the middle1, and I: says, And the girls are putting their hand up too,
okay and I: said, the girls are putting their hand up, too. So, the girls are put-
ting their hand up, too. So where is there a place where we find something
about putting the hands up . . . school. So first hypothesis, we might be in a
school.2 Were tough out West. So, between I: and P:, were tough out West.
Girl tackling boy in background. We do not know about P:, Were tough out
West. If that is in response to here ((turn 01)), P: might be a girl talking about in
reference to the girls. And those girls, more than hold their own. So this is a

1
It is almost unremarkable, but definitely an important recognition that the begin-
ning of the transcription does not constitute the opening of something. If it were an
opening, the analyst would recognize the work related to opening a conversation.
Here, the And . . . makes available that something else has occurred and been said
to which whatever follows the And is an addition. In a similar way, analysts rec-
ognize the work of ending a meeting. The early work in the field of conversation
analysis was precisely focused on those forms of work that people accomplish, for
example, in talk on the telephone, to open, maintain, prolong, and end the conver-
sations.
2
The analyst has taken the statement putting their hands up as documentary evi-
dence of some situation that is hypothesized here to be school. Taking the unfold-
ing transcript as a treasure-trove of documents that are evidence of some phenom-
enon that never expresses itself in its entirety but always in one or another
manifestation. These manifestations may actually appear contradictory and not
resemble each other in the way the members of the same family do, who may not
have any physical or psychological characteristic in common.
66 CHAPTER 4

comment to that.3 And P: . . ., were tough, . . . and the girls. See, because when
we look at the turns, we do not know what the turns, especially in the beginning,
are referencing. And because we are not part of that situation, we do not know
what the present subject is, what the talk is about. But here, if girls is the sub-
ject, and this is a reference to girls are tough in the West, and here we have a
girl tackling a boy, so confirming that statement.4 And this is further confirmed
by and those girls more than hold their own. So, I: makes that statement and
those girls than makes reference to that again. Some type of kids are playing in
the background.

04 I: Do you ever worry about getting hurt? [kids playing in background]

05 M1: No
06 I: Have you ever been hurt?
07 M2: Guys
08 M1: Ya
09 M2: Yeh you guys, were supposed to be playing bullrush (yelled from behind)

Do you every worry about getting hurt? So now, Do you ever worry about
getting hurt? So may be this I:there is a conversation with girls about girls.
And the questionthis ((kids playing in the background)) may or may not have
anything to do with the talk, because it is in the background. But here, Do you
ever worry about getting hurt? To girls that are tough out West and that hold
their own. Do you ever worry about getting hurt? M1: no, Have you ever been

3
The analyst says that turn 03 is a commentary on turn 02, which possibly is the
result of hearing the demonstrative pronoun those. In part, because the analyst
finds himself in the middle of a conversation, he cannot know with any precision
what the subject of the conversation is. Finding this subjectVygotskij (2005)
calls it the psychological subject as distinct from the grammatical subjectfrom
the documentary evidence actually provided, is part of the analytic work to be de-
scribed and exemplified in part B.
4
In this situation, the subject of talk is identified as girls, which then makes the
talk intelligibly a sequence of statements about girls. Those girls then refers to the
girls that have become, some time earlier in the conversation or with a lead in
from earlier parts of the interview the topic of this talk. That is, the demonstrative
pronoun those is one of the devices used to anchor one statement in the same situa-
tion and to the same subject of preceding talk. These anchors are what allows the
analyst to generate hypotheses about the possible situations to which that talk
might belong. If language games are irreducible ensembles of language-in-use and
the activities to which they belong (Wittgenstein 1997), then the analysis can be
understood as a kind of game where a situation is to be described when only parts
of it are visible, illuminated by the beam of a search light. If the distinction be-
tween knowing a language and knowing ones way around the world has been
erased (Davidson 1986), then it is possible to arrive, through rigorous data analy-
sis, at the identification of a limited number of societal activities that could have
led to the production of a particular conversation. The transcription of the conver-
sation then is a verbal protocol of that situation. That protocol exhibits the ways in
which participants to the setting account for their actions.
 DATA SESSION 4 (BULLRUSH) 67

hurt? Been hurt? Guys. So this is, we do not know whether the guys is in . . .
reference to that ((i.e., been hurt)).5 We do not know, there is no punctuation
giving away what it might be.6 M1:, Ya, Yeh you guys, were supposed to be
playing bullrush, Yea, Yeh you guys. Ya, actually, really.

09 M2: Yeh you guys, we're supposed to be playing bullrush (yelled from behind)
10 M1: Ya, actually, really . . . ya, I have really been hurt.

5
We can hear what the analyst says as a warning to make all-too-rapid conclu-
sions about the co-location of statements, or verbal statements and transcriptions.
Here, the fact that the transcription reads Guys does not mean that it is in refer-
ence to what has been said in the transcript before. The analyst does not articulate
any alternative situation description, but the Guys might actually belong to the
play in the background rather than to the conversation between I: and P:, which
now also includes M1:; as apparent from turn 09, M2: is yelling from behind and,
therefore, not likely to be part of the main conversation.
6
The transcriber would have produced the punctuation. In these considerations,
the analyst takes this transcriber in the same way a court of law takes a witness.
No longer are the data themselves at issue but how a more-or-less knowledgeable
witness sees and hears the situation based on his/her familiarity with cultural and
physical context. But the witness is not taken literal, that is, as somehow repre-
senting or expressing the truth. Instead, the very trustworthiness of the witness
has to be established and weeded out in the course of the analysis. We can look at
the situation through the physicists eyes (Fig. 4.1): She approaches her observed
data as filtered through an instrument, her task being to distinguish between hy-
pothesis 1 or hypothesis 2 as source of her observed data. When considering the
transcription, the analyst has to take into account the possible filtering that has
occurred between what has happened, what has been recorded, and how this re-
cording is expressed in the transcription. The analysts task is to find out which of
these hypotheses is more probable.

Fig. 4.1 A physicists view of the relationship between hypothesized phenomenon and observed
data given that what is observed (data) has been filtered by an instrument (witness).
68 CHAPTER 4

So this one ((turn 10)) responds to this one ((turn 09)). And then I:, What
happened then? So I: is, this is taken out of context. It could have also been
something like an interviewer, or a researcher, talking to girls about their expe-
riences. And what happened then? So I: is interviewing, it could be a teacher,
but it could be some other form of interviewer.7

11 I: And what happened then?

12 M1: I think I got up and dealt with it.
13 P: I think were wrapping kids up in cotton wool a little too much . . . um . . .
years ago a good parent was somebody who just let kids play. Nowadays a
good parent might be considered to be somebody [who] takes them to danc-
ing lessons . . . and rugby practice.
14 I: there may be some people, some parents who think, hmm, I dont want my
children, my little girls, playing bullrush! What do you say to those peo-
ple?

And what happened then, M1:, I think I got up and dealt with it. If M1:, if
this is about tough girls, Do you ever worry about getting hurt, M1:, No, then
we have M1: as a girl, Ya, actually. I think I got up and dealt with it.8 P:, I think
were wrapping kids up in cotton wool a little too much . . . years ago a good
parent, P:, parent, a good parent was somebody who just let kids play.9 Nowa-
days a good parent might be considered to be somebody who takes them to
dancing lessons and rugby practice. There may be some kind of people, some
parents who think, I dont want my children, my little girls playing bullrush!
So here is the kind of situation where we have . . . okay, about M2:, my hypoth-
esis is, so far, that M1: is a girl. Look here, I think were wrapping kids up in
cotton wool10 . . . here it could have been a girl ((turn 12)), but here ((turn 13)) it

7
In this paragraph, three hypotheses are generated about the nature of I:, who
might be some generic interviewer (e.g., journalist), researcher, or even teacher
talking to the girls about their experiences in the kind of setting of which the tran-
scription is a protocol.
8
The chain of reasoning here is based upon the current subject, which is stated as
the conditional: If this is about tough girls. That is, if the current topic is tough
girls, then M1: is a girl. This is a hypothesis to be dis/confirmed by means of sub-
sequent documentary evidence (data) in the transcription.
9
P:s statement about parents can be self-referential, P: as a representative of the
group parents. This is the hypothesis stated here. But the statement could also
be about parents without P: actually having to represent that group of the popula-
tion.
10
This statement, by being about kids, has a high likelihood to have been articulat-
ed by an adult. It is not just the content of the text that matters but also the context
that each sets up simultaneously and reflexively with the situation. That is, speak-
ers do not just articulate and thereby pick out something from the situation but
also they pick up the surroundings. In and through talk, an aspect of the entire
Lebenswelt (lifeworld) comes to be made visible. In fact, every little stretch of talk
reflects the entire speech situation, itself integral and constitutive part of the socie-
tal activity in its coursemuch as a raindrop reflects the entire world of which it is
 DATA SESSION 4 (BULLRUSH) 69

sounds more an adult, because this one is in response ((to turns 11 and 12)).
M1: No. Yea, actually really, I think I got up and dealt with it ((reading the M1:
parts)) some parents think, I dont want my children . . ., M1: If you think you
cant handle it, well then, dont play, thats pretty much it.11

15 M1: Well, if you think you cant handle it, well then, dont play . . . thats just
pretty much it.
16 I: So what do the parents think?
17 P: Some parents have come and asked about it. Theyve wanted to be reas-
sured . . . but I think generally, Ive had really good support from them.
[kids yelling and waving there arms in front of the camera]. I think our un-
derstanding of what is safe really means, is changing. And actually, kids are
safe doing things, that, maybe we have thought, werent safe . . . for quite a
few years.

So whats next? P: says, some parents have come and asked about it. Theyve
wanted to be reassured . . . but I think generally, Ive had really good support
from them. I think our understanding of what is safe really means is changing.
So, what we have . . . P: is a person to whom parents come and asked about the
rough-and-tumble play.12 What happened then . . . years ago a good parent . . .

an integral part (Vygotskij 2005). Phenomenological philosophers articulated this

insight in observations such as that of the pen, notebook, or ink well that are not
perceived as objects in their own right but always already as integral parts of the
writers office and desk (Husserl 2008). There is an integrated whole, both in
terms of the objects that surround us and in terms of how the past may become
present again when we return to the office where we had left it and had left off in
our work. These connections are part of what allows the analyst to make good,
highly likely hypotheses about the kind of situations that produce the transcribed
talk, which itself is an account of that situation.
11
Here, then, the analyst has distinguished the nature of the speakers M1: and P:,
the former likely being a girl, the latter an adult, both responding to I:. It is in and
as the talk that these characteristics are made salient: in the way the speakers are
positioned and position themselves with respect to the current subject of the con-
versation. If the subject is girls, and M1: speaks to the subject as a member,
then M1: is a girl. P: is talking about kids that we are wrapping up, where the
relationship to the subject is framed in terms of an adultchild juxtaposition.
12
Here, as before, P: not only talks about something, the content of the text, but
also co-articulates relations significant to the present situation. Although parents
have come and asked about it does not have to mean that the parents have come to
him/her, a few words later the person says Ive had really good support from them.
This increases tremendously the likelihood of the hypothesis that the parents have
actually come to the speaker. Although the current issue has to do with the rough-
and-tumble activities, the speaker co-articulates being in a position to whom par-
ents come to talk about these activities, and perhaps about the fact that girls also
engage in these activities in that specific place, that is, in a place where such activ-
ities are not normally expected. This talk therefore does not just make available a
70 CHAPTER 4

you see, P: talks both as a parent here ((in turn 13)), something, but here some
parents have come and asked me about so whatever they are talking about P: is
in a situation of responsibility that parents come to and ask and to provide sup-
port to kids. We see there are kids in the original situation; there are lots of kids.
It might be in a summer camp . . . one could imagine there is a summer camp
and P: is one of the persons employed or responsible there . . . but the first could
be just a throw-off those girls kids playing in the background. Then here, M1:
and M2: are responding . . . it might that be another girl is saying were sup-
posed to be playing bullrush, a rough-and-tumble game. Okay, yelling, so it is
something . . . oh here, look, there may be some parents they think, what are you
saying to those people so . . . M1:, if you think you cant handle it, dont play.
See, M1: is asked, this sounds like a question because this (({turn 14 | turn
15})) is a {question | response} pair, and here . . . what do you say to those peo-
ple? Little girls. I dont want my children, my little girls playing bullrush. So I:
is asking about those little girls. That would confirm to me that they are in a
girls camp or Girl Guides13 or some organization like this; and I: sounds like a
person interested, it may be a journalist, it may be someone interviewing those
[individuals] . . . because I: is asking the questions M: and P: responding. So it
is I: the outsider coming in, the M:s and the P: are part of that situation.14 P: is

specific fact. Instead, the statement becomes a raindrop reflecting the Lebenswelt
of P: as a whole.
13
Readers may note that a few statements earlier, the analyst marked as a poten-
tial throw-off the fact that there are girls playing in the background. Although
only girls are the topic of talk, this does not mean that boys cannot be part of the
situation. It could well be that the scene is part of a camp for girls, or a place
where Girl Guides meet; but it could also be that the girls become the topic of talk
because of the rough-and-tumble play that is uncharacteristic for the kinds of ac-
tivities that girls normally participate in.
14
Institutional relations, such as parentchild, teacherstudent, principalteacher,
employeremployee, or traffic-police-officerdriver tend to be used to explain
what is seen and heard in a situation. This completely hides the work that goes into
the making of societal relations. It is as if the institutional relations are taken as
boxes (roles) that determine how people behave. We step into a role and then
mechanistically enact itwe are presented as cultural or psychological dopes. We
have no agency left because we are just doing what the role prescribes. The result
is a completely deterministic view of society and the daily events we can witness,
where the members of society are but cultural dopes, doing what they are meant to
do by the very fact that they are taking specific roles. Here, cultural dope refers
to the man-in-the-sociologists-society who produces the stable features of the
society by acting in compliance with preestablished and legitimate alternatives of
action that the common culture provides (Garfinkel 1967, p. 68). Equivalently,
the psychological dope is the man-in-the-psychologists-society who produces
the stable features of the society by choices among alternative courses of action
that are compelled on the grounds of psychiatric biography, conditioning history,
and the variables of mental functioning (p. 68). Classical approaches to sociolo-
gy and psychology, thereby, fail to recognize the insight that human beings not
 DATA SESSION 4 (BULLRUSH) 71

in a situation of responsibility over these M1: . . . over these M:s . . . kids are
playing in ways that were not considered safe for quite a few years.

18 I: For these educationalists, the risks are far less than the risks associated with
an activity.
19 P: If you get a kid to test themselves when hes 7 years old on a scooter or . . .
a tree . . . climbing a tree. He is not going to have to test himself when he is
17 . . . behind the wheel of a car.
20 I: And they also say it makes for better students.
21 P: The only time they get into trouble is when theyre bored . . . and they really
dont get a chance to be . . . [laughter]

For these educationalists, the risks involved with a bit of rough-and-tumble

are far less than the risk associated with an activity. For these . . . For these . . .
well, maybe there is a little more going on.15 See this ((I:)), we dont know . . .
if this were a journalist who both asks questions to people and speaks to an au-
dience then we could get different comments of talk, almost like a change of
genre.16 But this comment, For these educationalists, which educationalists?
The educationalists who run the camp or whatever the girls do and wherever the

only are subject to conditions but also make the conditions to which they are sub-
ject (Marx/Engels 1958). Ethnomethodology represents a very different approach
to society in that it recognizes the performative dimensions that produce and sim-
ultaneously exhibit in recognizable ways the structures of practical action. Here
the analyst works in the opposite direction of analyses that implicitly treat human
beings as dopes. He derives institutional relations between I:, P:, and M1: on the
basis of the verbal protocol of their relation. I: is an outsider, whereas P: and M1:
are insiders to the situation in which they find themselves. I: is asking questions;
and P: is in a relation of responsibility over the M:s. This verbal protocol is an
account of the work of doing institutional relations. It is creative. A form of sociol-
ogy, ethnomethodology literally is the science of the mundane methods that pro-
duce the regularities of a people (ethno-).
15
The analyst twice emphasizes the demonstrative pronoun, which, by the very fact
that is demonstrative, not only makes salient something in the situation but also the
relational whole of the Lebenswelt that the members to the setting currently co-
inhabit. Constructivist scholars often point out that speakers have to construct
intersubjectivity while failing to attend to the pervasive levels of intersubjectivity
that any stretch of talk presupposes and does not have to be constructed through
conscious interpretation.
16
The analyst hears a possible change in the voice, an orientation toward a differ-
ent audience than M1: and P:. These educationalists may be about the people with
whom I: has just been speaking. That is, the phrase not only makes educationalists
the subject of a statement, which comes to be specified by whatever predicate is
following and not only sets up a subject that is related to an action and possibly an
object, but it also establishes a relation of the speaker to the surrounding. This
aspect of talk lies completely underneath the blind spot of those who analyze the
content of talk, which moreover tends to be attributed to individuals and their in-
ner makeup (i.e. the constitution of their minds).
72 CHAPTER 4

girls do these rough-and-tumble games and that P: is responsible for?17 Less

than the risks associated with an activity . . . If, this is interesting because it
might give us a little more. P:, if you get the kids to test themselves when hes
seven years old on a scooter or a tree climbing a tree. He is not going to have to
test himself when he is seventeen behind the wheel of a car. That might be a
give away or just a comment, when you are young, this could be just . . . if you
are young, and you . . . a scooter or a tree . . . not going to have to test himself
when he is seventeen. This seems to be a general comment about the kind of ac-
tivities that they do when you play rough and tumble when you are young age,
and then you do not have to show off or get into risk, risky kind of behavior
when you are older. Here, seven years does not mean that the girls have to be
seven years old although this might be if we had the whole context.18 And they
also say, it makes for better students, So I:, I am not sure about I: as a person
yet, and they also say that it makes for better students. Because sometimes the I:
statement are like comments and at other times they are like questions, ques-
tions that P: and M: answer and then here, almost like confirming.19 The only
time when they get into trouble is when they're bored . . . and they dont really
get a chance to be . . . so this is, something these girls may be interested in the
rough and tumble, and it may be a camp like situation. ((Long pause.))

22 I: And yes, before you ask, the kids do go back in the class after playing bull-
rush, with a bit of mud . . . but the full on mud sliding, well, thats before
they head home . . . to the washing machine I presume.

17
This statement shows that the analyst has not settled the issue of the circum-
stance, leaving it open whether it is a camp or something else. But he specifies that
the girls are engaging in rough-and-tumble activities and P: is in a relation of re-
sponsibility over the circumstance and is present in the function of an educational-
ist.
18
The analyst clearly articulates that the content of the talk does not have to be
specific to the situation. The fact that P: talks about seven-year-old children en-
gaging in rough-and-tumble play does not have to mean that those currently en-
gaging in rough-and-tumble activities or representing such play (e.g. M1: and
M2:) have to be of that age. Careful analysis is required to unpack whether talk
indexes the situation specifically, such as when demonstrative pronouns are used,
or whether the talk is about the subject more generically. Thus, if the girls in the
situation were to be seven years old, then the statement when hes seven years old
would be a give-away; but it would be just a comment if it were not indexing
this particular situation. The analyst determines the statement to be a general
comment.
19
This statement is another example of how the analyst maintains the hypotheses
rather than quickly jumping to definitive conclusions (i.e., p(H) = 1). The uncer-
tainty about the nature of I: continues even though the analyst appears to be cer-
tain that there are comments about the setting and interactions with the members
to the setting. That is, he appears to be certain that there are two levels of talk in
the same way that there were multiple levels in chapter 3, where Vicky talked in
class and also about herself teaching in class.
 DATA SESSION 4 (BULLRUSH) 73

And yes, before you ask, . . . when they are bored. This is throwing me a loop
And yes before you ask, the kids do go back to class after playing bullrush with
a bit of mud . . . but the full on mud sliding, well, thats before they head home,
here, this is an expression that I: knows what they are up to, but the interesting
thing is up above the person was in a questioning position. As if he or she was
interviewing. Whereas here, this is a statement that sounds as if the person knew
the situation already. So, this seems to be a different register this one ((turn 20))
and this one ((turn 22)) are different registers, as if that was spoken to a differ-
ent audience. The audience that follows the camera and the other things are ori-
ented towards the people. Do you see how this feels different than this, this one
((turn 16)) and this one ((turn 22))?20 This one ((turn 22)) is almost about that
situation and the person I: must be familiar with the situation and now is talking
for someone else. Because it is not a question, the person knows that after play-
ing bullrush the kids are going back. So the person already found out from pre-
vious parts of that workperhaps even before the camera was turned onthat
person has found out and is now speaking to whoever the audience is, whoever
the recipient is of that sentence. And the person is explaining that, Yes . . . it
may be a comment about and it is going back . . . it is a special kind of regis-
ter.21 There is a lot of focus on these outdoors things that balance off school and

20
It is quite evident that at this stage the analyst does not yet articulate what the
documentary evidence is on which this feel different is based. However, this feel
would be a first take on the situation that subsequently has to be substantiated in
and through rigorous analysis. Readers need to keep in mind that this is a very
brief analysis, a first look at a document the origin of which the analyst knows
nothing about. We may get an inkling of how much there might be even in the
shortest of video clip or transcription from the experience of interaction analysis
sessions (Jordan and Henderson 1995). In such sessions, analysts gather to talk
about a videotape that one of them, the owner, has brought for the purpose of
collective analysis. Participants in such a data session have to provide evidence
for any claim they make concerning the events observed; and they may not specu-
late about phenomena that could not be tested by observing the tape (e.g. state-
ments about mental content generally are not documented). These sessions tend to
be so rich that in two hours, the analysis does not proceed further than a few
minutes into the tape. Sometimes it may take two or three hours to work out what is
being done by the participants in the recording within a few seconds. If the analyst
had more time, he would be able to trace the connections within the text to a much
greater extent and attend to more of the possible connections between the talk as
text and its connections to the context.
21
If we can detect that there are different audiences for the different parts of the
text, which in itself is flat consisting just of words, then there is something about
text that makes it reflexive, talking about itself. Talk is explicating itself rather than
just stating facts or being about something. If it explicates itself and its relation to
the setting, talk does more than express what is in the mind of speakers. It estab-
lishes the situation as much as it is subject to it. For example, if this is an inter-
view, then the participants talk in specific ways because they are in an interview
situation; but the interview situation exists only in and through that talk. As I cited
74 CHAPTER 4

academics or some kind of situation where full classes are mixed with a lot of
these kinds of activities, a different kind of school, not a regular kind of school
that we know, where everyone is so concerned with what the courts of law
might say. But may be a private school or a charter type school that is organized
differently. But the full on mud sliding, well, thats before they head home to the
washing machine I presume.22

Marx/Engels (1958) to be saying above, human beings not only are subject to con-
ditions they constitute the conditions while being subject and subjected to these.
Because we actively constitute the conditions, we cannot be the cultural and psy-
chological dopes that traditional sociology and psychology make us to be in their
causal models. Moreover, we do establish the conditions in methodical ways
there is a method to what and how we constitute society, or rather, there is a uni-
verse of methods people (ethno-) draw on to establish the order of societal settings
in accountably rational ways and by accountably rational means. By orienting to
the reflexive accountability that members to the setting provide to each other in,
through, and with their talk, this analyst arrives at highly plausible hypotheses
about the kinds of situations that have given rise to a transcription.
22
The graduate student had downloaded and transcribed a news report from One
News in New Zealand about a school that ditches rules and loses bullies. The
subtext to the video says: Schools are playing bullrush again. Has the sky fallen
in? Are the accident and emergency wards full of school kids? Matt Chisholm finds
out. In the video, a journalist (I:) interviews a principal (P:) in the yard of his
school with children in the background engaged, during their recess period, in a
variety of rough-and-tumble plays, tree climbing, and other activities that are
banned from most schools because of the risk of injury (Fig. 4.2). The end of the
transcription coincides with the end of the report when the camera switches back
into the newsroom. The final turn 22 of the transcription actually is the TV anchor
person rather than the interviewing journalist.

Fig. 4.2 The journalist (right) interviews the principal of a school that has reintroduced rough-
and-tumble play as part of the activities students may engage in. The camera switches between
showing (a) the principal with or without the journalist, with or without children playing rough in
the background and (b) students in a variety of rough-and-tumble activities, including traversing
a hurdle with a scooter, climbing trees, and playing bullrush. (Source: http://tvnz.co.nz/national-
news/school-ditches-rules-and-loses-bullies-5807957/video?vid=4946899)
 DATA SESSION 4 (BULLRUSH) 75

Readers will note that the analyst uncovered very much what type of situation
had generated the transcript: A school that had introduced school rules that al-
lowed and encouraged rough-and-tumble play was visited by a journalist who in-
terviewed the principal and students. Some of the text represents the back-and-
forth of the interviews; other text appeared as the journalists talk in voice-over
mode. The graduate student had used the same I: in an attempt to trick the instruc-
tor-analyst.
5

Data Session 4 (Mikela)

01 P: Come on inside. Does she look like a little girl?

[looking at little girl, left hand pointing at little girl, brings hand towards
herself, looks down at page]
02 P: Heres Mikelas story
LG: [gets up and enters center of carpet stands and swings arms]

Come inside, does she look like a little girl? Looking at the little girl, left hand
pointing at little girl. Brings hand down herself . . . heres Mikelas story. So
we have no response across it all. But at the moment we hear one person. We
also see the question mark here and what I suspect is that the transcriber is a
competent speaker of the language and here, there is a question whether there in
fact is a question.1 Does she look like a little girl? Grammatically this is like a
question, but the question mark also could mean a prosodic feature, I mean if
the pitch has gone up.2 Heres Mikelas story. And we see, you actually identi-
1
Although the transcription contains a question mark, it is actually attributed to
the transcriber than to the participants in the situation. In the types of analysis
witnessed here, it is important to follow the actors and how they hear something,
an indication of which we get from the reply. What and how the participants hear
is reflected in their responses, which will have to be understood in terms of a holis-
tic event that begins with attending to and hearing the preceding speaker, which
continues into and through the speaking, and ends when everything has been said
and done by the individual whose turn it has been.
2
We hear some phrases because they have the grammatical structure of questions,
which could lead the transcriber to place a question mark independent of the into-
nation. But there are other situations where the grammatical structure is not that
of a question but the phrase nevertheless is heard as such. For example, a speaker
pointing to a chair and articulating chair with rising intonation may be heard
as asking whether the object pointed to is in fact a chair. Similarly, a phrase
grammatically structured as a question may not in fact function as such but consti-
tute an evaluation and invitation to engage in a repair. Thus, if a child were to
78 CHAPTER 5

fied the little girl, who gets up and enters center of carpet, stands up and swings
her arms. I might have to look at the actual video. So you have come inside,
Does she look like a little girl? and we have a little girl do something and we
see it here, enters center of carpet and swings arms. We might see this ((de-
scription of action in turn 01)) as an {invitation | acceptance} pair. We can see
that as an invitation to enter the circle and an actual entering of the circle. Then
here is the kind of story; something is set up. If the sequence is this, then we
have arrival, the first pair, and this ((action in turn 01, talk in turn 02)) would be
read as the next pair. So having entered, and beginning or introducing, Heres
Mikelas story is another sequence, apparently this situation is indicating read-
iness for the next thing, which is then announced here.

02 P: Heres Mikelas story

LG: [gets up and enters center of carpet stands and swings arms]
03 P: A little girl.
LG: [holds fingers]

So we have Heres Mikelas story, A little girl, holds fingers. We do not

know how she holds [fingers]. Does she look like a little girl? We might read
this now as an adult, it says little girl, but here it might be the adult commenting
upon what a very little girl might do. I dont know, my hypothesis is that it is
not a five-year-old or a six-year-old little girl, it might be a younger little girl.
That might be a working hypothesis, because I do not have any other infor-
mation. So I do not know. Heres Mikelas story. A little girl? Does she look
like a little girl? That picks up Does she look like a little girl? So the question is
this: What is the function of this?3 Of course, its uttered again, so it is different,
it is arguing against that this is already prepared here. A little girl. Now the
question is: Does this describe the little girl Mikela, who has come in? Or is it
something about Mikelas story? A little girl, Does she look like a little girl?
Now why would this little girl come here? Here the period indicates to me that
the transcriber heard this a statement or a description. A little girl. Does she look
like a little girl?4 It may be asking for confirmation.

point to a table and say chair, a parent might say in the next turn Is this a
chair?, which has the grammatical structure of a question. The phrase has func-
tioned as an evaluation and invitation for repair, when the child replies, Oh no, it
is a table.
3
The question concerns the function of the talk rather than its content. There are
situations where the content of a phrase or word is completely irrelevant to the
situation. A case in point is a brief text from the Diary of a Writer (Dostoevsky
1994), in which six drunken workmen use the same expletive six times in sequence.
Dostoevsky, who overhears the workmen, shows how each articulation expresses
an evaluation of previous talk rather than denoting a specific fact. The narrative
was taken up in theoretical considerations of language-in-use to underscore that in
any specific situation the dictionary sense is irrelevant to what it being done with
the sound-words produced (Voloinov 1930; Vygotskij 2005).
4
Readers note that already in his reading, the instructor-analyst is pairing speak-
ing turns or actions and speaking. It is clearly in the attempt to read the situation
 DATA SESSION 4 (MIKELA) 79

03 P: A little girl.
LG: [holds fingers]
04 P: Does she look like a little girl?
05 C: Yeah
LG: [turns head towards talking child]

Turns head towards talking child. So we need to have Does she look like a
little girl? Yeah. So we have here a {question | response} pair. But we still do
not know in this situation if this is about some little girl or if it is about Mikela,
the little girl. Here we have Heres Mikelas story. So for me, the reader, it is
not yet clear whether there is something about the little girl and Mikelas an-
other little girl or if it is little girl in Mikelas story, and Mikela and the little
girl are the same.5 Anyway, the teacher and the student produce a {question | re-
sponse} pair. {Does she look like the little girl? | Yeah}. So whoever she is,
whether or not it is some little girl or if it is Mikela. it will be visible as such in
the situation.6 The people present will act in ways that take this visibility for

through the lens of joint action. Such joint action cannot be reduced to individual
action. Anything individuals do can be understood only as integral part of joint
(social) action. It is identifiable as such only because the joint action is already
recognized as a whole. Joint action is not the result of the composition of elemen-
tary (individual) actions. There would be no necessity of an individual action to be
part of a joint action. What is this work that the participants here do collectively,
together? The idea of joint action does not come easy to most people, for we are so
used to thinking in terms of individuals. But some popular expressions point us to,
and are reflexive of, irreducibly joint action. Thus, for example, hand clapping
requires two hands. It is not one hand clapping and another one clapping
clapping is the joint, coordinated action of two hands. It takes two to tango is a
popular expression that highlights the irreducibility of a situation to individuals, as
it takes two hands to clap. At the end of the 19th century, . Durkheim suggested
that there are social facts that are of the same order as the facts of natural scienc-
es: The first rule and the most fundamental is to consider social facts like things
(Durkheim 1919, p. 20). Formal analytic studiesquantitative and qualitative
alikepresuppose the objectivity of social facts, even though they assume that the-
se facts are often hidden from the people and therefore require special methods to
be uncovered. Ethnomethodology, in contrast, exhibits the work we do in our eve-
ryday world to exhibit the structures of practical actions such that they become
accessible to participants in the setting in their factuality and, therefore, come to
be objective social facts (Garfinkel 2002). Precisely because social facts can be
and are exhibited in the public forum and, therefore, become witnessable, our ana-
lyst can reconstruct the institutional relations that both structure and are struc-
tured by the members to the setting.
5
The complexity of the talk arises from the fact that it could be about the girl in the
story or about Mikela, the author of the story. A phrase such as Does she look
like a little girl may be about Mikela the actor playing a little girlso that
Mikela could be an older girl as wellor about Mikela, the person.
6
The instructor-analyst is reading the transcription through the lens of a possible
situation where what participants say is the most rational thing to be said. Partici-
80 CHAPTER 5

self-evident. There is something in the situation that allows this pair Does she
look like a little girl? and Yeah to be a question and response pair.7 So we can
expect there to be something in this question. This question asks something that
is recognizable by all the participants in this situation, some little girl.8 Here we
have the turns head towards talking child, which is probably a little after the
yeah.

06 P: Okay. Would you be a little boy?

[brings left hand down towards lg2 and looks down at lg2]
LG2: [looks up at P and nods]
07 P: Okay . . . A little boy
[gestures hand towards center of carpet]
LG2: [gets up] [stands next to LG, hands by side]

Then we get here, Would you be a little boy? Now, Would you be a little
boy?, we here see something that the transcription says that it is a question, but
what we do not see in the transcript is the orientation of the teacher. That might
actually change how we read a given situation. So now the little girl gets up,
stands next to LG, beside the boy. Now the question for me is: Is this a question
and if so, where is the response? It is only a question if there is a response. So
here if this one ((action in turn 06)) is not simultaneous but follows a little bit
after, this could be actually read or seen as a {question | response} pair. {Would
you be a little boy? | Gets up, stands next to LG, hands by side of body?} So this
might be, again to be confirmed later on, a {question | response} pair. Would
you be the little boy?, the invitation to another girl to play the role of the little
boy. Then this would be a {question | answer} pair. Just by looking at this here,

pants are heard to be speaking for the purpose of that situation and to make salient
what they do for the specific needs of that situation and they will leave unsaid what
goes without saying, again in that situation.
7
Clearly the analyst does not know what the situation is. But taking the talk as an
account that the participants produce for each other provides clues about what the
situation is. That is, the talk is taken as having more than just literal content. We
can think about this in another way: The participants in the turn-taking are solving
a problem, through acting and talking; and the analysts task is to figure out what
problem the people are solving. Talk therefore also constitutes the context in which
the talk has to be understood: text also constitutes context (Roth 2010). And, most
importantly, it organizes the situation as a whole and the societal relation that
binds the participating subjects together into an intelligible unit. For activity theo-
rists, the unit would be one characteristic of society, a societal unit: activity under-
stood as a formation that produces something of value to society as a whole, which
also may be consumption, the reverse side of production (Marx/Engels 1958).
8
The analyst here uses the reflexive nature of language to hypothesize that the
question does not just ask about something but makes available that this something
is accessible in the setting. That is, this one phrase is doing two things simultane-
ously, asking a question about something and revealing that this something that
the question is about actually can be found in the situation.
 DATA SESSION 4 (MIKELA) 81

I do not have confirmation. But the videotape may provide me with further
clues, gestures left hand towards the carpet, Would you be the little house?

07 P: Okay . . . A little boy

[gestures hand towards center of carpet]
LG2: [gets up] [stands next to LG, hands by side]
08 P: Would you be the little house?
[looks down at lg3, left hand pointing towards lg3, nods]
LG: [walks towards LG2 and stands beside LG2]
LG3: [stands up, body swinging, hand to mouth]
LG2: [grabs dress, looks at lg3, walks away from lg3 behind lg to stand beside lg]

Here we have a sequence. There is something going on and they appear to be

knowing what is going on.9 But I do not yet have a clue. Here we have all little
girls, there is the question Will you be a little boy? This does not make sense:
Would you be a little girl? Unless: I can think of the question as an invitation to
a game where someone of those present will play the role of a little boy.10 This
would have confirmation here, but I need to find out a little bit more. What I am
trying to articulate is a form of analysis. So theres a lot of stuff that I do not
know. If I filled in detail [arbitrarily], then I may not have sufficient evidence to
confirm it. What I am trying to do is this: say only what is possible to be said: It
could be. I can generate hypotheses: this possibly is the case. I am trying to state
the next hypothesis that I could pursue. I do not know right now what is going
on. I am something like Sherlock Holmes or Miss Marple. Even if I know some-
thing, there is always something remaining open. You cannot just fill in these
open spots. What you are trying to do is this: You want to articulate those things
for which you have clear evidence. What I have clear evidence of is that which
is actually available.11 See, I have not talked about power, for example, between
teacher and children.12 I have talked about looking for pairs, {question | re-

9
It is not that participants in a situation merely appear to know what they are par-
ticipating in, at least, in most everyday situations where we find ourselves, which
we orient to, which calls upon us to act in specific ways, and which, in and through
our actions we both produce and reproduce.
10
We can hear this phrase as a hypothesis or as an articulation of a indeterminate
sense. In either situation, substantive analysis has to follow to provide documen-
tary evidence. In the case of the hypothesis, the evidence would function as data to
delimit the probability of the hypothesis p(H|data). In the case of the (vague) sense,
the analysis needs to work out how what is sensed can and does manifest itself in a
concrete way so that it is expressed and can be noticed not only by this analyst but
by analysts in general.
11
Clear evidence is that which Durkheim refers to social facts that are objec-
tively given and constitute the objects of sociology. This evidence is given in the
ethnomethodological sense, that is, as part of the ways in which human beings not
only produce society but also make this production and its results visible and ob-
jectively available to all members to the setting.
12
The analyst here (implicitly) discourages the use of power as an explanatory
category. If the term is used to denote the relative institutional positions of two or
82 CHAPTER 5

sponse} pairs. I am looking for things for which there is evidence that I can lit-
erally point to. I am not trying to fill in gaps by coming up with explanations for
which I do not have evidence. If I do not have evidence I still can say well it ap-
pears to me, I have a hypothesis. Maybe the teacher is inviting to a play that
they already know approximately. But you cannot say what the teacher thinks or
what the child thinks, because we have no evidence for it at this point in time.
All we have as evidence is what they do and say, what they exhibit to us for
us.13 What is important is to understand the flow of the lesson.14 So if [the ex-

more members to the setting, then it is because of certain relations. Relations re-
quire all their parts such that power-over is not something one person can have,
like the money one has in ones pocket or purse. It is a relation of ruling (e.g. Smith
1990a) that those who are determined to be ruling and those who are determined
as being ruled over participate in and produce.
13
It is important to retain the to and for us, because in any social situation com-
munication has intentional dimensions, where we articulate something directly
addressing another. But there are also other aspects that we do not articulate in-
tentionally but that are heard and seen as context that specifies what we are heard
to be really saying. Thus, for example, gesticulations accompany talk but are
not produced intentionally, though some gestures, such as the infamous middle
finger sticking out from an otherwise fist-forming hand, perhaps together with an
abruptly stopping upward movement of the whole arm, are in fact intentionally
produced. In gesture studies, these kinds of gestures are referred to as emblems.
Other examples are the middle and index fingers shaping a V for victory.
14
The significance of the flow of the lesson lies in the fact that this flow is not
generated from the outside and not by mysterious forces and causes. Instead, the
lesson unfolds in the manner it does as a result of the toing and froing tying to-
gether the speakers and participants. The flow is the result of how others hear
what we say and see what we do rather than what we intend them to hear and see.
The flow is produced endogenously. Thus, in the following exchange, the speaker
in turn 01 might be intending to ask a question but is heard to be nagging.

01 S: Did you clean the dishes?

02 H: Why do you always have to be nagging. Get off my case.

The participants now are confronted with a new aspect of the ever-evolving
situation. What S: has been said to have donein other words, what H: has formu-
lated S: to have donemay need to be addressed. For example, S: might say I
only wanted to see whether we are ready to leave. In this, S: provides an alterna-
tive formulation of what she has done in turn 01 in response to turn 02. That is, she
has indeed dealt with the new situation by re-formulating the nature of the intend-
ed work. The ordered and ordering turn sequences are at the heart of the flow of
the situation rather than what some analyst might suggest to be the contents of the
minds of the participants. What the two participants make available to each other
is all they and the analyst have. Neither participant is in control over the give-and-
take relation, which is a reason why situations can and do get out of hand: these
are never in the hands of individuals in the first place.
 DATA SESSION 4 (MIKELA) 83

hibited] is relevant to the flow of the situation, then someone else will comment
on, build on, use it somehow in their reaction. If not, then it is not relevant to the
flow. If there is evidence, for example, if the teacher said, I know you dont
want it, then you would have evidence that the teacher has heard something
that made her comment a rational response. If it is just you interpreting, then
there is obviously a problem: you bring in outside stuff irrelevant to the actors,
and then the readers will say, Oh, youre just introducing yourself, your inter-
pretation. What I am trying to say is this: Work with the materials that you
have, and, where you have evidence, use it. Really use the evidence. Here, the
videotape might provide us with much more evidence than we need for seeing
what is going on.

08 P: Would you be the little house?

[looks down at lg3, left hand pointing towards lg3, nods]
LG: [walks towards LG2 and stands beside LG2]
LG3: [stands up, body swinging, hand to mouth]
LG2: [grabs dress, looks at lg3, walks away from lg3 behind lg to stand beside lg]

So Would you be the little house? Now it is not Would you be a little house?
It is the little house. This is a possibility: here is Does she look like a little girl?
Would you be a little boy? A And here it is the, and that might actually turn out
to be significant in the sense that a boy and a girl, there is the indefinite article.
But here ((the house)) it is a definite article. And if the teacherI assume P: to
be suchuses the definite article whatever is to come may already be prefig-
ured or known to the participant namely there has to be, there is, there has to be
some house involved.15
To track back, when we record interviews and we transcribe them, we al-
ready know them. Then we tend to read the beginning of the event through its
end, its outcomes, which is the lens. What I am trying to do right now is what
we call first-time-through. What I am trying to do is to read these lines as the
participants experienced the situation not knowing what happened at the very
end. I do so because it could be that the lesson actually did not work. See the
teacher could actually have a lesson that did not work according to her method
that we are witnessing. But here, she does not know [whether the lesson will
have been a success]. She does not know it; the kids do not know it.16 As the re-

15
Here the instructor analyst begins to tease out the possible relevance of the defi-
nite and indefinite article. The definite article the is an indication that partici-
pants already know a house to be part of what is coming, whereas the indefinite
article a might be an index of the unknown status of what is coming.
16
This tendency to use after-the-fact evidence to explain social phenomena is per-
vasive and insidious. For example, I have seen time and again how researchers
take turns such as the following and already presuppose their effect/s. They might
quote innocently write that the teacher prefaces Jonnys story to come.

01 Teacher: Jonny, why dont you tell us what experiment you have done at home
and what you found out.
84 CHAPTER 5

searcher, and that is the situation that I am in, I do not over-interpret or wrongly
interpret that because I am not using what I will find out only later to look at
and understand what is happening right now. What I am trying to do is under-
stand the interview or hear this situation in a first-time-through manner. And at
this moment the teacher says, Would you be the little house? She may, unless
she reads from a script, she probably does not know the exact words that she
will speak just a few turns down here. If we read this part knowing where they
ended up, then we sort of falsify the way we understand that situation because
we no longer understand it like the participants do. They do not know whether it
will be a success or it will be achieving its purposes, especially the teacher. She
does not know whether she will succeed in what she might have put down in her
lesson plan. Given that that is probably a demonstration lesson, she probably
succeeds in whatever she wants to succeed.17 So what I want to do is get the
sense for what was it for these people when they participated, when they were in
this situation.

08 P: Would you be the little house?

[looks down at lg3, left hand pointing towards lg3, nods]
LG: [walks towards LG2 and stands beside LG2]
LG3: [stands up, body swinging, hand to mouth]
LG2: [grabs dress, looks at lg3, walks away from lg3 behind lg to stand beside lg]
09 P: Okay. A little house. How should Mikela since its your story how should
the house be how should she make herself?

So, Would you be the little house? Here, that would strike my interest: the
house and why not a house? So there seems to be something definite about what

However, at the time the teacher speaks, nobody could have known whether
Jonny actually will have told a story. In fact, Jonny might reply, I feel sick. Can
someone else tell what they did? In this case, the teacher might select another
one. Importantly, the turn will not have prefaced Jonnys story. It is only when
Jonny actually will have told a story of what he has done at home that turn 01 will
have been the prefacing of this story. As Nietzsche (1954) writes, causes are at-
tributed after the fact; they only can be attributed after the fact once the effects are
known. Nietzsche thereby explodes causeeffect reasoning as a fantasy. This at-
tribution is not just fantasy but has deleterious effects on social analysis as well as
on judicial cases, where people and parties frequently come to be blamed a poste-
riori even though they were never in a situation to anticipate (all) the effects that
their action would have. Heidegger, across much of his later work, criticizes tech-
nological thinking not because he is against technology but because of the underly-
ing, unrealistic metaphysics of causeeffect relation (calculation) (e.g. Heidegger
2006).
17
The instructor-analyst refers to the fact that the video was published online,
which can be read as an indication that whatever method was to be illustrated in
the video was in fact attained. But at the time when the event took place, nobody
could know whether what was in the process of happening would in fact be looked
back at some time in the future and considered a successful teaching demonstra-
tion.
 DATA SESSION 4 (MIKELA) 85

is to be and about what is to come. Walks towards little girl two and stands be-
side little girl two. Again if this is read like this, I might, and that would have to
be confirmed looking at the video, the teacher might have oriented towards the
little girl and invited her: Would you be the little house?18 Then, but then there is
actually a conflict for me, because she was asked, Does she look like a little
girl? So at the moment I, as the onlooker, as a stranger in this classroom or in
this group, I do not even know if it is a classroom, I do not even know what is to
come. Especially because I do not see where they are going.19 But if the event is
like this, then we have a {question | response} pair. An invitation or an {invita-
tion | acceptance} pair: the teacher invites, the student accepts the invitation to
be the little house, walks towards and stands beside, looks down at little girl
three, left hand pointing towards, nods. But at this point I am sort of confused
by my understanding of this situation because I thought it was dealing with a lit-
tle girl and a little boy and a house but the little girl now seems to be in the
house. She also looks pointing towards LG and now it is . . . it may actually be .
. . it may be that this ((2nd to 4th action in turn 08)) is a response to that ((talk
in turn 08)) and that this ((first of the four action lines)) assigns the little girl
three to the house and this little girl one and little girl two is the boy. So we will
have a pairing and the little house is the third girl.

LG2: [grabs dress, looks at lg3, walks away from lg3 behind lg to stand beside lg]
09 P: Okay. A little house. How should Mikela since its your story how should
the house be how should she make herself?

Grabs dress, looks towards LG three and walks, Stands up, body swinging
from . . . swinging, stands up, body swinging from side to side, hand up to
mouth, okay a little house. So whatever has happened, this, we can see this as an
event and the comment describes what has happened here, so whatever this con-
figuration tells us is being described.20 Okay, a little house. Then a little house

18
The you can be heard as specifically addressing a person in the setting, or it
might be used synonymous with one and thereby express generality. How it is
heard in the situation has to be revealed in the analysis of subsequent turns.
19
If the social world were as it is often depicted, the result of causeeffect rela-
tions, then it would be possible in principle to determine the future of an event.
Knowing what kind of event he is observing, the analyst should be able to say
where they are going. As a researcher and as a department head in a high
school, I frequently found myself in situations where I could not figure out from a
lesson where it was to go. So even though teachers apparently presupposed the
visibility of their intentions in the lessons and the visibility of some concept, phe-
nomenon, or focus, it was not apparent to me even though I was in many cases
more knowledgeable about the subject taught (physics, mathematics) than the
teacher I was observing.
20
In this situation, the instructor-analyst develops the rationality from the inside of
the situation. Although he does not understand right now, the approach is to locate
himself such that what he hears is the most rational thing to say in some situation.
In what kind of situation would the talk heard be the most rational thing to say,
86 CHAPTER 5

has been formed in this particular situation, and those present understand that
the little house has been formed. So it is both a description of what happened,
and a description of what was supposed to be: Would you be a little house? The-
se events have produced the little house. How should Mikela since its your
story how should the house be? How should she make herself? Okay here we
have more: it is one little house. How should the house be? How should she
make herself? This is a question. The question is directed towards the children.
We have a little house and we have a little boy and a little girl. It is Mikelas
story. The teacher somehow knows Mikelas story already, and what is to
come.21 So she appears to be scaffolding or whatever Mikelas story to be en-
acted in that particular situation. This is a hypothesis.22

09 P: Okay. A little house. How should Mikela since its your story how should
the house be how should she make herself?
10 LG: [looks down, then stretches both arms out to the side, then spreads legs
apart]
LG2: [hops once]

Lets get this right. Stretches both arms, should that be? How should
Mikelas, Mikela, since its your story, how should be the house be? How
should you . . . looks down, then stretches well this, if this is a {question | re-
sponse} pair, then this little girl tells us, in her response, how the little house
should be. So whatever looks down, stretches both arms out to the side and
spreads legs apart. Hops once. So this is a question, it is a multi-layered ques-
tion that is going on because in the hypothesis. . . . What I am building is a hy-
pothesis that can be disconfirmed as I go along, right?23 I am not saying that it is

assuming access to situational particulars that the analyst does not have to in the
absence of the video.
21
If the analyst can see that the teacher already knows the story, already knows
what projectively stated as something to come, then this is so because the talk ex-
hibits it as something for anyone else as well. It is therefore objectively available
for anyone who cares looking. As such, the story-assomething-already-known-
even-before-it-is-played-out-for-the-present-audience, is a social fact that can be
substantiated by evidence.
22
The hypothesis concerns the statement that she appears to be scaffolding the
telling of Mikelas story so that it comes to be enacted for anyone to see by means
of a performance. It is worthwhile to attend to the distinction between what actual-
ly is available, which is the knowledge of and about the story, and the hypothesized
situation that could have led to the manifestation of that knowledge. These two
aspects are part of the documentary method of interpretation described in chapter
8.
23
Building of a hypothesis is a pretty accurate statement of a process by means
of which the probabilities of different hypotheses Hi, i.e. p(Hi|data), are updated
continuously given what presents itself as data until one emerges from the process
that is (much) more likely than others. The probability of a hypothesis before the
inclusion of some data is called the prior [probability] and the probability follow-
ing the consideration is referred to as the posterior [probability]. In the case of the
 DATA SESSION 4 (MIKELA) 87

like this. When I tell the story then I can be more definitive because I have all
the evidence. But right now I am open in this respect. So, as a response to how
should the house be? we get this ((transcription)), which is actually not suffi-
cient evidence that girl one is the little house. Because we have the little girl that
was being invited and little girl three, she stands up swinging side to side and
then walks away from behind little girl to stand beside little girl. So okay we
have the little pair, and we have a house. It is Mikela; she is both. So far we
have sufficient evidence, we have no contrary evidence for the hypothesis that
Mikela had a little story and in that story she is herself, a little girl. There is a
little boy involved and there is this third person playing a house. And so we can,
we may anticipatethis is another hypothesissomething is going to happen
where there is a little boy and a little girl and something related to the house.

10 LG: [looks down, then stretches both arms out to the side, then spreads legs
apart]
LG2: [hops once]
11 P: Alright would you do that? Like that.
LG3: [swings side to side puts arms partially out to the side]

So then what next, Alright. Would you do that? Like that? So the question, so
whatever this event is in here. Something happened in here, something is a
commentary or a questioning or an asking for a reflection perhaps or Alright.
Would you do that? Whoever that question asks, if it were asked of that, then
the question would be Would you, the little boy Hops in the way the little boy
hops, or it might be that the question is to little girl three playing the house, ask-
ing Would you do what Mikela just, showed us the little house has to do?,
like that, in this very way that Mikela has shown.24

analyses provided here, the probabilities are not actually calculated and numeri-
cal. Instead, the reasoning involves increasing and decreasing the relative proba-
bilities given only in some more or less fuzzy way. This could be modeled mathe-
matically using the formalism of fuzzy logic. Fuzzy logic is a way of mapping fuzzy
observations and fuzzy criteria onto categories. For example, I was able to show
how examiners arrived at ratings for the performance of pilotse.g., between 1
(below standard) and 5 (very good)based on observations about their communi-
cative performances such as leaning heavily on the other pilot or pretty open
discussion (Roth and Mavin 2014).
24
As already observed above, there is no unique way in the possible functioning of
a phrase. Would you do that? may be an invitation, especially when appearing
with a please. It could be part of a negative evaluation, if the function of the
query would be to ask another child whether it would engage in the action thus
marked. It may also an invitation to a child to produce an alternative, especially if
heard as Would you do that in this way? What matters is not a sentence and its
semantic properties, which might be specifiable by means of dictionary sense, but
the place a phrase takes in an ongoing language-game. This place, the tight con-
nection between language game as a whole and the specific language-in-use is not
attended to in much of social research, concerned as it is, among others, with the
beliefs, knowledge, conceptions, or attitudes of individuals. What is social about
88 CHAPTER 5

11 P: Alright would you do that? Like that.

LG3: [swings side to side puts arms partially out to the side]
12 P: And do you go inside the house?
[points at lg]
LG: [nods]

Swings side to side, puts arms out to the side, And do you go inside the
house? And here we have a nod. Now here we see a possible {question | re-
sponse} pair. {Do you go inside the house? | Nods}. So if she goes inside the
house, she cannot be the house. So my question that I have above theremy
uncertainty, is she the house or is she the little girl?is resolved here: No she is
the little girl and she is asked Do you go inside the house?, because it makes no
sense to ask a house if you go inside the house. So it is the question to a little
girl, Would you go inside the little house?
So what I am doing is I am working step by step through the analysis and I
describe what happens, and I use all the evidence that I have. If there is a cough,
if there is a face, and if that face is related to something else, for example, you
might think that the teacher is asking, You want to play the house? And the
girl will go uh, you know goes like this ((pulls face)), then there is some con-
sequence something has to, is happening next. The next is responding to this
first pair, so it might be no I dont want, not today. Why not the other? The
girl makes a face and then the teacher says, Oh well, if you dont feel like it to-
day, then someone [else] might play, we observe a consequence. For you to be
able to say what is going on, collectively it has to matter within this unfolding
event. Otherwise talk about what the teacher does and makes available to these
children. It could be an orientation.
As I said, I only hypothesize, I do not have the video. But somehow, if she
says, Alright, would you do that? while he snapped like that. My hypothesis is
that Would you do that? Like that? Do the house in that way? is sort of a possi-
ble way to see what happened.25 But she also is oriented toward someone. While

talk and relations is completely stripped from the talk: nothing but some abstracted
sense remains.
25
The analyst is not elaborating on this situation, which he might well do in an
extended analysis. There is something in the situation that makes it rational to ask
whether what is being done to depict a house is in accordance with some referent
that is here not made explicit. For example, if it is Mikelas story, she might have
had the house expressed in a particular way. The teacher now is asking whether
the girl playing the house is actually doing what Mikela wants the house to do, in
her story. That is, rather than saying This does not make sense, the analyst asks
the question, In which type of situation would it make sense to say precisely this,
not more and not less? In our everyday lives, there are many situations where we
complain about the non-intelligibility of what others say or do: This makes no
sense! Perhaps we say, This is illogical or irrational. Yet we could make a
different assumption and take everything people say and do as rational in some
situation or from some point of view. Rather than describing others to be irration-
al, or to have some kind of cognitive or social defect, we would be trying to figure
 DATA SESSION 4 (MIKELA) 89

analyzing the video, that orienting to someone can be used as part of the evi-
dence that she oriented, she addressed a particular person and did not just speak
generally. Would you do that like that?, she might be oriented to the third girl
then this could be the response. If she is oriented to the first and does not even
see the third girl, then you cannot say it is a question and response pair. So the
swinging may be completely irrelevant to the question. It may not be. But unless
you have further evidence you cannot say.
We can say she makes a face, but using the word grimace might be
judgmental. The question is not whether or not it is a face or a grimace. The
question is: Does it have an effect? Does this action have an effect on this les-
son? If the teacher just says to the next girl Do you want to play the house?
then we know that the teacher has acted upon this face.26 Or she might say,
Come on, its just playing a house. But the teacher somehow has to pick up on
this grimace. If she does not and the girl gets up and then does it, then you can
simply see her action as an {invitation | response/acceptance} pair. It might be
grudgingly, and it might come up later, if you have further evidence that there
was a grudge. The face may indicate something else. The kid might not feel
well. If the teacher said, Aw come on, do it, play the house, and the girl says,
Well I actually dont feel so well or My stomach aches or something like
that, then you would say it is not a grimace or grudge. The face would be an in-
dication that she does not feel well. What you have seen as a grudging look may
actually be an expression of not feeling well. I would therefore look for more

out within which horizon the actions and statements make sense. Once we have
identified one or more situations, we have arrived at hypotheses about the context
within which the text observed does indeed make sense. It provides us with an un-
derstanding of the worlds of others.
At this point, it is useful to bring into the discussion the notion of language-
game (Wittgenstein 1997). The term denotes a whole consisting of language and
the activities with which it is interwoven (p. 5). This then allows us to ask of any
statement: Within which language game does it make sense? The analysts task
then becomes one of working up a description of a language game as a whole in
which specific statements are integral parts. The analyst does so based on the con-
crete pieces of documentary evidence, actions and words, that not only imply the
activity as a context but without which the activity could not be, at least could not
be in this (actually observed) way.
26
The analyst describes a strategy that moves the work of interpretation from the
analyst to the participants. Descriptions such as making a face or grimacing
are culturally (value) laden and may be highly idiosyncratic. All research where
the weight of the interpretation is taken on by the researcherwhether it is an in-
terpretation by a qualitatively working individuals or by a coder in a quantitative
studyis subject to mis-taking the in situ function of the facial movement. What
matters to the unfolding of the observed event is how the facial expression, if there
had been one discovered by other participants, was taken up and thereby mediated
the course of the events. The facial may be of the same type as a grooming gesture,
having no other reason than some physiological one (e.g., being the response to an
itch).
90 CHAPTER 5

evidence that the look was actually treated as a grudging look by participants ra-
ther than just expressing something else.
I am just trying to think of this in the context of an interview. 27 Like if you
are conducting a direct interview, one-to-one with your subject, then most of the
things you are going to put down is going to be text but are you going to take in-
to account everything? I remember when I was going through some videos,
there was a girl that they were interviewing for wrestling and she was like
((gasps)) and looking up as if she was trying to think of a response. Do you in-
clude those kinds of things? These may not actually matter at all to your analy-
sis. It depends on your question. What I am not going to do is attribute inten-
tions to her unless I have them. If she said, I want to ask you a question or I
want you to play with us or something then she articulates a want. If she does
not, then we do not know what is going on in her mind. She might feel terrible
and articulate for herself, Oh I wish this lesson was over. But we do not know
this. So what I am trying to say is this: work with the evidence! And here I try to
work just with the evidence that I have.
If the timing of talk and action is irrelevant to your analysis then you do not
need to include it in the transcription. If it is relevant, then you want to show it
because if your child moves immediately that is an indication that there is a
shared anticipation of what is to come. Even though the teacher seems to pro-
vide the main narrative, the children act as if they already knew what was com-
ing.28 So it is not the teacher telling what is to happen, the children, they partici-

27
A distinction is made as to the relevance of transcribing what the transcriber
perceives. Thus, if the event is an interview where the researchers is interested
only in the what of the Said, then there tends to be little need to transcribe repeti-
tions, coughs, false starts, mumbles, pauses, and the likes. Whereas we may view
such research and the way in which it produces transcription as a legitimate en-
deavor, most researchers do not make explicitto themselves or the readers
what the underlying assumptions are that ascribe some text to a specific individu-
al. For example, it assumes that we produce texts independently of the audience;
and it assumes that what we say is constrained by the language or the recipient.
That is, there is a gross reduction in play that many researchers are unaware of
when doing such research (Edwards and Potter 1992). This approach, for exam-
ple, overlooks that language speaks (Heidegger 1985) and that we find ourselves
speaking a language rather than designing or inventing the language that will be
speaking (Rorty 1989).
28
The traditional approach to the relation between a plan or instruction and the
action that implements or realizes the plan is taken in a deterministic way. Thus,
for example, if a cookbook recipe reads, knead until smooth and elastic, what-
ever the (novice) baker does is thought in terms of causeeffect relation. However,
this view of the relation between plan/instruction and situated action is problemat-
ic because we know what we have done only after having done it (Suchman 2007).
Even the most highly trained scientists, following 30 years of doing a particular
dissection, may find after a fruitless day of work that their dissection in the morn-
ing had not done what they intended it to do (Roth 2009). Thus, plans/instructions
are appropriate accounts of the actions when the results are successful; and they
 DATA SESSION 4 (MIKELA) 91

pate in acting. Then I would understand the teachers text more like a commen-
tary on events that are at least in part already anticipated by the children in col-
laborating and producing it. The teacher collaborates in the sense that she pro-
duces a narrative that goes with what the kids do. If the teacher did not, if the
teachers text was unnecessary, she would simply say, Go play Mikelas sto-
ry.
The fact that she speaks allows us to assume that it is for a purpose in that
situation. Nobody does or says something that is unnecessary.29 In fact it would
be seen as strange if you asked me a question, and each time you ask a question
I said, Oh, are you asking a question? So youre asking a question right now?
You would likely think: This guy is strange. Why would I do that? I am ask-
ing, because it is not necessary. So if the teacher talks, then it is for a purpose. It
is not just willy-nilly or at random.
You do provide evidence that the children act without pause; it is almost
simultaneously. Then it is important to show this near simultaneity. You do so
because it tells us that the children anticipate and may know what will happen.30
Simultaneity is very important for understanding this social happening. It is not
just some willy-nilly event; and the children do not just act because they are or-
dered by the teacher to act in a particular way. At the same time the teacher ap-
pears to provide this narrative because without it something would be missing.
So in this narrative we also find the reason for that teaching to go on. So I am
sure if I was to look down further in the transcriptioneven though I do not
know the details of the teaching methodI might find sufficient evidence to tell

are inappropriate accounts when the actions have failed to achieve what they were
intended to achieve. At best, therefore, plans/instructions are accounts against
which an actual course of action is judged after the fact. Any one statement, such
as She is making the house may serve as a prospective organization of what is
to come, as a concurrent account of what is happening, or as an a posteriori ac-
count of what has been done.
29
The very act of speaking is taken to be important because of the default situation
according to which talk is unnecessary when something goes without saying.
Moreover, what is said is relevant because it figures against the virtually infinite
realm of what could have been said. There is a {figure | ground} relation between
what is said and what remains unsaid.
30
If the teachers description is concurrent with the childrens actions, then we
hear the talk as an account of what is currently unfolding rather than as an in-
struction for what is to come, an account that will serve as referent in establishing,
after the fact, whether what has actually been done is conform with what was pro-
jected to come. A concurrent account may serve to assist children to discover in
their actions something of theoretical relevance, to associate what they are doing
with something that is normally done in the domain with which they are only in the
process of becoming familiar. This concurrent account would then be an aspect of
teaching a language-game, using relevant language in the course of an unfold-
ing activity and thereby tying it to the specifics of the activity. It is in this way,
then, that the knowing a language and knowing ones way around the world be-
come indistinguishable (Davidson 1986).
92 CHAPTER 5

you about what that teaching is about. See I could ask, Why this narrative?
Children do not know. In fact they do know. So they anticipate what is coming.
But the narrative has a purpose. She might help the children to tell stories in a
particular order and act stories in a particular way, or to tell story by acting. Her
text helps the children to organize their enacted story. I would need some more
evidence. But I bet I could come up with some hypothesis that I could not dis-
confirm. There might be multiple hypotheses, as you saw. There was a point
where I could not say whether child one was the house or little girl. So I have a
hypothesis. Theres a possibility that she might she might actually be the house.
Until this hypothesis . . . was confirmed as I went further down.31

31
The graduate student who had brought the transcription to the class subsequent-
ly described its origin in this way:
The data analyzed in this paper was from a video which filmed Vivian Pa-
leys storytelling and story acting in a preschool classroom. The video used
was one of three created by The Child Care Collection, which is part of Ball
State University in Muncie, Indiana. Vivian Paley was invited to visit a Pre-
school in Muncie, Indiana, where she was filmed for two days working with
the children. The footage from the two days was used to create three films;
the film used in this paper was The Boy who Could Tell Stories. In the video,
footage of Paley working with the children, scribing their stories, and the
children acting out their stories using her method of storytelling story acting
was seen. The video was a combination of the footage with the children, and
interviews with Paley where she comments on her methods and the ways in
which they are good for children. (B. Pearson, April 2010)

Fig. 5.1 Photograph displaying what the graduate student saw and described: Kindergarten
children acting out their stories.

As part of her analysis, the graduate student presented the following original
transcription. She described Vivian Paleys method as one in which teachers cre-
ate the appropriate space and opportunities for students to use multiple modes to
tell their stories.
 DATA SESSION 4 (MIKELA) 93

We notice that already around turn 09, the instructor-analyst provides, as a

hypothesis, a description of what is happening here merely based on the words and
description of actions. He also suggests that is a demonstration of a particular
method, which, in the case of Vivian Paley, is identified as proper to her. The hy-
pothesis is further elaborated near what will have been the end of this 40-minute
analysiswhich, for any more experienced analyst, would only have been a first
beginning. There is always more to be said, dimensions overlooked in the first
quick look, and which is worked out in the course of extensive deepening analyses.
Regular participants in data sessions know that a group of analysts might send two
or three hours to unfold what is happening even in a two-minute clip. In the ses-
sions, the analysts articulate what normally tends to go without saying and what is
implicit in the relations of the members to the setting.
6

Data Session 5 (Kiana)

So the idea is that this kind of analysis . . . you do in a very . . . rigorous way,
you are not just making it up. You are trying to find out what is going on there
without making inferences that go beyond the data, without speculating. So,
Kiana what did you do today? Er . . . cooking.

01 M: Kiana, what did you do today?

02 K: Er . . . cooking. [She avoids the camera and looks at the other side]
03 M: Cooking! What else did you do?

So, Er . . . cooking so . . . then we already get some further description,

avoided the camera and looked at the other side. See, you are already giving me
hints that I did not actually want. But one, Kiana what did you do today, Er . . .
cooking. So the first thing, we would, we have a turn sequence. And there is a
response. We have a {question | response} pair because there is something that
we can hear as a response. There is a hint, of course, a name. It is not a typical
English name I am familiar with. If I wanted to find out I might look up on the
Internet what nationality or what kind of language background that name comes
from. The response is kind ofit is interesting. What did you do today?, and
then we get cooking. So that might give us a first hint of, about what kind of re-
lation there might be. So, Er . . . Cooking, what else did you do? So okay, what
is going on? We have here, cooking, there is a response, there is then a repeti-
tion of the response, almost like an acknowledgment, and then what else did you
do.1

1
The analyst notes that the first part of turn 03 repeats what has been said in turn
02. Such repetition is relevant; it is part of what can be described objectively in the
sense that it is indisputable that the sound-word cooking was produced twice and
by different speakers. As the unfolding analysis shows, this repetition provides
clues as to the type of event that might have produced it. Implicit in the analysts
procedure is the fact that repetition is something special and does not generally
96 CHAPTER 6

Now, we can stop there now our analysis for a second and ask, What is go-
ing on here? Why . . . so there is a question, what did you do today? The se-
cond person articulates something, the first person sort of confirms, so we have,
repeats the term, and then what else did you do? So why would that come? You
know, one could have expected that the first person then might explore that is-
sue whatever it is that [was said].2 So, what else did you do? It is as if, I am say-
ing as if, simply a listing of things was being asked for: the production of a list.3
So, what else did you do?

03 M: Cooking! What else did you do?

04 K: [She turns back to the camera]

occur as a third turn that follows a {question | response} sequence. Consider the
{question | response} sequence {What time is it? | five-twenty}. When asked
by one pedestrian of another, there may not be any continuation. But if a repetition
were to occur, then the range of situations in which the verbal exchange might
have taken place is more constrained. In which way the search for an appropriate
situation is constrained depends on other aspects of the talk. For example, the
third turn might be Five-twenty, already? or Five-twenty? I missed the bus. A
repetition articulated with falling intonation so that it can be heard as a confirma-
tion of an appropriate response can be found in school contexts or in childparent
exchanges.
2
The first few turns would thereby become a reflexive account of the work that has
been opening the production of a more-or-less extended narrative concerning,
here, cooking.
3
The work of which the transcription is a protocol is that of producing a list of
items. The list does not just unfold on its own. Once everything has been said and
done, there will be a list of items recorded in the video and available here in the
transcription. What the transcription does not show is the living work, which the
analysis is in part intended to provide. How do the two participants produce this
list? What is the living work that leaves the list behind as a trace, in the video and
associated transcription? Thus, for example, a researcher might say that the ver-
bal exchange produced a list of activities that a child had engaged in during the
course of the day. Producing a list of activities would then be an account of
what is observed as having been done. But this does not constitute the work of pro-
ducing the list. The two, the account of the work and the work itself must not be
confused. To make the distinction salient, Garfinkel and Sacks (1986) suggest us-
ing expressions such as doing [producing a list of activities]. Doing refers to the
actual, lived, social work praxis that is denoted (i.e., accounted for) by expressions
such as producing a list of activities. Whereas formal methodswhether the
research is qualitative or quantitative/experimentalall focus on the second part,
the expression (account). Ethnomethodological studies focus on the work. What is
the work of opening a conversation? What is the work accomplished in starting a
topic? What is the work in producing a list of the activities of the day that is com-
ing to its end? This transcription is a protocol of the joint work that the two, Kiana
and M:, are doing to establish a list of things that Kiana has participated in during
the day in question.
 DATA SESSION 5 (KIANA) 97

05 Er [She avoids the camera again and spends about 3 seconds to say it]
06 [She turns back to the camera again]
07 [She clenches her fist and waves her right hand]
08 M: Whats that?
09 K: Crayoning.
10 M: Crayoning. So drawing.

Now we get the person K:, Kiana, she turned back to the camera. You can
frame it as a hypothesis, and then look for the evidence. What kind of evidence?
So, there is an interrogation, and it might be the kind, it does not have to be, but
it could be, we might expect something, especially when you go a few turns
down, what else did you do? It is sort of, as if an adult was interacting with a
child or a younger person or a person in a teaching position asking another one
about the kinds of activities they have engaged in.4 And important here is the
listing. And then we have here some further evidence, okay avoided, and then
we see some back and forth and in turn 07, clenches her fist and waves her right
hand, and then M: goes, Whats that? And K:, Crayoning. Well, we can see, we
have two items listed and 09 is actually also a response to 01, but we do not
know at the moment. Why would we get in turn 03 this question for another
one? Was the first one insufficient? Is the production of a list of items im-
portant? Then we get a kind of relationship between a person who is not ESL
((English as second language)) and another one who is ESL, we do not yet
know what the purpose of this conversation might be, but we can hypothesize
possibilities. If it is a teacher with a small kid, it might just be of one type; if it
is at a college with a language instructor, with another one, it might just be some
of the order of getting the other person to talk English, to practice literacy
skills. So what we mobilize here is our everyday familiarity with the world.5 Do

4
This statement captures a hypothesis about the type of situation that could have
produced this concrete transcription. The transcription containing the actual
words exchanged would be the documentary evidence of a general phenomenon
that is concretized in the particular (actual) exchange between M: and Kiana. The
logic is that of abduction generally and undercoded or creative abduction specifi-
cally (Eco 1984; see chapter 10). In abduction, the investigator takes the documen-
tary evidence and invents a general rule (here a possible, hypothesized situation).
The rule, working itself out in the special case (also unknown to the researcher),
produces the actual data at hand. In undercoded abduction, the researcher selects
among a series of equiprobable alternatives (p. 42). In creative abduction, the
researcher has to invent the rulehere the hypothesized kind of situation
explaining the data ex novo. The approach seen here is a mixture of undercoded
and ex novo abduction because at the beginning, there is no hypothesis already
given, especially not a set of multiple hypotheses with equal probability. The ana-
lyst has to invent plausible hypotheses on the basis of the data and, through further
research, select among those produced the one that has the highest probability.
5
This statement is an index towards the transitional nature of the work from crea-
tive abduction to undercoded abduction. On the one hand, the analyst begins with-
out any one hypothesis. But then, given the data at hand, the plausible hypotheses
offering themselves derive from the researchers familiarity with the world and the
98 CHAPTER 6

you see how what we already know is used for the analysis? What we do not
want to do is jump too far; you want to provide evidence for the kinds of things
that you are saying. So what is that? Crayoning, so M: Crayoning. So, drawing.
See how M: repeats again, first she has done it before with the cooking.6 So in
that kind of interaction, the M: person repeats the same words that the other one
has said. Here she does not only repeat it, but also offers an alternative. In lan-
guage teaching, what do teachers, whether they are parents or official teachers
or others, what do they do?: They might repeat it as sort of acknowledgment and
then offering the new word.7 So then K: goes, Yeah! So drawing, so drawing
may, we can hear that so drawing, Yeah! So drawing is the term that K: wanted
to use. So we have that.

10 M: Crayoning. So drawing.
11 K: Yeah! [She nods and smiles]
12 M: Yeah!
13 [She turns her head away]
14 And . . . [She turns back to the camera]
15 [She turns her head away touching her mouth]
16 [She continues touching her mouth and starts touching her tongue]
17 ???? [Japanese? She starts talking again and turns back to the camera right af-
ter that]
18 M: Thats all?

Okay, and then Yea, and then M: Yea, and we have Okay and she turned her
head away; and then we get lots of[the transcriber] gave away another hint,
Japanese words, and she lived in Japan. I was supposed to figure this out. You
gave it away: Kiana is Japanese. Okay, thats all? Even if we did not have this
information so far, what we have is a very limited, quote unquote conversation:
there are just one-word items. Thats all? Kiana nodded and stared at the cam-
era. I saw maybe you went into the pool today, did you? So now we have some-

kinds of social situations that are plausible candidates for describing the specific
situation that has produced this transcription.
6
The statement that M: has repeated the word crayoning is a fact; so is the state-
ment that this is a repetition of something that has occurred earlier: the repetition
of a word. The two facts are slightly different, for in the first case it is the repeti-
tion of a specific word that comes to be noted, whereas in the second case the repe-
tition is noted as an invariant across words and situations. It is a pattern in the
actionsit is a practice or method of doing something. What is being done, the
actual living work, is one of the central questions in the endeavor of this analyst. It
is his problem to which he seeks a solution; but the transcription is the solution to
the problem that the members to the setting (Kiana, M:) are solving, which is the
production of a recognizable social situation. The transcription therefore provides
the analyst with a problem and its solution.
7
In these three statements, the analyst has gone from the concrete data, the repeti-
tion, to the pattern across different instances in the list of words beginning to
emerge, the repetition as a pattern, and to the statement of kinds of situations in
which such repetition of words is known to occur.
 DATA SESSION 5 (KIANA) 99

thing that is being offered as a question, but we do not know how it is being
heard.8 She nodded, and M: Yea. Is it a big pool or a small pool?

18 M: Thats all?
19 K: [Kiana nods and stares at the camera]
20 M: I saw maybe you went into pool today, did you?
21 K: [She nods and leans forward]
22 M: Yeah? Is it a big pool or small pool?
23 K: [She picks her nose with the left hand and she uses a pinch gesture with her
right thumb and index finger]
24 M: Whats that?
25 K: Small pool. [She still uses the pinch gesture and smiles]
26 M: A small pool! Did you have fun?
27 K: Yeah! Yeah! Yeah! Yeah! Yeah! Yeah! Yeah! Yeah! Yeah! [She starts wav-
ing her hands in the air]

There is again something offered as a question, actually ((walks to projection

screen)). It is a {question | answer} sequence; but the responses are limited.
There is only limited number of options. We got this hint that she is Japanese.
Now, here we have a turn sequence, we have a statement, one as it was before, I
would have seen it as an affirmation. Now there comes a question, How about
you? Is this the first time you have used one, and you get Yeah! And now there
is something strange, there is something I cannot make sense of, is this the first
time you have used one? Yea, and now this one, they are very different. And
how old might this person be? From the kind of reaction, it is probablywe
might have the hypothesis that is probably be a younger person, we got this hint
that this is Japanese, and how old will that person be? From the kind of reac-
tions we see, it is probably, we might have the hypothesis that it is a younger
person.9 Okay, so small pool. Okay, so one person sort of is leading the conver-
sation, the other person is very short [in her answers], either making gestures or
giving very short [verbal] answers, and, what we had stated what is going on
hereM: is asking for what the person has done but not following up.10 It is

8
Social action involves two people, two voices, as a minimum unit. A question nev-
er exists as such, in itself. It is as if the left hand in a handclap could be understood
on its own. Thus, {question | response} or {invitation | acceptance} constitute pairs
(wholes) and need to be understood as such. The parts always are parts of a
whole. Consistent with this idea, the analyst suggests hearing a particular state-
ment as a possible offer of a question, leaving open for the moment whether in that
situation it was actually an accepted offer. In this way, the analyst does not commit
to a hearing of the statement, leaving it open to the participants themselves to ex-
hibit their hearing and uptake of it.
9
We see here the tentativeness of the tracing that results from the analysts work.
The hearings are always probable, and always lead to hypotheses, even though
their likelihood may not be equal, as it would be in the case of undercoded abduc-
tion (Eco 1984).
10
The statement is explicitly linking several of the facts and hypotheses articulated
earlier, thereby limiting the range of kinds of situations that would lead in specific
100 CHAPTER 6

almost like a listing, now we have a third thing that the person that the person
has done, and in fact, the third was offered as an option by M:, Yea, Yea, a small
pool. Did you have fun? See, Small pool, very brief, Small pool, same type
again. Kiana, the person says something, and then we get the M: person repeat
the phrase. It is almost like an acknowledgement, A small pool, A small pool
and, Did you have fun, yea yea yea yea yea and then we get further here ((walks
to projection screen)) and what did you eat for lunch?

27 K: Yeah! Yeah! Yeah! Yeah! Yeah! Yeah! Yeah! Yeah! Yeah! [She starts wav-
ing her hands in the air]
28 M: And what did you eat for lunch today?
29 K: Er . . . [She spent about 3 seconds on it]
30 [She turns her head away]
31 Lots. [She turns her head back but avoids looking at the camera]
32 M: Hmm . . . lots.
33 K: [She looks at the camera again]
34 M: Like what?
35 K: Er . . . (mom?) . . . Kianas home. [She gazes away from the camera]
36 [She looks at the camera]
37 M: You ate lots of food at Kianas home?
38 K: Yeah! [She keeps staring at the camera]
39 M: What did you eat in kinder today?
40 K: Lots. [She still stares at the camera]
41 M: Lots . . . And what are you gonna do now?
42 K: Er . . . I dont know. [She looks towards the other direction for a little while
and turned back]

It ((lunch)) is a new item again. What we dont see is a person identifying a

topic and then engaging with it. They are jumping from topic to topic, or nam-
ing an activity and then moving on. It is a practice. So, here, we do not see,
here, we are at 28, and we have listed three words, and here we are into the
fourth, What did you eat for lunch? What did you eat for lunch? Lots. Er. and
then Lots. And also looking away, you may have others like that because turn-
ing the head away may be something like shyness. But we do not know. It is a
hypothesis that needs to be tested. Lots. If this one relates to this, it is kind of
non-traditional. What did you have, what did you eat for lunch? Lots. Lots is
not an answer to a What, it is an amount. Hm, Lots. Lots. We do not have the
age yet. From that type of response, from that, turning the head away, then there
is that picking your nose, who else would pick [as possible individual]? Lots.

instances to the concrete protocols available here in the form of this transcription.
Whereas the previously articulated patterns may occur in other types of situations
as well, such as the repetition of a word or the pattern of repeating articulated
words, the combination of patterns reduces the range of situation types that would
explain this verbal exchange. Whatever it is that the two are doing, the methods of
doing it include the repetition of a word or statement. This repetition is invariant
across instances within this situation. It is characteristic, as the analyst points out
above, of a variety of educational situations as well as of a variety of everyday
situations.
 DATA SESSION 5 (KIANA) 101

Lots. There is that repetition, that repetition again. Looks at the camera, as if,
and then we get a repetition to the prior M:, to the What rather than to a How-
much?11 Er, Kianas home, so the person now, what Kianas home, if this is a
response, we hypothesize that this is an ESL, may be the What and Where are
confused here? This could be a hypothesis, because here Kianas home, so it is
not at home, it is not her home, here is something about (mom?), so mom is an-
other person, we hypothesized initially that it could be mom, it could have been
another mom, but then she could have been an ESL, it could have been at the
same level. It might be a kindergarten teacher, if you, then, see Kianas home,
so you ate lots of food at Kianas home. So we know it is not Kianas home, it is
a child not speaking English or something like that. But what we have is this
true articulation, under M: of what the child says, or a repetition, as if she was
saying, Yea, this is a good word.

38 K: Yeah! [She keeps staring at the camera]

39 M: What did you eat in kinder today?
40 K: Lots. [She still stares at the camera]
41 M: Lots . . . And what are you gonna do now?
42 K: Er . . . I dont know. [She looks towards the other direction for a little while
and turns back]
43 M: Are you tired?
44 K: No. [She shakes her head and smiles]
45 M: I think so. What do you do after you have a shower and bath?
46 K: [She looks towards the other direction]

Now, what did you eat in kinder today? We are not in kindergarten, but the
child is probably attending kindergarten. So now we have, it is not the kinder-
garten teacherno, not it is not, it is probably not the kindergarten teacher.12
We have a person, we have a third person, not a mom, not the teacher, it could
be a tutor. And then what we have here is the response. What, in which, to
what? And, What are you gonna do now? I dont know, Are you tired, No. I
think so, What do you do after you had a shower? Now, this person is, here ((at
the projection screen)), we hear that as a question. I think it is a cue that the
transcriber has heard it as a question. But the transcriber, if she were not very
familiar with the rules of the game, might actually have used the prosodic in-
formation [to detect a question]. I think so. After you had a shower in the bath.
Why would the person ask the little, Kiana, about a shower in the bath? Is she
going to give the shower in the bath? If that were to be the case then it might be
the nanny.13 See how we can draw out certain hypothesesunless, but no. I am

11
Readers will have noticed the repetition of the repetition. There is a possible
question asking for a qualitative What and the reply contains a quantity. The ana-
lytic move is from a description of an action to the description of a patterned ac-
tion.
12
Here, after the analyst has articulated a definitive statement of what is, he is
actually taking it back to state the claim in hypothetical form.
13
In stating the term nanny, the analyst in fact limits the range of persons, or age
range for this person. Kiana is someone who is being taken care of by a nanny. It
102 CHAPTER 6

not familiar with anywhere in the world that you have school where you take a
shower or bath in the morning.

46 K: [She looks towards the other direction]

47 M: Whats next?
48 K: [She looks back to the camera and lifts her fist to her mouth and then opens
her mouth]
49 M: Whats that?
50 K: Brush your teeth. [She still looks at the camera]
51 M: Yes. You gonna brush your teeth. And what we gonna do before bed tonight?
52 K: Er . . . lots play. [She looks down and does not look at the camera until she
says play]
53 M: What? No! I dont think so. How about we. . .?

What will you do after you have your, it could be. Now, we dont know, it is
a limited conversation, we cannot draw much information about the tense, so we
do not know what you do after you have a shower or bath, after you will have
had, so we cannot. So there is a question about what you have after you have a
shower in the bath. This is, we might hypothesize that this is coming up. Or it
could just be a conversation about a situation where you have a shower in the
bath. Probably you are going to bed. This is the thing that definitely happens.
So, M: whats next, whats next. So M:, whats next? So M: mightis she at the
end of her options for the conversation? Is that a signal? Is this a sign that we
can hear as this person being out of resources for what to do? So this could be
the reasoning that Are you tired? No, but then we get, I think so. We do not get a
response, I think so, well and this would then be sort of anticipating that shower
and bath is next. That could be: this is a hypothesis. And which, what comes
next after that? is go to bed, which would then be in continuity to that I think so
youre tired and then Brush your teeth. Oh, we got a response: Yes you are go-
ing to go to bed and What are you going to do before bed tonight? I am still
puzzled about this. It could have been the mother, but there was some infor-
mation, that needs confirmation if this is the mother or the nanny. Play. Brush
your teeth. This is actually interesting, because this is the longest phrase so far.
Play. No I dont think so. How about we . . . what, Lots of play. So there is
something about lots of play, no. I dont think so. How about we, see what we
have about here it seems to me something that . . . it feels like getting a kid
ready thinking about going to bed. You are tired? No. The person says, Yea I
think so. And then, what comes after the bath? Uh brush your teeth. . . uh and

is a person who is accompanied by a caretaker to taking a shower in the bath. Tak-

ing a shower, possibly prior to bedtime also limits the kinds of situation that the
participants are currently in because the talk is a projection of what these individ-
uals are going to do next. The {Whats next? | Brush your teeth} pair is part of
formulating what is to be done next. These formulations, more than other aspects
of the transcription, articulate what is being done; and, either explicitly (when
formulated) or more implicitly, they specify the situation as what it actually is un-
derstood to be.
 DATA SESSION 5 (KIANA) 103

then what are we going to do before going to bed tonight? There is here a prep-
aration for the bed theme, lots of play.

53 M: What? No! I dont think so. How about we. . .?

54 K: Hmm. [She shows a shy smile and looks down]
55 Little bit play. [She looks at the camera and bites her bottom lip]
56 M: Hmm . . . little bit play and maybe read some books?
57 K: Yeah! [She smiles and raises her left hand]
58 M: Sing a song?
59 K: Yeah! [She raises her left hand again]
60 M: Alright! And then (well go to bed?).
61 K: Hmm. [She nods]
62 M: Ok. So you have fun today?
63 K: Yeah. [She nods slightly]
64 M: Alright. Bye!
65 K: [She leaves the sofa and comes close to the camera staring at it]
66 Bye!

This person says, No, I dont think so. How about we . . . so what is going to
come next? Hmm little bit play. Hm little bit play? M: might be mom. Then that
comment we heard earlier on, the one about what did you eat at Kianas place,
is interesting, because if it is mom then that would be a fairly strange way to
talk about your own place.14 But the other thing is if it is ESL . . . or it is . . .
yea. Something like, if this is a Japanese so we have, it seems to me that we
have from this type of conversation, this turn taking, is getting someone ready to
bed. The child does not want, offers Lots of play, M: says, No perhaps not. May
be read some books? Interesting is the . . . what we can see as a limited lan-
guage. If it is not ESL, then it is a very young child. Or, I was thinking about the
child, it could be a visiting child. Or it could bethere is not much we have,
this is a difficult tape because there is not much we get back out [Kiana]. Sing a
song or make ((looks at watch)). We have to bring this [session] to a close so
that we have time for others. So I am scanning a little bit. So you have fun to-
day? Yea, Alright. Bye! So the best we have. We do not have a lot to go by, but
it sounds like, we have sufficient to say that it is an adult with a very young per-
son who is ESL, if it is an ESL, it could be an older, but it is probably a pretty
young kid being prepared to go to bed.15

14
The analyst is not just generating hypotheses but evaluating earlier hypotheses
and observations in terms of the newly emerging ones. Thus, as soon as M: is hy-
pothesized to be the mother of Kiana, the analyst returns to one of the phrases he
has read where M: asks K: what she has eaten at Kianas place. The analyst de-
notes it to be strange, to referring to the home as Kianas place when the cur-
rent exchange takes place in that very place. It is not very common to refer to ones
home in such a roundabout way.
15
The graduate student who prepared the transcription had downloaded the vide-
otape from YouTube. It lasts 2:21 minutes and features a girl sitting on a couch,
sofa, or lounge chair sometimes squarely gazing at the camera, sometimes looking
away (Fig. 6.1).
104 CHAPTER 6

Fig. 6.1 The videotape features a conversation between a mother, who is doing the taping, and
her child Kiana, raised in Japan, but here speaking in English to practice immersion in this lan-
guage. (Source: https://www.youtube.com/watch?v=c6wW4EAYuUA)

The graduate student describes the videotape in this way:

The two participants involved in this study are a three-year-old bilingual
girl and her mother. The girl, Kiana, lives in Japan with her Japanese fa-
ther and Australian mother. The whole process of their conversation is
videotaped by her mother and posed on the YouTube. According to the de-
scription given by mother, Kiana can speak both Japanese and English;
however, her Japanese is better than her English. In their daily life, both
her parents use their native language to communicate with Kiana, and the
mixture of Japanese and English is used between parents, as they are able
to speak two languages. This conversation happens in Kianas home, and
Kiana sits on a sofa talking with her mother in the whole course of this
video clip. (J. Zhang, March 2014)
 PART C

Rigor and the Pragmatics of Relations

In part B, we observe an experienced instructor-analyst at work, in which he at-
tends in a rigorous manner to the pragmatics of the unfolding verbal exchanges. He
attends to how, when, and where participants exhibit something to and for the at-
tention of others and, thereby, make some thing rather than another thing salient
bringing it to bear to the matters at hand. We also see the analyst attend to the rela-
tions, and, by doing so, show the work that lead to differential knowledge-power.
In his approach, he situates himself at the level of the actors themselves, stepping
onto their playing field, elaborating what they say without saying it in so many
words. That is, he does not develop concepts as an expression of the privileged
access of the analyst; instead, he articulates the very ground that allows the ob-
served verbal exchanges to unfold and on which the work of the participants is
based in which the ordered and orderly world comes to be seen and ordered. In my
subtext, point to the ways in which the analyst moves from the undeniable facticity
of the verbal texts to what forms of relations come to be expressed in and produced
by talk. He shows how that talk itself thereby comes to be documentary evidence
in the concrete situation where the recorded talk led to the transcriptions. He also
attends to the temporal relations between what is the presence of the talk and the
relation of the talk to past and future events, which thereby leads to a constitution
of temporality itself. In the chapters of the present part C, I take up, articulate, and
expand on these and related issues. I do so with materials from the transcriptions
that figured in part B but also with additional materials that I selected from pub-
lished resources or from my own texts and data sources. All of the chapters in this
section exhibit the same kind of rigorous attention that the instructor-analyst dis-
played in respect to the particulars of how social relations are brought about and
maintained in and through talk. In fact, it is only because the relations and inter-
subjectivity exist that the participants can engage in the conceptual talk they do
rather than the other way aroundas this is so often assumed and practiced in
much of qualitative research.
7

Turn Sequences

Human beings talk in specific settings that always are characteristic of society.
Students talk (whisper) during lectures, in coffee shops, or (gossip) over a drink.
Professors present at conferences, interact with the audience following these
presentations, discuss data analysis in breakout data sessions, over coffee exchange
ideas about their governments social policies, or chitchat in the evening while
dining. In all these situations, they do not just speak and act as social and psycho-
logical dopesi.e., as if becoming part of a giant machine where they become a
little cogwheel of society, mechanically turning in response to other turning
wheelsbut creatively participate in making each setting what it is and becomes.
This is so in particular when affect is considered, for the schema that figure in so-
ciological and psychological theories never are intimately tied to the inherent af-
fective reflection of events that accompany the intellectual and practical reflec-
tions. Even if a situation is completely novel, members to a setting find ways to
organize it, on the fly and without reflecting too much or at all, without an associ-
ated conceptual schema, and produce a successful encounter. I lived an example of
this while working on this book, when two U.S. computer science colleagues came
to stay with me for three days to engage in data analysis sessions for the purpose of
jointly authoring several research articles. I had met one of the two only twice be-
fore, and had no close personal relations with either. Nevertheless, we pulled off
the stay at my home, organized the days, the meals, our discussions, and the anal-
yses, even though we had never related and worked in this manner before. How did
we do it? In much the same ways in which people creatively produce everyday
encounters in such a manner that not only the resulting organization reveals itself
but also the organizational work is made apparent in the most innocent contribu-
tions to the verbal exchanges. The instructor-analyst in section B shows how re-
searchers may approach transcriptions of such encounters to make visible the ways
in which people exhibit to each other the very aspects of the organizational work
that they enact.
Conducting data analysis in a rigorous manner requires us to have some model
of conversation that makes intelligible how the different forms of approaching a
110 CHAPTER 7

piece of talk or text make different assumptions and produce different analytic re-
sults. We need to understand the more-or-less gross reductions these approaches
produce and how these reductions affect what is and can be said about the talk and
its content. Take, for example, Fragment 7.1, an excerpt from Transcript 1 involv-
ing Heidi, David (Suzuki), Amanda, Ashley, and Michael. In most research tradi-
tions, it is legitimate to focus on the contents of the turns where Heidi speaks and
to attribute the content to her. In such common analyses it is assumed that the
words are an expression of her mind, conceptual framework, cognitive structure, or
personal opinion. Whichever one of the foregoing theoretical terms is used, the
outcome is the same: the statement is tightly linked to the individual bearing the
name Heidi. This is so irrespective of the fact that we are observing a conversation
in a particular setting for a particular purpose and with a double audience: Heidi is
speaking with, to, and for the benefit of the three young individuals in the context
of the recording of a documentary that is to be televised.

Fragment 7.1 (from Transcript 1, Heidi, David Suzuki)

19 Heidi: Not that much although he did get a bit of a facelift, hes lost his double-
chin. But, uh, were really concerned that the cap rock on the hump of
Fred may fall off. That ironstone. And if that falls off the hump could
erode away very quickly.
20 David: But erosion is natural, its going to fade away over time.
21 Heidi: Fred is naturally going to erode away. But if he ever lost his hump, all
wed do is change his name to Humphrey the camel.
22 David: Awwwwwwooooooo
23 Heidi: Ha ha Ha Ha HA Ha
24 Amanda: [Ha ha
25 Michael: [Ha ha
26 Ashley: [Huh ha
27 Heidi: Sometimes it takes a while. Hmmm Hmmm Hmmm

We observe something else happening following turn 21. All five individuals
appear to be laughing. The transcription does not provide information about the
temporal delay between the teenagers laughing and the end of turn 21. But in turn
27, Heidi says, sometimes it takes a while, and then appears to chuckle again.
Why would they all laugh? Why at this point? Some readers may have noted what
is happening over the course of turns 1921: There is something in preparation that
then is taken up as an offer to a joke, but one that is taking a while. It is a joke be-
cause there is an offer followed by laughter (Roth et al. 2011). There would not
have been a joke if the conversation had simply continued. We would see that a
statement was treated as an offer for a joke had a next statement been I dont
think its funny. That is, the second turn can now be heard as a recognition and
rejection of the offer of a joke. We cannot therefore say that Heidi made a joke, or
that she is funny, or make any other statement about Heidi independent of the situ-
ation. That statement was produced for and directed towards the other three or four
individuals present. (David may have heard Heidi talk about the camel-shaped
mountain before, in which case he might have anticipated what was coming next.)
Heidi says that if the mountain lost the hump, she, and perhaps others in the out-
doors center, would rename it; and then she articulates /hmpfri/. This sound can
be heard as the word Humphrey, but, in the case of a realized joke, has been heard
 TURN SEQUENCES 111

as hump-free. Now that I have explained the joke, it no longer is a joke. In

fact, it is when both possible hearings are heard simultaneously in the context of
the mountain having camel shape, or in recognizing that what is offered as a name
can be heard as an adjective, that the (unexpected) contradictory hearing is integral
to the laughter.
To make statements such as Heidi told a joke or Heidi is funny requires
reductions. First, something that clearly is socialclearly recognizable in situation
where a statement may have been a joke in one frame is an insult or a racist com-
ment in anotheris reduced and attributed to the individual. Whether a statement
is a joke or a racist comment depends on the audience. In fact, the statement in and
by itself is nothinguntil it is called something in and by some response. It
therefore does not help us a bit to say that it can be interpreted differently. This
is precisely the same situation that K. Marx discusses in the context of use-value
and exchange-value as manifestations of the value of a commodity. The traditional
approach to the question of what a particular commodity ise.g., a shirtis to
state that it depends on whose perspective we take (Marx/Engels 1962). For the
buyers it is use-value, because it is something they will wear and therefore use; for
the sellers it is exchange-value, because they receive some other good (when it is a
barter exchange), money (in modern economies), or some even more symbolic
(e.g., credit card, bit coin exchanges) in postmodern economies. Marx suggests that
it is not the perspective that one takes, not ones personal interpretation that deter-
mines whether the commodity has use-value or exchange-value. Instead, the two
value forms are manifestations of value. But these manifestations are different be-
cause value is not something in itself but is difference as such: value is not ()
identical with itself. There are very different traditionse.g. the social-
psychological tradition of activity theory started by L. S. Vygotskij (e.g. 1934) and
the tradition started by the group of scholars referred to as the Bakhtin circle (e.g.
Voloinov 1930)that treat words specifically and language generally in the same
holistic way Marx treated value and commodity. Once we have a holistic framing
of words and language-in-use, we can then ask questions about how different
forms of (un/acknowledged, un/recognized) reductions lead to different forms of
analysis and mediate the rigor of the analyses.

A Holistic Model of Language-in-Use

Children do not intend to speak. At some point in their development, they find
themselves speaking. The human species did not decide to invent language so that
humans could speak. Instead, at some time in the evolution of the human species,
in and integral to anthropomorphosis and already living in proto-societies, they
found themselves speaking. In both situations, language-in-use is not that of an
individual but always already something that characterizes verbal exchanges. The
expression verbal exchange is fortunate because it marks a family resemblance of
verbal with economic exchanges. This is not just some statement that I ask readers
to accept. Instead, I conducted an experiment that involved taking Das Kapital
(Marx/Engels 1962) and placing the term sign each time the work employs the
112 CHAPTER 7

Fig. 7.1 This model of speech situations combines sociological (dotted box) and psycho-
logical dimensions (solid box). In the former, vertical dimension each statement is under-
stood as belonging to both speaker and recipient. In the latter, horizontal dimension, the
response begins with active listening and ends with what will have been the last word of the
statement.

term commodity, and inserting an example of a sign each time Marx uses an
example of a commodity. The result are texts that bear a great degree of similarity
withto the point of being indistinguishable fromtexts produced in the late 20th
century by language philosophers (e.g., J. Derrida and P. Ricur) (Roth 2006). In
fact, the Russian znaenie translates not only signification (meaning) but also
value and function, which coincides with the fact that both Vygotskij (1934)
and Voloinov (1930) emphasize that language generally and the word specifically
does not belong to the individual but is a phenomenon characteristic of a group.
The word, Vygotskij says referring to the philosopher L. Feuerbach, is an impossi-
bility for the individual but a reality for two persons in a verbal exchange. Let us
investigate this question by considering a two-turn fragment from the transcription
featuring Heidi the naturalist.

Fragment 7.2 (from Transcript 1, Heidi, David Suzuki)

05 Heidi: Its funny
06 David: As long as youve got a good imagination

The fragment, as presented, already constitutes a reduction without that we may

become aware of it. It does so by attributing a particular phrase to an individual.
Thus, for example, Its funny (turn 05) is attributed to Heidi and As long as
youve got a good imagination (turn 06) is attributed to David. We now focus on
turn 05. In the original situation, there was indeed some sound objectively pro-
duced in the sense that all those presentincluding David, Heidi, Amanda, Ash-
ley, Michael, camera person, producer, etc.could hear it and the camera could
record it. It is easily established that the vibrations of Heidis vocal cords produced
the sound. However, it is only the latter that is articulated explicitly in the tran-
scription (Fragment 7.2). What is not explicit is the fact that those vibrations reso-
nate in the inner ear of the others. That is, in this verbal exchange, the sound-word
is a reality for two in the way Vygotskij (1934) stated. The sound-word thereby is
positioned between those participating in the verbal exchange in the same way that
the commodity is positioned in the hands of buyer and seller during the economic
exchange. This positioning is represented in an alternative rendering of the two
turns (Fig. 7.1), which marks the received words and statements in grey. In the
 TURN SEQUENCES 113

instance of turns 2426, there would be a temporal spacing between Humphrey

and the laughter, but Humphrey would appear in the turns attributed to all three
teenagers. Considered in this way, any response, reply, or second turn consists not
only in the words said but also in the words heard. This has immediate conse-
quences for the value (signification, sense) and function of the word in the
same way that it has for the value of the commodity.
First, in an economic exchange, the value is different for seller and buyer,
which is paralleled by the different value (signification, sense) that the sound-
word has for speaker and recipient. Second, in dialectical logic the different values
(signification, sense) are the result of an inner difference of the word and lan-
guage-in-use. Some readers might want to argue that the dialectical approach is
just another way of theorizing the classical logical approach. This is not so, how-
ever, for the classical approach can be shown to result when the dialectical case is
reduced based on an implicit assumption. Classical logic is premised on the self-
identity assumption and different from everything else: p = p and p p ( syn-
onymous with not). In dialectical logic, everything is understood to be different
not only from everything else but from itself as well. Nothing is . . . because the
universe and life are continuously becoming. A thing or phenomenon is different
not only from every other thing or phenomenon but even from itself. This is so for
the commodity in the economic exchange as it is for the word (language-in-use) in
verbal exchange. There is continuous transition that manifests itself twice: the
movement of the word (language-in-use) in its unfolding (temporally) and towards
the other (spatially). These two manifestations or expressed in the model (Fig. 7.1)
along two dimensions: a sociological and a psychological.
The position of classical logic with respect to language philosophy is achieved
by reducing the movement of the word to stasis. It can then be attributed to one
person, the speaker. Moreover, because it is assumed to be self-identical, the dif-
ferent senses (significations) of a word are then attributed to the different
dis/positions of the different participants in an exchange, much as the difference
between use-value and exchange-value of a commodity was attributed to the dif-
ferent positions individuals take in an exchange relation.1
The sociological dimension of the model takes into account that language-in-use
exists for speaker and recipient simultaneously. The language-in-use is not self-
identical, which manifests itself in the different ways participants hear it. In fact,
we need to insist that the Saying must not be reduced to the Said, for the latter can
be said to exist only when Saying has ended. Even then, the Said does not consti-
tute a fixed entity for the speaker or the audience, for, as Husserl (1928) shows,
some past thing or event continuously changes because it is seen through the veil
of the intervening experiences. This constitutes the psychological dimension (Fig.
7.1) of language-in-use, which is such that no statement ever is self-identical. That
is, we are never in a position to state what really has been said because it changes
while Saying is happening, and when Saying comes to a stop the Said changes
because always seen through the veil of what occurs between the end of the Saying

1
In 2004, Michael G. Hoffmann, a classical philosopher by training, spent about six months
as postdoctoral fellow in my research laboratory. He explained to us this position of classi-
cal philosophy on Marxs category of value.
114 CHAPTER 7

and the experience right up to the present. Stating that someone has said something
definite and definitive requires a reduction (abstraction from).
Doing data analysis rigorously means making explicit the instruments of scien-
tific production. Our theoretical and methodological language-in-use is a prime
candidate for such investigation, because it is in effect an immense repository of
naturalized preconstructions, and thus of preconstructions that are ignored as such
and which can function as unconscious instruments of construction (Bourdieu
1992, p. 241). Without such an investigation of what our presuppositions do, here
pertaining to the reductions concerning language-in-use, means that researchers
literally do not know what they are doing. I am not saying that it is illegitimate to
engage in such reductions but rather that one should be more or less fully aware of
the effects that the reduction produces that cannot be ascribed to the phenomenon
itself. Consider the following fragment from what we now know to have been an
interview involving a journalist talking to the principal and some students of a
school that has introduced a policy allowing and encouraging rough-and-tumble
play during recess. In Fragment 7.3, we find the interviewer (I:), the school princi-
pal (P:), and two female Maori students (M1:, M2:).

Fragment 7.3 (from Transcript 3, Bullrush)

06 I: Have you ever been hurt?
07 M2: Guys
08 M1: Ya
09 M2: Yeh you guys, were supposed to be playing bullrush (yelled from be-
hind)
10 M1: Ya, actually, really . . . ya, I have really been hurt.
11 I: And what happened then?
12 M1: I think I got up and dealt with it.
13 P: I think were wrapping kids up in cotton wool a little too much . . . um .
. . years ago a good parent was somebody who just let kids play. Nowa-
days a good parent might be considered to be somebody [who] takes
them to dancing lessons . . . and rugby practice.
14 I: there may be some people, some parents who think, hmm, I dont want
my children, my little girls, playing bullrush! What do you say to those
people?
15 M1: Well, if you think you cant handle it, well then, dont play . . . thats
just pretty much it.

Some researchers might be interested in finding out about attitudes and beliefs.
They might then attribute a particular attitude and belief to the principal, who, in
turn 13, says that todays kids are wrapped up in cotton and talks about how par-
ents in the past let their children just play whereas today they bring them to dance
lessons (girls) and rugby practice (boys). But if we made such an attribution to the
principal, that is, if we were to reduce the collectively produced text, cut it into
pieces, and assigned bits and pieces to different individuals, then this attribution
comes at a high price. This is so because we lose the connection the piece of text
had to all the other pieces of the text constituted by the transcription and, ultimate-
ly, by the lived situation that is never made present again in its entirety in and
through the videotape or transcription.
 TURN SEQUENCES 115

One of the dimensions we lose is the fact that the stretch of talk is part of an
exchange. It has use-value for the recipient, to employ the term from political eco-
nomics. It is a text in a language that the principal has received from the general-
ized other that the culture represents and that, in the exchange, is produced for the
other, to whom the language therefore returns. As the transcription shows, whatev-
er is stated in turn 13 appears to have been unproblematic, as the interviewer takes
up and continues the current topic. Therefore, turn 13 was intelligible. But if so,
that is, if turn 13 was intelligible, then whatever the statement is about and how it
is formed has already been a possibility for the two participants; in fact, it has been
a possibility in this language community as such. If it was possible, it was possible
for at least the two speakers. The phrase and its content, therefore, represent more
than just what the principals beliefs. It represents a way of talking about a particu-
lar topichere the ways in which children are raised and the kinds of activities
they engage in and how much more todays children are protected as compared to
their peers in a previous generation. As a way of talking about and constituting the
topic at hand, the principal does not own this stretch of talk (text). This text comes
from a collective possibility of talking. In reducing the text and attributing it to the
principal, in making it the effluent from his mind, we lose sight of the talk as a
collective possibility of talking. The stretch of talk, as shown in the model (Fig.
7.1), belongs to speaker and recipient. We lose sight that a person does not have to
have thought or talked about the topic ever before to be able to produce it. We may
find it intelligible even if we had never thought about it before. That is, we may
talk this way without having formed an idea, thought, conception, mental structure,
conceptual framework, or belief. Through the principals mouth, language itself
speaks allowing us to hear what it has to say.
The forgoing example shows that we lose theoretical rigor together with the loss
in method-related rigor; and by giving up method-related rigor, we lose theoretical
rigor. We make attributions for which we do not have the slightest evidence. Time
and again, we may hear interviewees tell interviewers that they have not thought
about some topic introduced by the latter (Roth 2008). This means that they could
not have constructed a mental framework in this respect. And yet, the interviewees
have no trouble talking about the topic. If they can do so it is because language
itself provides topics and ways of talking: language itself speaks (Heidegger 1985).
Moreover, even if the interviewer I: had never heard such talk as that articulated in
turn 13 before, it would have been intelligible. It would be intelligible even if, and
precisely because, a research might categorize a stretch of talk as misconception,
a description and explanation of something at odds with the current scientific can-
on. Something can be said to constitute a misconception precisely because it is
intelligible; otherwise those sound-words would only be gibberish.
When we reduce talk and attribute it to individuals, we also lose sight of the fact
that we do not have to construct intersubjectivity. Language-in-use presupposes
mutual intelligibility. It is when this presupposition appears to be violated that we,
as witnesses of the situation or as analysts of the videotapes or transcriptions, will
be able to see trouble being formulated. This is precisely what the analyst in part B
is attentive to, what the participants themselves make available to each other and at
what point they might make available to each other potential trouble in the presup-
posed intelligibility of talk. Thus, between the end of turn 21 and the beginning of
116 CHAPTER 7

the three childrens laughter (turns 2426), 6.54 seconds pass. The statement in
turn 27, Sometimes it takes a while, therefore can be heard as a description of
what has just occurred, between the offering for a joke and its realization (ac-
ceptance) as such. It notes that there had been a slow uptake, and, indeed, it may
connote that a joke rather than something else had been offered, and the intended
recipients had been slow in their uptake. Turn 27, therefore, articulates a recogni-
tion of an initial incomprehension followed by comprehension, made available in
and through the laughter that is heard as paired with, and acceptance of, the offer
of a joke. Rigorous data analysis allows us to recognize what has been going on in
this situation, whereas less attention to the particulars of the exchange; and concern
with what is in the heads of these individuals may lead us to fail to recognize how
the participants work together to exhibit what is currently happeninghere the
slow uptake of the joke in the {turn 21 | turn 2426} pair sequence and a faster
uptake in the {turn 21 | turn 22} pair.
In this case, we another reduction can be seen at work, which makes it more
difficult to know what was happening in the situation. When the soundtrack of the
documentary is analyzed, its full phonetic transcription is /'hmpfri/. However, the
name Humphrey tends to be pronounced as /'hmfri/, that is, without the phoneme
/p/. For this reason, the unexpected relation between Humphrey and hump-free
upon which the joke as irreducible social phenomenon is builti.e., the concurrent
hearing of the two words based on the same soundhas disappeared in the tran-
scription.
The two types of reduction described so far can be understood as occurring dur-
ing a particular stage in the trajectory from the lived-in world that comes to be de-
scribed, theorized, or used as example in research articles (Fig. 7.2). Here this re-
duction occurs in the transition from the videotape to the transcript. First, on the
videotape, there is sound. This sound and the associated sound-words are available
to participants in the situation and to the researcher. Although generally available,
the sound-word is attributed in the transcription to individual speakers. Once at-
tributed to the speakers, the language-in-use is treated as an emanation from the
individual speakers minds. That is, what is denoted in grey letters in Fig. 7.1
comes to be omitted. The fact that sounds are not only produced but also received
is lost in the transition from the tape to the transcription. With this reduction is
therefore lost the inherently shared nature of any stretch of talk and the underlying
presupposition of intersubjectivity. That the shared nature of statement underlies
any verbal exchange can be seen from the fact that queries such as what do you
mean by . . .? tend to be infrequent. In the everyday world, there is a popular ex-
pression to denote such situations: the interlocutors are not on the same page. Inter-
locutors normally tend to engage in exchanges that get them back onto the same
page. This effort itself requires a certain level of being on the same page.
Second, in the transition from sound to sound words, the transcribers make se-
lections. Thus, rendering /'hmpfri/ as Humphrey constitutes a reduction in the
sense that it makes difficult to read that one can clearly hear the /p/ and, therefore,
 TURN SEQUENCES 117

the difference from the normal pronunciation of the Christian name, which is
/'hmfri/.2
The point of the foregoing discussion is not to critique reduction. It is evident
from the model of the movement of research (Fig. 7.2) that there are unavoidable
reductions occurring when we record some everyday situationthe first reduction
being that the camera is operated by a person or has been positioned in some
wayand this recording inherently is associated with blind spots: Why this rather
than another camera angle? Why this rather than another angle of inclusion? Why
this rather than another zoom? Rigorous data analysis has to keep in mind and be
aware of the reductions that have occurred and the implications these unavoidably
will have for the analysis and the claims that are reported. Believing that the tape,
transcription, analysis, or article can get us back to the original happening in one or
another way is part of a nostalgia dynamic, a term used to denote the desire,
regularly felt (and encountered in others), born perhaps, of a Sartrean mauvais fois
[sic], for greater simplicity, authenticity, and directness (Ashmore and Reed 2000,
para. 22). According to these authors, this nostalgia dynamic, which is from the
right towards the left in the model (Fig. 7.2), frequently tends to follow the surfeit
of sophistication and irony, an overdoes of ramified interpretations; simply, just
too much rightward-ness (para. 22) that is associated with the analytic amplifica-
tion of articulating the universal and general that can be observed in any particular
real-life happening.
A holistic model of language-in-use considers it in its relation to everything else
available to and presupposed in the everyday conduct of affairs. Taking language-
in-use in this way allows the analyst in part B to reconstruct the type of situations
that could have led to the concrete videotape and transcription in the concrete case
of the event recorded. The analyst cannot get us back to the event. No analysis ever
can get us back. Tape and transcription allow us to make present again, in reduced
form, aspects of what has happened. On the other hand, such a reconstruction be-
comes near impossible if the language-in-use is treated as if it merely were an em-
anation of individual human minds reacting somehow to other individual human
minds.

The Dual Orientation of Talk

In conversation analysis (i.e., the method), the emphasis tends to be on the back-
ward relation of a turn to the one that it relates to complete a turn pair. This, in the
above interview involving a journalist, a school principal, and some Maori female
students, turn 10 can be heard in relation to turn 06 because, together, the two turns
complete a {question | reply} pair.

Fragment 7.4 (from Transcript 3, Bullrush)

> 06 I: Have you ever been hurt?

2
Some online services for pronunciation keys that do not use the rules of the International
Phonetics Association transcribe the sound of the name Humphrey as hum-free.
118 CHAPTER 7

07 M2: Guys
08 M1: Ya
09 M2: Yeh you guys, were supposed to be playing bullrush (yelled from be-
hind)
> 10 M1: Ya, actually, really . . . ya, I have really been hurt.

The underlying dynamic is that of attributing a causal role to the first turn, ques-
tion, as a consequence of establishing the effect (reply): effect ! cause (Nietzsche
1954). This dynamic had been established as the fundamental way in which cause
effect relations really are constituted rather than the presupposed cause ! effect
that underlies classical logic and scientific reasoning alike. This dynamic, however,
is only part of the picture. As articulated above, there is also a forward orientation
in any form of speech is for the recipient, a fact that is discussed in conversation
analysis under the term recipient design. There is therefore both a backward orien-
tation and a forward orientation in each turn, which, in rigorous data analysis, has
to be retained and attended to. Take a look at turn 11 in the interview situation.

Fragment 7.5 (from Transcript 3, Bullrush)

10 M1: Ya, actually, really . . . ya, I have really been hurt.
> 11 I: And what happened then?
12 M1: I think I got up and dealt with it.

From the turn pair perspective, turn 11 constitutes the common point of the
pairs {turn 10 | turn 11} and {turn 11 | turn 12}. Turn 11 is the second member of a
pair. It constitutes the active uptake of turn 10, which in fact is heard. A fuller tran-
scription of those three turns would include the hearing:

Fragment 7.6 (from Transcript 3, Bullrush)

10 M1: {Have you ever been hurt?} Ya, actually, really . . . ya, I have really
been hurt.
> 11 I: {Ya, actually, really . . . ya, I have really been hurt.} And what hap-
pened then?
12 M1 {And what happened then?} I think I got up and dealt with it.

Here, each turn includes the hearingi.e. the active listeningand talking parts
of the response. The hearing itself is not available to anyone else other than in the
way it mediates the Saying that follows. That is, the second, unbracketed part of
the turn is an index to the bracketed part. Because the Saying is developing, what
has been heard is itself to be understood as under developmentbecause we would
be back at the causeeffect nostalgia otherwise. Fragment 7.6 also makes evident
the recipient design aspect of the utterance, because whatever appears in the se-
cond part of turn 11 becomes the first part of turn 12. Thus, not any text does the
trick.3 Whatever comes forth is based on the assumption that it is intelligible to the

3
I am often wondering about the contradiction that is apparent to me between two practices:
(a) New researchers often tell me that they have written an article and then ask me where to
submit it; and (b) the same new researchers will talk about what they have done on a given
day differently to a colleague, their spouse, and their children. The two situations are con-
tradictory, because in the latter case, these individuals clearly design their accounts for the
 TURN SEQUENCES 119

intended and designated recipient in this situation. That is, recipient design also is
situation design, for whatever we say to a colleague will differ whether it is during
a faculty meeting, during a data analysis session, in the evening over dinner, or
while writing an article in which we analyze our own verbal exchanges in one or
the other of the preceding contexts (e.g. Antaki et al. 2008).
It is only because of this double orientation of each turn that we actually get a
conversation as a whole event, a unit, rather than as something composed of indi-
vidual statements (elements) the addition of which make up some whole. In this
latter way, conversations tend to be treated in many analyses of interviews that
attribute statements and their conceptual contents to individuals. Such approaches
also fail to realize that verbal exchanges do more than continue some topic. They
are integral part and reflections (protocols) of the relational work jointly accom-
plished by the interlocutors. Thus, I: and M1: are not just responding to someone
elses talk in some abstract way. Instead, the journalist had come to the school and
had made arrangements for producing an item to be featured on the local news.
Both I: and M1: are oriented towards the event as one where an interview is to be
produced that can be featured on the local news. They therefore collude to produce
something that after the fact can be designated to be an appropriate evente.g.,
shown as part of the report on a school in which rough-and-tumble play has been
introduced as an integral part of the school day. In a somewhat overstated way,
each person produces statements such that they are projected to be suited to the
kind of situations that they collude in producing. This constitutes a limited subset
of statements from the set of possible statements a person may produce. Thus, we
have to anticipate the reply to a query to be different as a function of the activity: a
person will respond differently to the statement How are you today? when ar-
ticulated by the cashier in the supermarket, by her psychoanalyst at the beginning
of a session, or by the spouse after getting up in the morning. It is precisely this
recipient-in-situation design of any verbal articulation that allows the analyst in the
chapters of part B to reconstruct with high fidelity the type of situation that has led
to the concrete transcription in hand.
The double orientation of each articulation is facilitated when researchers also
subscribe to theoretical models consistent with it. Thus, for example, if researchers
subscribe to the way in which talk is understood within the sociology of emotion,
as integral part of an interaction ritual chain (Collins 2004), then they are (have to
feel) encouraged to attend to the chain-like nature of conversations. A chain con-
sists of interlinked chain links rather than of individual joints (Fig. 7.3a); it is is a
nice metaphor for a conversation and the way it is analyzed/theorized in different
traditions of research. Many received theories and methods treat a turn as an ele-
ment from which the conversation is constituted. But (oval) rings do not make a
chain link (Fig. 7.3b). What makes a chain a chain is the linki-ness, the fact that
there is an uninterrupted chain of interlocks (Fig. 7.3c). Each lock projects back-
ward and forward making a whole link only with the previous and the following
lock. But it needs to be understood in terms of the whole of which it is integral

recipients, both in content and genre or style. But in the former case, they have authored a
text independent of the consideration of intended/projected audience and the kinds of styles
or genres it may find acceptable and intelligible.
120 CHAPTER 7

Fig. 7.3 a. A chain consists of interlocked rings. b. A ring does not retain the chainli-
ness of a chain. c. The link retains the interlocking as a fundamental feature and, therefore,
the chainliness of the chain.

part: the whole exists only in and through the parts, and the parts are parts only
because of the whole they constitute. In a strong sense, the chain is not a whole, the
chainliness of a chain comes to be exhibited in situation where it is used as a chain
but no in situation where it is used as a hammer or as a weapon to fend of or ag-
gress another person. The notion of language-game (Wittgenstein 1997), which is
the irreducible unit of a practical activity together with its associated language-in-
use, is designed to retain this interlocking within a conversation and of the con-
versation with the practical activity at hand.
This orientation towards the overall activity is evident in Transcript 3 (Bull-
rush), where participants act such that we can recognize the mutual rolesas iden-
tified by the analyst in chapter 4. In talking and in staffing positions in turn-taking
sequences as they do, the participants produce an interview for the purpose of fea-
turing it on the daily news. The transcription is a protocol of the work that leads to
that initially only projected product. Close and careful attention to the turn-taking
sequences and the regularity in which speakers staff the different available posi-
tions allow the analyst to hypothesize that it is an interview conducted by a re-
searcher, journalist, or other person interested in the place where rough-and-tumble
play has been re-instituted as a feature of relations between children. Close atten-
tion to the ways in which talk not only projects content to come but also projects
who the intended audience is allows the analyst to detect that there are in fact dif-
ferent recipients for the text attributed to I:.
This same attention allows the analyst to detect that the transcriptions are frag-
ments taken somewhere from the inside of the verbal exchange rather than from
other parts. It is this orientation that allows the careful analyst to detect an end-in-
the-making in turns 6466 of Transcript 5 (Kiana).

Fragment 7.7 (from Transcript 5, Kiana)

64 M: Alright. Bye!
65 K: [She leaves the sofa and comes close to the camera staring at it]
66 Bye!

In the turn-taking sequence, turn 64 constitutes the invitation to the ending (of
the recording), subsequently accepted in turns 65 and 66. There is therefore an
 TURN SEQUENCES 121

{invitation | acceptance} to produce the ending of the recorded event, which comes
with the turning black of the YouTube video screen. It also flags the possibility
that M: orients Kiana to address the future audience of the video, which may be
one or more familiar individuals (e.g., an aunt or the father currently away from
home) or some generic audience that is assumed to view the video online.
8

Knowledge-Power and Institutional Relations

We should rather admit that power produces knowledge (and not simply by
encouraging it because it serves power or by applying it because it is useful);
that power and knowledge directly imply one another; that there is no power
relation without the correlative constitution of a field of knowledge, nor any
knowledge that does not presuppose and constitute at the same time power
relations. These power-knowledge relations are to be analyzed, therefore,
not on the basis of a subject of knowledge who is or is not free in relation to
the power system, but, on the contrary, one has to consider the subject who
knows, the objects to be known and the modalities of knowledge as so many
effects of these fundamental implications of power-knowledge and their his-
torical transformations. In short, it is not the activity of the subject of
knowledge that would produce a corpus of knowledge, useful or resistant to
power, but that power-knowledge, the processes and struggles that traverse
and constitute it, would determine the forms and possible domains of
knowledge. (Foucault 1975, p. 36, emphases added)
One must think relationally. Now, it is easier to think in terms of realities
that can be touched with the finger, in a sense, such as groups or individu-
als, than in terms of relations. It is easier for instance to think of social differ-
entiation in the form of groups defined as populations, as with the realist no-
tion of class, or even in terms of antagonisms between these groups, than in
the form of a space of relations. (Bourdieu 1992, p. 228, original emphasis)
Concepts are an important aspect in the classification of circumstances that histori-
cally provided an advantage to the human species; they are integral to the human
will to power and the more or less accurate anticipation of results associated with
specific actions (Nietzsche 1954). However, concepts also operate like horoscopes
in the sense that they constitute a form of confirmatory bias, subsuming experienc-
es even if other concepts might be better suited. In my graduate seminars on re-
search methods, I observe this time and again when playing some videotape for
analysis. The novice researchers tend to use the categories teacher and student to
124 CHAPTER 8

explain the relations in classrooms rather than focusing on who learns what, when,
and how. A classical example arises when novice and experienced researchers are
familiar with the theoretical concept of zone of proximal development, which tends
to be understood in terms of a teacher providing opportunities to students, who first
constructs something in the social sphere prior to construction it for themselves
(internally) (Roth and Radford 2010). How then do teachers become better at
teaching? In fact, if teaching makes us better teachers, then the analysis of class-
room relations ought to be able to detect the learning on the part of the teachers
arising in relations with students, who thereby contribute to the development of
teachers. But the concept zone of proximal development, understood and used as it
traditionally is, biases researchers attention on the institutionally designated stu-
dent-learner. In this context, power tends to be ascribed to the institutionally
designated teacher completely failing to acknowledge that in any case, classroom
events are joint achievements rather than the result of a god-like demiurge control-
ling every aspect of life. In the opening quotation, Foucault orients us in a different
way to the analysis of human relations.
In part B of this book, we see the analyst work in accord with the writings of
Foucault. The analysis exhibits the nature of the relations and shows how these
produce gradients of knowledge-power. That is, because the analyst does not know
beforehand about the institutional positions of the participants, and knows nothing
a priori about these participants, he is forced to investigate the particulars of the
relations and what these constitute and how. This approach to data analysis reveals
how any power is exercised rather than possessed, and how it is not a privi-
lege, acquired or preserved of the dominant class, but the effect of the ensemble of
its strategic positions (Foucault 1975, p. 35). That is, rather than falling prey to
the confirmatory bias that comes with the rapid subsumption of events to reigning
concepts, data-driven analysis more rigorously derives the particulars of relations
and shows what these actually rather than presumedly produce. Just as scientists
are critical of those who read and see confirmed the daily horoscope, the rigorously
working data analyst eschews rapid application of concepts to explain what is hap-
pening attempting instead to show how relations themselves give rise to differ-
ences along a knowledge-power gradient. This also allows us to understand why
lessons often are not successful at all despite the preparations and best intentions of
a teacher: it takes the living curriculum as the result of the relations rather than
viewing the curriculum in terms of the plans that a teacher may have written our or
received (Roth 2014a). Rigorous analysts ask what is really rather than presumedly
happening. They then will find out that any power does not apply itself purely and
simply, like an obligation or an interdiction to those who do not have it and in-
stead find that it invests them, is passed by and passes through them; it rests upon
them, just as they themselves, in their struggle against it, rest upon the holds it has
over them (p. 35). In his work, the analyst featured in part B exhibits this orienta-
tion to derive differences in knowledge-power from the relations rather than ex-
plaining relations in terms of differences in knowledge-power. Consider Fragment
8.1, which is taken from Transcript 1.
 KNOWLEDGE-POWER & INSTITUTIONAL RELATIONS 125

Fragment 8.1 (from Transcript 1, Heidi, David Suzuki)

09 Heidi: Well, it does. Some people claim it might be a kind of hallucination
now watch out for the cactus. When we get around here, you want to
take a look around and see if theres any landforms that look like some-
thing that would be familiar to you, not just like a rock. So what do you
think that landform over there is? Does this one look, look like anything
to you?
10 Amanda: Hmm. Oh! That rock right there looks like a camel
11 Heidi: Oh . . . no . . . thats it! We actually have a name for this guy. We call
him Fred the camel . . . see the hump . . . see the big droopy lips point-
ing to the left. And if you look off in the back can you see anything
else?
12 Amanda: ?
13 Michael: No.
14 David: Ill give you a clue. Where are camels found?
15 Ashley: Egypt.
16 David: Very good!

We observe, as the analyst has done, that there is a query-reply-evaluation se-

quence across turn 09 (beginning) through turn 11 (end). That is, the third turn in
this sequence shows that the query was not just any query, but a query to which the
correct reply was already known. The actual reply in turn 10 then was assessed
against the initially hidden correct answer. In being designated as correct, the actu-
al reply also comes to be denoted as the correct one, which, in this move, reveals
itself to all those not in the same position as Heidi. The same sequence is repeated
in an expanded form when the reply is not forthcoming. Thus, can you see any-
thing else is treated in turns 12 and 13 as a question to which Amanda and Mi-
chael do not know the result. David, in stating that he is giving a clue co-articulates
knowing what the correct response to the intended question is. In fact, David for-
mulates what follows as a clue. He not only makes a statement that has the func-
tion of a clue but in fact states what could have functioned as another question
Where are camels found?to constitute a clue. By saying that he is providing a
clue, David makes publicly available the fact that he already knows the answer.
Readers recall that this led the analyst to describe the relation between Heidi and
David as being in cahoots. The sequence concerning the clue itself takes the que-
ry-reply-evaluation form that we observe in previous case of the camel-shaped
landform.
In Fragment 8.1, knowledge with respect to that part of the world (the Canadian
Badlands) comes to be marked as differentially distributed: David and Heidi know
and Amanda, Ashley, and Michael do not. With that differential knowledge also
comes differential power. David and Heidi not only come to be marked as knowing
but, simultaneously, as those in power to assess the knowledge of others. That is,
knowledge and power gradients are constituted simultaneously. The participants
themselves do so. Equally important, they make it known to everyone else witness-
ingwho may be directly witnessing the event or vicariously accessing it, as the
analyst in part B or as the readers and I herethat there are in fact such differences
produced and reproduced simultaneously. These differences may be reproduced
because of presumed pre-existing differences, but these are simultaneously pro-
duced because prior to that particular part of transcription, the differences in
126 CHAPTER 8

knowledge with respect to the landformscamel and pyramidwere not known.

That is, these differences were not known and, therefore, could not have been the
causes of the way in which the relation (conversation) unfolded. These differences
came to be known in the very ways in which the relation unfolded in and through
the contributions of the participants. Thus, the surprise we can hear in the opening
of turn 11 is an indication that Amanda, too, is now recognized as part of the club
of those who know and have seen the camel-shaped landform. Rigorous attention
to what happens in the data, therefore, allows us to understand with Foucault that it
is not the activity of the subject of knowledge that produces a corpus of
knowledge, useful or resistant to power, but that it is in fact the relation that simul-
taneously produces power and knowledge gradients in and through processes and
struggles that traverse and constitute knowledge-power. These processes and
struggles also determine the forms and possible domains of knowledge that come
to be established or contested.
Continuing with this line of argument, we may actually approach knowledge not
as a definite category but always as something contested and as the product of
struggles. The contested nature may become invisible without disappearing in
some situationssuch as we can observe in the relation between Heidi and David,
on the one hand, and Amanda, Ashley, and Michael, on the other hand. The strug-
gles, too, might be more usefully considered to constitute a continuum rather than
being part of a dichotomy opposed to struggle-free events. We can then see the
relation itself as a struggle for establishing what constitutes relevant knowledge
and who knows versus who does not (Vygotskij, 2005, uses the term drama in-
stead of struggle). Turns 1114 do in fact ascertain that Heidi and David know,
whereas Amanda, Ashley, and Michael do not. In contrast, the turns 0911 also is
treated as a struggle in which a knowledge or perceptual gradient could not be es-
tablished, for it turns out that Amanda does in fact see what there is to be seen and,
simultaneously, that Heidi is not the only one over the related knowledge (i.e.,
landform = camel-shaped).
To further exemplify the rigorous, data-driven route to analysis, I draw in the
following section on a database that I established for a project concerning graphs
and graphing. Attending to the details of interactions over the graphsassociated
with rigorous data analysis that did not import institutional positions of people to
make these explanatory resources employed to account for the interactionsI
started to make some interesting discoveries (Roth and Middleton 2006). That
same approach also allowed me to make some interesting discoveries when taking
a look at teacher-student relations with respect to the concept of zone of proximal
development frequently mobilized in this context (Roth and Radford 2010). For
example, I came to see that not only institutionally designated students learn and
institutionally designated teachers teach but also students teach while learning and
teachers learn while teaching. The students teaching and teachers learning had
become invisible when using zone of proximal development as a lens, especially
when it is regarded as being controlled by those who know and becoming a learn-
ing space for those who do not.
 KNOWLEDGE-POWER & INSTITUTIONAL RELATIONS 127

Knowledge-Power: Gradients and Struggles

In this section, I exemplify how rigorous data analysis might proceed. For didactic
purposes, I neither reveal the institutional identities of the two participants, their
gender, or any other background information other than what constitutes the cur-
rent topic of their talk. When I first conducted these analysesalso including D.
Middleton in a three-week, 6-hour/day analysis of one 25-minute tape, I bracketed
what could be seen disallowing any background information to enter into the anal-
ysis as analytic resource. This approach is typical, for example, for phenomenolog-
ical research where not only everyday understandings and concepts are bracketed
but also scientific understandings and theories (Husserl 1976a). That is, I voluntari-
ly put myself in the situation of the analyst in part B. Bracketing is a form of
doubting required to deal with the preconstructed that lurks and besieges social
analysts everywhere they turn (e.g. Bourdieu 1992). Everything, however much we
are convinced by it, however much we are sure to have an adequate understanding
thereof, is subject to doubt and requires inquiry.

The Joint Production of Relations

The Topic. The verbal exchange analyzed below is over and about the contents
of a large sheet of paper, which features three graphs, an explanatory text, and an
instruction (Fig. 8.1). To assist readers in understanding what undergraduate stu-
dents taking an introductory ecology course would be required to know, I provide
the following information all the while emphasizing that the participants possibly
know neither what I explain nor how the other is located with respect to this way
of talking about the graph. All three graphs represent the amount of plant growth as
a function of two nutrients that may be provided in different amounts. For each
combination of the two nutrients (R1, R2), we can tell, by looking at the graph what
the plant growth is. To do so, it helps seeing each of the three graphs as two-
dimensional presentations of a three-dimensional graph. Thus, for example, graph
(a) can be envisioned as something like the part of a pyramid (Fig. 8.2) with the
edge running through the corners seen in Fig. 8.1. For each pair of nutrients on a
grid spanned by the pyramid base, one has to go up until hitting the surface (see
dotted line in Fig. 8.2). Whatever the corresponding height would be the plant
growth. Thus, in the example, the plant growth (dark circle) would be greater than
50 but considerably below 100 units. Looking from straight above onto Fig. 8.2
along the vertical axis one would precisely see graph (a) in Fig. 8.1.

Transcription and Analysis. A and D are talking over and about the sheet on
which the contents of Fig. 8.1 are printed. Readers note that in Fragment 8.2, there
is no beginning work observable, which supports the hypothesis that we are some
stretch into the verbal exchange. The point of this section is to demonstrate what
can be recovered from the situation even without knowing who the two partici-
pants are. In fact, rigorous data analysis begins in this manner, thereby bracketing
128 CHAPTER 8

Fig. 8.1 D has asked A to complete the task, which is typical of introductory ecology
courses at the college or undergraduate level.

presumed knowledge, status, institutional relations, and so forth. Comparing turn

01 with the contents of the graphs, we first note that the initial two lines reproduce
the task statement, which A is in the process of re-reading. We then note a phrase
syntactically structured like a question or a statement that also queries its own ac-
curacy: Is this the amount of growth, 20 units, 50 units, and 100 units. In the
context of the preceding turn, turn 02 begins with what comes to reify a {state-
ment-query | evaluation-reply} sequence: Its fine, followed by an offering of
rephrasing: levels, plant growth, and a repetition of the task, it says plant
growth 20, 50, and 100. (The transcription conventions are found in appendix B.)

Fragment 8.2a
01 A: okay. (5.09) discuss the effects of different levels of two nutrients on
each amount of plant growth. (1.58) twenty fifty and a hundred. (14.42)
i:is thIS the amount of growth. (0.28) twenty units, fifty units, and one
hundred units.=
02 D: =um (0.67) ts fine. .hh uh (0.53) levels. (1.45) plan plant
growth yea so it says plant growth twenty fifty and a hundred
03 A: its just showing it moves this way? ((Gestures diagonally from lower
left to upper right in graph (b) of Fig. 8.1.))
04 (0.31)
05 D: um, well, twenty fifty and a hundred are the different (0.72) um (0.91)
theyre the diff (0.57) the (0.32) the the amOUNts of plant growth.
06 (0.77)
07 A: amOUNts of plant gro[wth.]
08 D: [yea.]
09 A: each amount of plant growth. (0.60) [ ok]ay.
 KNOWLEDGE-POWER & INSTITUTIONAL RELATIONS 129

Fig. 8.2 A spatial model of graph (a) in Fig. 8.1.

There is then another {question | reply} sequence, in which the first turn offers a
particular reading of the graph, something (it) moving diagonally, and the reply
offering again what is stated in the task. Although there is no explicit marker of
evaluation, the repetition can be heard as a way of reigning in the preceding alter-
native statement. In turn 07, A does in fact return to the way in which the three
numbers are defined, amounts of plant growth. The pair {turn 07 | turn 08} con-
firms the statement, which is reconfirmed in the pair {turn 08 | turn 09}.
We may gloss this first part of Fragment 8.2 as a way of grasping what the task
is asking generally and the role that amount of plant growth plays in the lan-
guage game offered. The repeated question parts fall to A, whereas the replies and
evaluations fall to D. That is, in contrast to the IRE sequences typical for schools,
the statement-queries here are genuine and it falls to D to assert the correctness of a
statement or to respond. In this fragment, therefore the relation affirms a
knowledge differential, whereby D is knowing more and A is knowing less than
the respective other in as far as this task is concerned.
Turn 09 is followed by a statement of something that might be helpful to A:
reading the task (graph) as providing a growth function g, which is equal to some
g of R1 and R2. Readers familiar with calculus may immediately hear what D is
stating as a functional relation that in written form tends to be expressed as
g = g(R1, R2 ) .1

1
Readers unfamiliar withcalculus may have had difficulties generating hypotheses with
respect to this stretch of talk. This is an important method-related point: Analysts have to be
familiar with the language-game-in-play, that is, the ongoing material activities together
with the language that not only constitutes objects and knowledge but also is the tool for
relating to others. Analysts unfamiliar with the ways of verbally articulating a mathematical
function, therefore, might be passing over this part of the transcription or produce readings
that will turn out to be inappropriate. This actually happened when I originally looked at the
transcriptions together with D. Middleton, a social and discursive psychologist. Unfamiliar
with the mathematical language-game-in-play, he came to different conclusions as to what
130 CHAPTER 8

This statement is followed by a confirmation/acceptance in the form of the affirma-

tive adverb yea. This is followed by another turn that in part restates what has
been said and, in addition, states that g to be a function of R1 and R2 in the way that
we have already heard the preceding statement. Again, there is an confirma-
tion/acceptance turn in the form of the adverb right (turn 15).

Fragment 8.2b
09 A: each amount of plant growth. (0.60) [ ok]ay. [
10 D: [okay] [(0.55) so (0.39) so (0.32)
what might be helpful s (0.63) you have this plant growth function of
whats called GEE (0.65) so GEE is equal to some GEE of r=ONE and r-
TWO.
11 (0.35)
12 A: yea.
13 D: GE is a function of r=ONE and r-two.
14 (0.42)
15 A: right.
16 D: so um
17 (3.20)

In this second part of Fragment 8.2, we therefore have a statement of a mathe-

matical form that is not present in the task. D thereby not only states something in
mathematical form, but also makes known knowledge of this particular aspect of
mathematics and mathematical modeling. In the turns falling to A, this knowledge
statement is uncontested. Contestation being at 0, the knowledge statement comes
to be accepted, at least for the time being. D comes to be constituted as someone
not only knowing about the task statement but also about how to transform the
statement into mathematical form. This transformation is juxtaposed to that offered
in turn 08, which had fallen to A. That is, that earlier statement had not been ac-
cepted, in fact contrasted with the original task statement. In Fragment 8.2b, D
comes to make a statement that offers help, which comes in the form of a mathe-
matical restatement of the task. Turns 12 and 15, both falling to A, acknowledge/
accept the mathematical statements. As before, a knowledge differential comes to
emerge from the joint actions and the level of contestation falling to the acceptance
as opposed to the rejection end. The restatement of the original text in turn 05 de
facto rejects its preceding part, which was an offering of a way of reading the de-
scription of the graph in the task statement.
There is then a long, 3.2-second pause before A, following some hesitating,
long pausing, and audible inspiration and expiration, states not being able to see
this other than as garden flower beds. Brief pauses, an interjection, and an abort-
ed statement ensue before D comes to articulate that it will be a dangerous way.
When considered as the second part of a pair begun by turn 18, we can hear a
statement of what can be seen and an evaluation that this seeing constitutes a dan-
gerous way (turn 23). The next pair constitutes an {evaluation-statement | ac-
knowledgment/acceptance} pair: I know, I know (turn 24), followed by the as-
sertion that this is the way in which it [presumably graph] looks like.

was going on. He heard the participants differently and talking about different things than
those who have familiarity with the institutional characteristics that A and D represent.
 KNOWLEDGE-POWER & INSTITUTIONAL RELATIONS 131

Fragment 8.2c
17 (3.20)
18 A: <<p>um> (0.31) this is (4.66) HHhhh .HH i cA:Nt help but sEEing this
as gA:den flOWer beds (.) frankly
19 (0.27)
20 D: hum
21 A: i mean that's
22 (0.28)
23 D: pdt i think itll be a dANgerous way
24 A: i <<f>knOW i know.> but thats what it looks like to me. (2.74)
<<p>okay.> (2.05) <<f>SO> (0.33) this is the amOUNt of nutrients. the
availability of nutrients r=one. (2.05) and this is the availability of
r=two.

We may gloss this third part of the fragment, as a statement about a way of see-
ing an aspect of the task, an evaluation that this is a dangerous way, and an ac-
ceptance. Again, there is a knowledge differential instantiated according to which
D is in the know and A is not. Both actors contribute to the articulation of the
statement and to its factual nature. Simultaneously, there is a power differential.
For example, turn 23 is the second part of a jointly produced {statement | evalua-
tion} pair, which is followed by an {evaluation | acceptance} pair. The evaluation
turn falls to D, whereas the statement and acceptance turns fall to A. That is, D is
instituted as an individual in power of making an evaluation and A as the individu-
al who has offered something to the evaluator, an evaluation that then is accepted.
Here, the knowledge differential is indissociable from the power differential. The
two are instantiated at the same time and by the irreducibly joint work.
Readers note that we do not have to draw on the institutional relations that tie A
and D. In fact, in the discussion below, the actual relations that tie the two individ-
uals may be surprising to the (unsuspecting) reader. By voluntarily doing the anal-
ysis without further background descriptions, we actually bracket out the institu-
tional positions that the two occupy as explanatory resources focusing instead on
the real relations that they come to produce in the course of doing whatever they
are doing and by engaging in the corresponding verbal exchanges. The transcrip-
tion is a protocol of what they are doing. Rigorous data analysis recovers this do-
ing by bracketing the rapid application of concepts that may turn out to be inap-
propriate, constituting little more than confirmation bias.
Above I state that it is useful to think about relations as struggles (drama) even
when what we commonly consider to be a struggle is invisible or at a zero level.
Taking struggle as a dialectical concept, we can then anticipate the possibility that
the relations of dominance can overturn instantly and, like the relations of
knowledge, perpetually remain open. This also would keep our focus on the joint
nature of the work required rather than thinking in terms of knowledge or power as
something that a person can have, a view that Foucault explicitly rejects as noted in
the introductory section of this chapter. It further appears that A is engaged in do-
ing the task and D is suggesting helpful pointers. The two or differentially posi-
tioned with respect to the task, one apparently trying to solve it but not knowing
how to do so and the other not only knowing what is required but also in the posi-
tion of being able to offer assistance. What constitutes more relevant knowledge in
132 CHAPTER 8

a situation and who is more in the know at that point than others present is the re-
sult of the (ruling) relations and associated struggles rather than being given with
the relative institutional statuses of the participants. Thus, if some person said
What do you mean by . . .? we can hear this not only as a request for elaboration
but also as an evaluation of the preceding turn that is, by this offered question,
qualified as being unintelligible or only partially intelligible. If the query What do
you mean by . . .? falls to an institutionally designated student and the reply to an
institutionally designated teacher, we then come to understand that the teacher, too,
is being evaluated in the situation. We might say that the students qualify/assess
the teachers content pedagogical knowledge, which was insufficient to make a
statement such that the students actually could understand what the teacher meant
to say without actually doing so. To highlight that it is useful to think about every
instant of a jointly produced relation in terms of struggle, we turn to another part of
the meeting that had brought A and D together.

Knowledge-Power Struggles

The Topic. In this part of the meeting, A and D are considering another set of
graphs with associated description and task statement (Fig. 8.3). Two line graphs
can be seen: one depicting death rate in relation to population size N, the other one
showing birthrate. It is another example of a graph that can be found in introducto-
ry biology courses. When the birthrate and death rate graphs intersect, the rates are
in fact equal and, therefore, the net change is zero because there are as many indi-
viduals dying as there are born. That is, to solve the task such that the instructor of
the course from which it was culled would have rated the response as correct, the
two rates have to be compared. When birthrate exceeds death rate, then the popula-
tion is growing. When the birthrate is less than the death rate, then the population
is decreasing in size. This has as the consequence that if the population is smaller
than at the left intersection, that is, if N < NL-INTERSECT, then it will diminish in size
until it has disappeared. (This could be the case when the organisms no longer find
mates.) On the other hand, the population will grow in size when it is to the left of
the right intersect (i.e. N < NR-INTERSECT) and will decrease when it is to the right of
the right intersect (i.e. N > NR-INTERSECT). The left intersect, therefore, constitutes
something like an unstable equilibrium, because the population increases in size
when it is a little larger and collapsing (going to zero) when it is a little smaller. On
the other hand, the right intersect is a stable equilibrium, because to its left the
population will be increasing in size, but when to the right, the population size will
be decreasing.

Transcription and Analysis. Upon reading turn 01, we may note that A is talking
about the graphs, rendering a verbal description of what can be seen on the sheet
(Fig. 8.3). The transcription describes A to hold and lay down the paper and some
gestures associated with the talk. In turn 01, we observe the use of an adverbial
conjunction so following the verbal description, as if an implication was offered
(so they are both . . .). There is another adverbial conjunction so, followed by
 KNOWLEDGE-POWER & INSTITUTIONAL RELATIONS 133

In the derivation of a logistic model, we assume that, as N increased, birthrates decline line-
arly and death rates increase linearly. Now, lets assume that the birthrate follows a quadrat-
ic function (e.g., b = B0 + (kb)N (kc)N2), such that the birth and death rates look like the
figure. Such a function is biologically realistic if, for example, individuals have trouble find-
ing mates when they are at very low density. Discuss the implication of the birth and death
rates in the figure, as regards conservation of such a species. Focus on the birth and death
rates at the two intersection points of the lines, and on what happens to population sizes in
the zones of population size below, between, and above the intersection points.

Fig. 8.3 These graphs relate birthrate and death rate of some population to its size (densi-
ty) N.

the qualifying adverb presumably and an implicative statement that means . . .

There is a pause, and then A speaks again saying Is that right then?, which not
only has the syntax of a question but also is pronounced with a rising intonation as
tend to find in the first part of {question | reply} pairs (turn 03). This is followed
by a longish pause, then a drawn-out interjection um that we often find associat-
ed with hedges, followed by another pause. Then we hear another statement, ap-
parently articulated with rising intonation and, therefore, allowing us to hear the
offer of a question: round this region?

Fragment 8.3
01 A: SO (1.60) ((puts paper down on table)) here ((pencil to the graph)) we
have the (2.11) death rate increasing ((traces death rate)) (0.68) and the
birthrate increasing ((traces birthrate)) and the birthrate is increasing
(0.76) faster (0.87) than the death rate. (1.80) so they are both increasing
but the birthrate is faster increasing than the death rate so presumably
that means that the population is increasing.
02 (0.93)
> 03 A: is that right then?
04 (0.96)
05 D: u:m::
06 (0.41)
> 07 A: round this region?
08 (0.77)
09 D: well yea if you take well birth and death, the birth minus the death
(0.56) well, the birth plus the death which is negative, you are gonna get
something positive (0.73) um growth rate, right?=
10 A: =yea [I::]M looking at
the slopes of the curves [
134 CHAPTER 8

11 D: [so]
[oh okay.

The sequence can be heard as if A took D as not replying to the query in turn
03, for whatever reason this might have been, and then specifying or clarifying
what the preceding query was actually about, namely the correctness (is that right
then? (turn 03) of the statement presumably that means that the population is
increasing? (turn 01). After a pause, D begins to speak making some statement
relating birthrate and death rate (turn 09). Upon first reading, the statement may
sound confusing, and the confusion may persist for many readers upon rereading.
However, having obtained an MSc in physics, this kind of talk is familiar to me
because in that field there are particular conventions concerning the conceptualiza-
tion of quantities. Thus, both birthrate and death rate would be signed quantities.
Birthrate would be a positive quantity; death rate is a negative quantity. When the
two are added, then the effect on the population size is positive in the case of
birthrate, whereas it is negative in the case of death rate. In turn 09, then, we can
hear first a statement of a pertinent issue in terms of the common way of talking
about the situation, taking birthrate and subtracting the death rate to find the over-
all rate of change; we then hear a restatement this time articulating an addition ra-
ther than a subtraction. The subtraction would be typical in the everyday domain
and in fields such as biology and ecology; the addition would be more typical of
conversations in mathematical physics and related fields. In fact, in some situation
D might be heard as showing off, exhibiting the forms of knowledgeability that
generally are associated with a more esoteric academic field and, in so doing, of-
fering up a claim to membership in that field.
This time, the next turn is not of the kind that we observe above, that is, accept-
ing. This time there is an emphatic Im followed by the formulation of an action,
looking at and the articulation of the object of the gaze: the slopes of the
curves. That is, we observe a statement of what to do with the heights of the
curvesi.e. subtract or add themjuxtaposed by what D is looking at, the slopes
of these same curves. This turn comes to be the first part of a {statement | ac-
ceptance/acknowledgment} pair. The second member of the pair, Oh, okay, in
fact begins with an interjection that frequently marks surprise (as well as another
range of emotional states, which, depending on intonation, include frustration, dis-
comfort, disappointment, and hesitation); this is followed by the affirmative and
adverbial interjection okay, that functions as a marker of acknowledgment or
acceptance.
The notion of struggle is useful because we can then view turn 09 and turn 10 as
competing. The former is offered in response to the query whether the earlier im-
plication is right, the latter offered as a statement to the inappropriateness of the
statement because D had looked at something other than is implied by the state-
ment in turn 09. If D is heard as claiming membership in a fieldwhich might be
mathematical physics, given the family resemblance with the discourse of that
fieldD articulates at least a similar claim if not having some superiority in rec-
ognizing the inappropriateness of the claimed knowledgeability to the case at hand.
It is better, therefore, to think in terms of knowledge-power gradients and in terms
 KNOWLEDGE-POWER & INSTITUTIONAL RELATIONS 135

of struggles of participants in locating themselves relative to others: not in absolute

terms but in terms of the issues at hand.
In turn 09, D not only states what needs to be done but co-articulates being
knowledgeable in a particular way, that is, with respect to the addition of two
quantities rather than their subtraction. However, A, too, makes available knowl-
edgeability. Thus, turn 10 marks not only inappropriateness of turn 09 but de facto
constitutes a statement of not only being knowledgeable in the contents of turn 09
but also in a statement of its inappropriateness given what A has offered as solu-
tion in turn 01. In that statement, A constitutes a level of knowledgeability at least
similar to D; perhaps the relative knowledgeability is in favor of A because it also
includes a statement about the appropriateness of turn 09 in the context of what
turn 01 has articulated. D therefore also produces an evaluation (turn 09).
We can hear and see what is going onat least state as hypothesisas struggle.
A apparently offers a solution statement and asks about its correctness. D offers a
description of what is to be done, which is evaluated to be inappropriate given that
A has been looking at something other than what D states. D has offered some-
thing as pertinent that A evaluates as impertinent, an evaluation accepted by A.
Thus, whereas turn 03 can be heard as an invitation of an evaluation, and, there-
fore, the production of a knowledge-power differentialD not only knowing the
answer but also being in a position of evaluating ita reverse knowledge-power
differential is produced with respect to the pertinence of turn 09 to the issues at
hand.
The need to think and use the category of struggle along dialectical lines be-
comes even more clear a little later in the face-to-face meeting of A and D con-
cerning the birthrate and death rate curves. Fragment 8.2 begins with a {requesting
| acceding-to-request} pair of a clue. The request for the clue falls onto A, D is
making a statement about what is confusing A. D formulates that A is thinking or
rather talking about the birthrate as the slopes of the curves seen, which, as those
knowledgeable about calculus know to be different than the value of the curve at
each point.2 That is, in this statement As thinking is formulated as taking the term
birthrate to refer to the slope of the curve labeled birthrate rather than taking it
as an indication of the values of the curve. Turn 02 constitutes an offer of a de-
scription of how A thinks, an offer that comes to be accepted in turn 03, which in
fact overlaps and, in so doing, stops the continuation of the contrastive conjunction
but at the end of turn 02.

2
The following figure shows how the value of a curve at point x differs from the slope at
point x. In this situation, again, hearing the participants in the way they apparently hear each
other requires an understanding of some of the basic principles of calculus.
136 CHAPTER 8

Fragment 8.4
01 A: give me a clue
02 D: yea=i think whats confusing you is you=re thinking (0.40) of (0.34)
you=re tal (0.30) you=re talking about (0.34) the birthrate (0.87) as the
slopes of those um (0.18) curves you are talking about (0.39)
b[ut ]
03 A: [ye]s:
04 (0.22)
05 D: but those curves:: (0.33) are the rates of change.
06 (0.82)
07 A: oh okay
08 (0.28)
09 D: s:o the (0.45) the sl[ope]
10 A: [this] is the rate of change of birth? (0.23) <<p>i
see.>
11 D: the rate of change of tho:se curves would be like (0.23) the
[derivative o:f: ]
12 A: [derivative the second de]rivative, the
[derivative of rate ]
13 D: [they=re both I don=t know if it would be the] second derivative but it
would be the derivative of the rates with respect to the population densi-
ty
14 A: rig[ht so] its actually the second derivative of the po[pulation]
15 D: [so ] [yea ]
16 A: den[sity.]
17 D: [so i ] i think thats key (0.26) ak (0.14) key thing to understa:nd

In turn 05, we find the continuation of turn 02, which now states what I antici-
pate in the preceding paragraph: the curves are the rate of change [of the popula-
tion], which the tasks states to be birthrate and death rate. This turn is the first part
of an {offer | acceptance} pair, which is completed by the interjection marking
(among others) surprise Oh, and the adverb of agreement okay. Turn 09 begins
with the implicative connective so and an indication of what is to come as per-
taining to the slope, but, as turn 10 overlaps, turn 09 comes to an end thereby leav-
ing space for turn 10 to unfold. It is in the form of a statement but intonated as a
question, followed by a marker of insight intonated affirmatively: I see. In that
overlap, the two voices vie for access to the speaking floor and, therefore, consti-
tute an explicit form of competition (struggle). It is not just here that we have to
think struggle dialectically. Instead, every stretch of the verbal exchange is struggle
so that who has a turn may be overturned at every instant. Here, D is losing the
turn to A. The struggle along the knowledge-power line manifests itself in the fol-
lowing lines.
In turn 11, D makes a statement about what the rates of change of these
curvesbirthrate and death ratewould be: the derivative of the curves.3 Turn 12
constitutes a restatement of just what the slopes would be: the second derivative or
derivative of a rate curve. In the pair {turn 11 | turn 12}, we therefore find an offer
of a description together with a counter-offer of a description. In turn 13, which

3
In calculus, the values of the slope of a curve can be plotted giving another curve. This
other curve is referred to as derivative.
 KNOWLEDGE-POWER & INSTITUTIONAL RELATIONS 137

constitutes an uptake of the counter-offer, we first find a formulation of not know-

ing about the second derivative and an uptake in the form of a re-statement of the
second part of turn 12. That is, there is a partial {offer | acceptance} pair. Turn 14
constitutes an implication and evaluative statement so right, and then a restate-
ment of the earlier description of the derivative of rates to be a second derivative.
Turn 17 then constitutes an acceptance of the offer delivered in turns 14 and 16.
Here, turn 11 constitutes not only a description of how to think about the rate of
change of the curves depicted in the task (Fig. 8.3); instead, in introducing a tech-
nical term derivative, it constitutes a marking of membership in a community
where the specialty term has currency. Turn 12 then introduces an even more spe-
cialized term, second derivative, which is a response to the challenge of group
membership. Whereas high school students already are introduced to the notion of
derivative in their first calculus class, the notion of a second derivative is part of a
language-game university students come to play in their advanced calculus and
mathematical physics courses. In turn 13, the speaker (D) acknowledges not know-
ing whether the current topic also constitutes a second derivative, whereas turn
14/16 restates it to be a second derivative.
As a result of the developments, we observe what appears to be a somewhat
strange situation, which we may gloss in the following way. A appears to be asking
for a clue, and D for providing it. D formulates what A is thinking and what is con-
fusing A, as if an instructor was talking to a student. In turn 17, D makes another
statement, as if summarizing the entire fragment: what they have been talking
about is the key to understanding [the graph]. That is, in the statement, D comes to
be positioned as someone knowing what it takes to understand whatever the task
demands: Discuss the implications of the birth and death rates in the figure (Fig.
8.3). In As asking D for a clue, D comes to be positioned as the individual in the
know, and therefore comes to be vested with the power over giving or not giving
the clue requested in turn 01. Thus, on the one hand, we observe the joint work of
investing D with knowledgeability and power. But, on the other hand, we also ob-
serve struggle concerning knowledgeability with respect of taking the derivative of
a rate, or the derivative of a derivative. We may gloss the situation in this way: D
offers the description (turn 11), D marks not knowing about it (turn 13), and A
summarizing that D indeed stated what the second derivative is about (turn 14/16).
Thus, D appears to have the upper hand with respect to the discourse about second
derivatives; and this upper hand comes to be jointly achieved.
The results of these analyses therefore suggest that there is not some clear
knowledge-power differential that is produced and maintained. Instead, our anal-
yses become more rigorous, better adapted to the data, and less coarse if struggle
and knowledge-power are approached dialectically. This allows any instant of
practice to overturn and become a different practice and any instant of knowledge-
power to overturn and invest someone else has differentially more invested than
others with respect to knowledge-power. The most rigorous forms of analyses
begin by imposing the least constraints and presuppositions and then work out the
work being done to establish, produce and transform, any differential along what-
ever line of interest. This is equivalent to stating that we allow every common and
scientific notion to be bracketed and, thereby, it is equivalent to preventing all-too-
quick classifications and explanations of the events that we employ to understand
138 CHAPTER 8

in and by means of the analytic work that we conduct. We might therefore begin
with the less stringent assumption that there is struggle everywhere; and when it is
not apparent, we take it to be invisible, as if there were tacit agreements for taking
and accepting certain knowledge-power differentials, at least for the time being.
Struggle, disorder, is less stringent than order; contestation of and for knowledge-
power is less stringent than assuming (permanent) order.

Ethnography of Institutional (Ruling) Relations

By now readers may be very interested in finding out more about A and D, their
institutional relations, and what brought them together. Even though I had known
about what I reveal here before conducting the analysis, bracketing meant not al-
lowing this knowledge to enter and thereby affect the analysis and what we pub-
lished about the event that had brought A and D together (Roth and Middleton
2006).
Daniel, an undergraduate student with physics and anthropology majors, was
completing an internship in my research laboratory generally and in a project on
scientists use of graphs specifically. In that project, we asked scientists to think
aloud doing a range of tasks from undergraduate biology (ecology) courses and
equivalent graphs from physics. Daniel had recruited 21 individuals from the phys-
ics department, most being faculty members but also including some postdoctoral
fellows, to serve as experts. The research design called for expert think-aloud pro-
tocols. One of the physicists was Anne, a faculty member with over 30 years of
teaching experience recognized for her excellence in her teaching awards.
This constellation provides an interesting situation. Institutionally, Anne is
higher up the ladder and, as faculty member, is positioned differently from Dan-
iel, who not only is (but) an undergraduate student but also not among the higher
achieving ones in that department. From that institutional perspective, therefore,
Anne would have to be considered being in the know and in terms of the
knowledge-power relations superior to Daniel. Why then would Anne ask Daniel
for hints, clues, and evaluations? She has had a long history of teaching, had ob-
tained a PhD, and conducted research that was published in the acknowledged
peer-reviewed journals of their field. At the time, Daniel was only in his third year
and not among the stronger students in the physics program based on his course
grades. Anne has accepted the invitation to serve as an expert on graphing. And
yet: she is asking Daniel in a way that provides him the upper hand along the
knowledge-power differential.
In addition to their formal institutional relations, there is a second one. The two
are actually part of a different form of activity: research on graphs and graphing.
Daniel, serving as a research assistant currently collecting data, may be assumed
rightly or wronglyto know what the tasks are asking for. Under this assumption,
then, he would have the upper hand, at least with respect to some, along the
knowledge-power differential.
The preceding analysis shows that neither one of these descriptions would make
for good starting points to hypothesize how the meeting would unfold. Using these
 KNOWLEDGE-POWER & INSTITUTIONAL RELATIONS 139

differential positions a posteriori to explain what happened simply is bad science.

If these distinctionsstudent : teacher or researcher : researchedare to have any
explanatory power, then the ought to be used to predict how the meeting will un-
fold. This, however, as the analysis shows, is quite impossible. As the physics pro-
fessor who also acts as an expert on graphing, Anne ought to have the upper hand
with respect to knowledge-power. On the other hand, as a researcher, Daniel might
be granted to have the upper hand at least situationally. But the fact that Anne has
been invited as an expert, it is still possible to anticipate greater knowledgeability
on her part when compared to a research assistant. This is the case when psycholo-
gists ask historians to interpret historical texts or when cognitive scientists ask a
Nobel laureate of economics to interpret an economics graph. Saying after the fact
that Daniel is more knowledgeable because he is a research assistant whereas Anne
does not know the responses because she is the person under examination does not
have any explanatory power. Either we accept the causeeffect figure as explana-
tion, then the causes have to be stated beforehand; or we accept the critique of the
causeeffect figure, which means causes, because identifiable only as a result of
the effects, must not be used at all. We must not say that Anne wins over the con-
test over the second derivative because she is a physics professor who would know
more about this topic than the undergraduate student only to say that Daniel is as-
sumed to know more or all about the tasks because he is a research assistant. In any
event, whatever partial knowledge he might have exhibited, it had emerged in ver-
bal exchanges with me, who had explained to him the tasks and how to read them.
This data is nice and instructive because it allows us to understand why institu-
tional positions have little explanatory value. Explanations work because they may
function as predictors for future events. The events described could not have been
predicted along either of the institutional relations. Using these relations after the
fact is but another manifestation of a phenomenon that in sports is referred to as
Monday morning quarterbacking. The approach exhibited by the instructor-analyst,
an approach further articulated, exemplified, and described in the other parts of this
book leads us to a more rigorous approach: reconstructing the situation based on
the analysis of the real relations and differential positioning with respect to
knowledge-power as the result of the relations rather than as its antecedent.

Returning the Horse in Front of the Cart

The proposal here is instead that it is the workings of the phenomenon that
exhibit among its other details the population that staffs it. (Garfinkel 1996,
p. 5)
In the preceding section, we see continuous development emerging from the give-
and-take relation, which, as one of its results, differentially positions the partici-
pants. It is not that A or D position themselves in one or another way. What we are
witnessing is in fact irreducible, social and joint work so that both individuals are
positioning and positioned. In other words, rather than the individuals doing work
to bring about the phenomenon, it is, as the introductory quotation suggests, the
140 CHAPTER 8

phenomenon that exhibits the particulars details of the population that staffs it. It is
the IRE phenomenon that exhibits an asymmetrical distribution of turn taking, the
differences between teacher and student; and the phenomenon of knowledge-power
relations exhibits the details of the population that staffs it: researcher-researched,
professor/teacher-student, school principal-teacher. This is so because the social,
joint work is in excess of individual contributions. We not only contribute to but
also are conditioned by social, joint work. This is why it makes sense to think par-
ticipation dialectically (e.g., Goulart and Roth 2006; Lave and Wenger 1991).
Thus, to turn to an educational example, if learning is the topic then it behooves the
researcher to show where, what, and when learning occurs and on whose part. Us-
ing institutional positions such as student and teacher to point at relations and sug-
gesting that the student learns and the teacher teaches is using what is to be ex-
plained (explanandum) as a resource in and for explaining (explanans). Much of
the research that we can find reported in the literature constitutes little more than
Monday morning quarterbacking: teachers teach because they are teachers, learn-
ers learn because they are learners. The real challenge is to work out in and
through analytic detail, how teaching and learning occur in the joint work of peo-
ple and the form in which their relations are realized. To work out what is happen-
ing without subjecting oneself to preconstructions or to Monday morning quarter-
backing requires rigorous data analysis that bracketsto the extent that this is
possibleconcepts that are used for explanatory purposes. If there are concepts to
be used as explanations, then these should allow predicting any situation, recorded
in the past or anticipated to occur in the future for which the analyst does not yet
know the actual outcome.
To constraint of not knowing the actual outcome may be difficult to implement.
Generally, researchers know the outcomes of the events that they record and ob-
serve. In this situation, adopting the use of a first-time-through policy is advisable.
First-time-through means that the analyst never allows something known to partic-
ipants only at some later stage to be imported to the analysis of an earlier stage of
the event. Thus, even though we might know before the events that led to Tran-
script 4 (Mikela) that the teacher is the well-known childhood educator Vivian
Paley in the attempt to model her approach to narrative and story telling, we ex-
clude it as a causal explanatory precedent. This then allows us to understand that
sometimes a lesson might not turn out so well. We will then not find ourselves in
the position to have to try to explain why we observed one of Paleys lessons that
did not work out as intendedwhich someone might be tempted to do by saying,
again after the fact, that she had a bad day, did not sleep well the night before, or
whatever else people say (after the fact) in the attempt to explain why something
did not turn out the way they expected. Bracketing this knowledge, which was the
point of departure in chapter 5. Here, the instructor-analyst did not know Vivian
Paley or that the P: in the transcription had the institutional role of a teacher or
that this teacher role was occupied by Vivian Paley.
9

On the Shop Floor and Playing Field

The relation that establishes itself between observer and observed is a special
case of the relation between knowing and doing, between interpretation and
use, between symbolic mastery and practical mastery, between logical logic .
. . and the universally prelogical logic of praxis. (Bourdieu 1980, p. 37, orig-
inal emphasis, underline added)
In the introductory quotation, a difference is apparent between (the theorists) in-
terpretation and the practitioners doing, between a symbolic grasp of social action
and practical mastery in social situation, between the logic of the theorist and the
logic of the practitioner. We observe the difference everyday, for example, in the
proper speaking of a language on the part of individualse.g., young children and
many working class peoplewho do not know or know (very) little grammar.
Here, individuals exhibit competent practical linguistic performance without hav-
ing a theoryi.e., the rules of syntax and semantics. The converse also can be
found daily, for example, in (sports) journalists or sports fans who do not know to
play more than at rudimentary levels and yet comment on and often denigrate the
performances of professional and highly trained amateur athletes. Whereas such
individuals often exhibit symbolic mastery they lack to practical mastery required
from the participants.
The relationship between the analyst-theorist and the participants in social re-
search tends to be practiced along the line of differences between theorists and
practitioners. A first indication of this can be found in the methodological specifi-
cations that authors are required to produce to ascertain the reproducibility of the
findings. The very fact that research methods have to be specified points us to their
difference from the normal everyday ways of acting that is required in the social
situations that analyst-theorists describe. If the methods of producing social struc-
ture and describing/theorizing social structure were the same, then restating them
in the methods section of an article would be redundant and odd. The difference
between these two types of methods has been made thematic in the difference be-
tween formal analytic methods and ethnomethods (Garfinkel 1996). Although
142 CHAPTER 9

Bourdieu has been quite critical of ethnomethodology (e.g., Bourdieu 1992), the
fundamental distinction at the heart of the two projects appears to be the same. The
distinction between formal analytic methods and ethnomethods does not inherently
say that the former is worse or deficient with respect to the latter. Ethnomethodol-
ogists do recognize that formal methods have described and provided instructions
how to see a vast range of phenomena of order. Instead, ethnomethodology asks
and proposes to work out what more there is to ordered and orderly phenomena
that formal analytic methodswhether qualitative or quantitative in nature
describe. Thus, [ethnomethodology] asks what more is there that users of formal
analysis know and demand the existence of, that [formal analysis] depends upon
the existence of for [formal analysis]s worksite-specific achievements in carefully
instructed procedures (Garfinkel 1996, p. 6). This statement, which many readers
might experience as dense, requires perhaps some unpacking.
The important aspect of the statement is the suggestion that users of formal
analysis appear to know more, in fact, that they presuppose and demand existence
thereof, and upon which formal analysis depends to produceusing the carefully
described methodsresearch findings. Take the example of a queue. A social sci-
entist studying queues has to have the (practical) competence of recognizing a
queue and distinguishing it from other forms social gatherings. But the people lin-
ing up to purchase movie tickets do form the queue, recognize someone who at-
tempts to squeeze in ahead, know when the next in line apparently fails to see that
it is his/her turn, and so on. That is, the researcher has to have the same competen-
cies for lining up in a queue as those whose queuing is studied. That is, whatever
qualitative or quantitative method the researcher is using, the everyday methods
underlying the knowledgeable production of queues is presupposed and even re-
quired. There are studies precisely exhibiting those presupposed competencies at
work.
The first example is a study of graduate sociology students asked to code the
entry of the records of a psychiatric outpatient clinic (Garfinkel 1967). The purpose
of the study was to find out by what criteria treatment applicants were selected. A
traditional interrater reliability procedure had been performed to establish con-
sistency in the coding.1 It was noted early on that to accomplish the coding, cod-
ers were assuming knowledge of the very organized ways of the clinic that their
coding procedures were intended to produce descriptions of (p. 20). That is, the
coding process had as a prerequisite the very practical competencies that were to
be the research outcomes. The second example is a detailed study of natural scien-
tists in the course of producing an explanation for the shape of the graph; they had
to draw on the very understanding of biology that their study was supposed to pro-
duce (Roth 2013c). Without recontextualization of their work in the very settings
where their specimen derived from, scientists struggled producing an interpretation
of the results that emerged from their study. A familiarity with the original settings
from which the data derive, and with the totality of the transformational trajectory
(their work!) that produces the graphs that they obtain, scientists are unable to pro-

1
To establish interrater reliability, researchers compare the ways in which two (trained)
individuals code the same data. Interrater reliability might then be stated in the form of a
percentage of cases where the raters agreed.
 SHOP FLOOR AND PLAYING FIELD 143

duce explanations. Thus, experienced research scientists have been shown to fail
producing explanations of graphs even when these were culled from introductory
university courses of their own field (Roth 2012; Roth and Bowen 2003). Both
examples exhibit the circularity of the research activity.
There are different ways to distinguish the different orientations that practition-
er-participants and analysts have. Some conceive this difference in terms of games,
where those on the playing field are subject to different logic and temporality than
the removed researcher-analyst, who, symbolically making events present again
representing themhas all the time required for a thorough analysis that is una-
vailable to the player during the game (Bourdieu 1980). Others distinguish what
happens on the shop floor where social order is produced versus shop floor theoriz-
ing (Garfinkel 1996). Others again write about the difference in terms of observing
street life through a window (from above street level) and stepping out of the door
and becoming part of street life (Ingold 2011). I take rigor to pertain to the ethno-
graphic adequacy of the researchers account of what is happening on the shop
floor, playing field, or street:
The ethnographers adequate account of what natives do together must fol-
low from the way in which the natives structure a situation to allow their par-
ticipation with each other from one moment to the next. The ethnographer
must articulate the same hesitant and momentary contexts that the natives are
displaying to each other and using to organize their concerted behavior.
(McDermott et al. 1976, p. 246)
These authors suggest four criteria for descriptive adequacy: (a) members to the
setting tend to articulate, in situ, one or another aspect of the context of and for
their behavior; (b) members tend to organize their bodily orientation and move-
ments such as to configure or position what they are communicating, often without
actually saying so; (c) members orient to the order produced in and through their
actions, and they make this order available to each other as part of the ongoing
affairs; and (d) members not only provide accounts of their actions but also hold
each other accountable for what they do, and they do so in ways that are consistent
with the ongoing activity and the particularities of the setting. Rigorous analysis
attends to these, the members ways, which therefore leads to descriptions what
members recognizably and accountably do and how they produce recognizability
and accountability in doing it. Ethnographic adequacy takes us to the shop floor
where social reality is produced, making the analyst look at what is happening
through the eyes of the participants. Ethnographic adequacy also recognizes that
the researcher has to have the same competencies as the members and use the same
methods to see occurring what the members themselves perceive.

On the Shop Floor of Social Order

Much of research is concerned with the meaning that social actors are said make
while participating in whatever societally organized setting in which they take part.
However, this (subjective, individual, socially constructed) meaning is extracted
144 CHAPTER 9

from the data using special methods, specified in the methods section. There ap-
pears to be a contradiction in the sense that (a) this meaning is held to be subjec-
tive, negotiated, or constructed by the participants, who, apparently, produce
knowledgeability for each other to be recognized as such and (b) special methods
are required for exhibiting it to the readers of the article. This way of doing re-
search and data analysis is not as rigorous as the methods sections of journal arti-
cles make it out to be. This is in fact another manifestation of the circularity de-
scribed above: research exhibits order that is a prerequisite for the process of
exhibiting. We may change our research orientation and analytically place our-
selves on the shop floor of social order, presupposing nothing but the competencies
in the same practices that the participants exhibit. The difference with other re-
search lies in the orientation to the work that produces order rather than to the order
that is the result of the work. Rather than describing the order of a queue, its be-
ginning and end, the by-passers, the next turn, and so on, the living work itself is of
interest, the methods by means of which we, members of (immortal) society, staff
the relevant social phenomena (e.g., a queue). One such phenomenon is the joke.

Production and Organization of a Joke

It does not matter what the research-analyst thinks or interprets: the only real ques-
tion is what the people in the situation themselves do or not to that leads them to
laugh and describe what has been said as funny or not. That is, rather than situating
ourselves like the manager observing the movements on the shop floor, we posi-
tion ourselves on the shop floor where the action is occurring. We throw ourselves
into the live event, becoming by-standing participants witnessing whatever is hap-
pening. Consider the ending of the fragment with Heidi and David (Suzuki) tran-
scribed to include pauses and prosodic information. (For transcription conventions
see appendix B.)

Fragment 9.1
21 Heidi: frED IS nATurally going to erode away, bUT (0.45) if he EVER lost his
hump. (0.59) ALL we=do is CHAnge his name to hUMp=frEE the cam-
el.
21a (0.62)
22 David: [gnaw]wwwwwooooooo
[(holds nose to produce something like the sound from a saxophone or a
snort of some animal))
23 Heidi: [a.H <<dim>HA HA HA HA HA HA HA> ha ha ha .HEE]
[ ((3.05 seconds)) ]
23a (1.70)
24 Amanda: [ha ha ]
25 Michael: [ha ha ]
26 Ashley: [huh ha]
26a (0.31)
27 Heidi: sometimes it takes a while. .H hm hm hm hm
 SHOP FLOOR AND PLAYING FIELD 145

In this situation, we may gloss what has happened as the joint making of a joke
and the different participants staff this phenomenon. We may gloss the event (cur-
sorily) by saying that in turn 21, we have the offer to a joke. There is a pause. Then
David produces a sound together with body movements that acknowledges and at
least partially accepts the offer in a way that may also be heard as an evaluation of
a contorted joke. It is Heidi who then laughs extendedlythe transcription shows
that it is for over 3 seconds (turn 23). Then there is another longer pause before the
three children laugh. As the video image shows, it is a tentative rather than loud
laughter. Finally, there is a turn that we may hear as an evaluation that might be
glossed as a slow uptake of the joke. This gloss will allow reader to recognize
the phenomenon that is quite common: something offered as a joke takes some
time to be recognized as such by the recipients with the evaluation on the part of
the individual who made the offer, marking and commenting on the slow uptake.
In this situation, we do not need to interpret or define the methods for doing so.
The participants themselves staff the production of the phenomenon. By focusing
on the joint workoffering, slow reception, and evaluation of the slow recep-
tionwe actually situate ourselves right where this production occurs. Readers
will note that this phenomenon differs from the way in which professional comedi-
ans participate, who tend not to contribute in a way that we might gloss as laugh-
ing at ones own joke. If researchers were to investigate the function of joking and
laughing in a particular context, they would then seek to identify variations. Such
variations might be situations glossed by the terms this is not a joke, not getting
a joke, or a distasteful joke. We may frame what is happening in terms of a
formalism that makes thematic the work of practical action and the ways in which
it is, or may be, properly glossed by participant (Garfinkel and Sacks 1986). We
may then write the shop floor production, as viewed from the shop floor, as doing
[. . .], where the bracket encloses the gloss. Thus, the foregoing glosses would
then be investigated in terms of doing [a funny joke], doing [a distasteful
joke], or doing [not getting the joke]. Whatever is happening, participants ob-
serve it and make it happening; they may report it, and may, but do not have to do
so by means of the text that appears in the bracket. Thus, Davids sounds and ges-
tures is an alternative way of making available to others his evaluation. It is open
and undetermined because no further evidence exists for the uptake in that situa-
tion. The researcher attempting to provide an interpretation of what David might
have wanted to express would be on shaky grounds and not conducting rigorous
data analysis. But the statement sometimes it takes a while is a description and
evaluation of what has taken place. It is a formulation of the phenomenon as one
where an offer to a joke takes some time to be recognized as such: doing [telling a
joke that takes some time to be recognized].
To exemplify how researchers might approach data analysis that focuses on
joking and laughing, I reproduce a slightly modified analysis of laughter in science
classroom (Roth et al. 2011) and provide a running commentary about what the
analysts are doing in their text, thereby producing a similarly structured text as in
part B. It is but one example that we may find in the literature where researchers
exhibit a pragmatic orientation in their focus on the function of jokes and laughter
in specific settings, such as the organization of an emergent ethnic joke in a profes-
sional meeting (Markaki et al. 2010). The analysis of the science classroom epi-
146 CHAPTER 9

sode below was done to describe a phenomenon of laughter that has a particular
function in the reproduction of science as serious business. The phenomenon there-
fore is a societal one, and the individuals appearing constitute the staff or produc-
tion cohort. The analysis situates itself on the production floor, following what the
actors do to recognizably produce the phenomenon. The analysis reports on how
this phenomenon is organized and, therefore, how it can be recognized as a societal
phenomenon. The analysis exemplifies a rigorous approach because it does not
impose interpretations from the outside, one among many possible ones, but de-
scribes the phenomenon staffed by the specific people, the teacher Victoria and her
students. That is, the analysis does not seek to get into the heads of Victoria or her
students. Instead, it works out the structure of the phenomenon and its operation.
The analysis then shows how science as serious business is reproduced in and
through the inversion that it undergoes when it momentarily becomes a laughing
matter. Joking and laughing are better approached as holistic social phenomena,
the result of irreducible joint work, in which participants find being (unwitting)
part. The very point of joking is its unexpected nature. An ethnographically ade-
quate analysis has to work out what arises from this temporal nature of joking and
laughing, which in part derives from the very social nature of the phenomenon
rather than reducing social events to the individual minds.

Dialectics of Laughter in an Extended IRE Sequence

An important interactional form that teachers use to reproduce schooling gener-

ally and the differential institutional knowledge-power relations specifically fol-
lows a particular turn-taking routine: the teacher initiates the turn with a ques-
tion, a student responds, and the teacher evaluates the response.2 This sequence,
the interactional function of which is control and the reification of a positiv-
istic conception of knowledge, is generally referred to by the acronym IRE
(i.e., initiation, reply, evaluation). Control is no laughing matter: The tenants of
the old truth and power tend to be gloomily serious . . . they do not know how
nor do they want to laugh (Bakhtine 1970, p. 213). The IRE sequence tends to
be serious business because it asserts that the teacher already has the answer,
which students often do not know; and the sequence asserts not only who is in
the know but also who is in the position to provide the evaluation of
knowledge.3 Not yet investigated has been the question whether laughter, which

2
In this opening phrase, the text sets the reader up to approach a particular turn-
taking routine that has a recognizable pattern typically found in schools but com-
mon in other situations as well, such as parent-child relations.
3
The two phrases set up where the research findings will locate themselves. On the
one hand, science is serious, and this seriousness is reproduced in the very way it
is taught, that is, by emphasizingin contrast to some other school subjectsthe
scientifically correct ways of doing and thinking the subject. On the other hand,
laughter can be observed in school science classrooms. There is therefore a possi-
ble tension between the seriousness of science and the laughter that nevertheless is
 SHOP FLOOR AND PLAYING FIELD 147

is a way of undermining and inverting existing structures of knowledge and

power, has a function in the reproduction and transformation of the IRE struc-
ture.4 In the following, I produce an extended analysis of one IRE episode, iden-
tified as an episode by the teacher who initiates a review for all those students
who had been absent during the previous introductory lesson to the unit. The ep-
isode is completed with the review and when the teacher begins to introduce
students into the present lesson.
The IRE exchange begins in turn 06, when Victoria formulates what is to
come as to refresh the memory of those people who were away followed by
the question to be answered What is engineering, who are engineers? The epi-
sode ends when Victoria evaluates what turns out to be the final answer as
beautiful, then flags a summary to come by saying okay, so, which is fol-
lowed by the summary proper, engineers are very cool people. In turn 41, she
gives an example of the more efficient that she has earlier articulated. In this
episode, there are several answers. Victoria deals with them in different ways,
provides different kinds of evaluations. In turn 12, she repeats Coreys answer
with falling intonation like a statement, followed by the evaluative term beauti-
ful, so that the answer thereby is evaluated positively. She then solicits another
response from Terra. Victoria repeats his answer, they are incredibly good
looking, grins, and laughs, referring to an instant where this has been stated be-
fore. The following account breaks the IRE sequence into four fragments, iden-
tified by the markers that the participants themselves use: The teacher names or

observed. The text communicates that the author is familiar with the analyses that
the Russian literary scholar Mikhail Bakhtin has conducted on the function of
laughter during the Renaissance generally and as it appears in the writings of the
French author Franois Rabelais in particular. Bakhtin shows how laughter and
carnival constitute both an overturning of the normal, serious order and the reas-
sertion of this order by the limitation of laughter to particular dates in the year
(e.g. fairs and carnival). There is therefore a historical precedent that this study
draws and capitalizes on.
4
Here the text states the white spot on the research map and, therefore, where
and in which way it contributes to the science of science education. Rather than
just naming a situation as IRE, which is what much research does, this text de-
scribes the work of doing IRE, that is, the relational work people do that profes-
sional analystslinguists, educators, social psychologisthave come to denote by
that term. The people themselves do not talk about IRE while doing what they do.
Moreover, they do not explain what they do in terms of the IRE concept. Thus, for
example, Victoria might say afterwards that she was reviewing what we had
talked about the week before; we might then characterize their practice as doing
[reviewing what we had talked about the week before]. In this way, what the par-
ticipants in this science classroom are doing is recognizable by the members to the
setting, Victoria and her students. On the other hand, these members will likely
find non-intelligible when researchers tell them that they were doing [IRE]. This
latter is the account that the researchers provide within their community, with a
decided focus on the same work but glossed in a different way, within the context
of a different community.
148 CHAPTER 9

points to a student, which marks the beginning of a fragment, and, by naming or

pointing to another student, transitions to another lesson fragment.

Reviewing and the Setting up of the IRE. The total episode has the IRE struc-
ture, with a few additional turns until the anticipated/expected response has
been provided. During the episode, there are instances of student contributions
that give rise to laughter on the student and on the teacher part, that is, some of
these instances are student initiated, others are teacher initiated. However, the
teacher-initiated instance comes after the IRE sequence, after the teacher has
summarized the exchange in her words and gone on to lead to the next issue in
the unfolding of the curriculum.
Fragment 9.2a
01 V: ((questioning look)) okay (0.72) we talked about what
is engineering lASt week DIdnt we.
02 L: uh hm.
03 V: ye[ss;]
04 S: [yea];
05 (0.67)
06 V: to refrESh the memory of thOSe people who were away:.
(1.22) whAT is engineering; who are engineers.
07 (0.60)
> 08 T: <<p>wes>
09 V: what type of <<flicker of a
smile>pEOple>. yes krissy.
10 K: people who build stuff and desIGn
stuff.
11 (0.42)
12 V: people who bUIld stuff and desIGn
stuff; bEAUtiful ter[ra]? *

In the first fragment of this episodeitself defined by the teacher who an-
nounces a review of what the lesson has covered before, to refresh the memory
of those who were away to the point of summarizing what the students have
said here and in the preceding lesson. There is a first offer of an answer, Wes
(turn 08), the name of Victorias husband, which is not taken up. Whereas Wes
de facto is an engineer, so that his name is a correct answer to the question,
Who are engineers? it may not be such here, not as a possible response to the
other question offered twice, What is engineering? Rather, Victoria makes
another offer, What type of people? (turn 9). Here, the request is made to
name a type of people rather than a specific person. It can be heard as a rebuke
of the previous utterance, an indication that it is inappropriate. The next student
utterance, then, reifies a questionanswer sequence, People who build stuff and
design stuff (turn 10).5 Victoria repeats what the student has said and then ut-
5
In this writing, the text refers to (e.g. by naming) individual actors: Victoria or
one or more students. However, this is not mean to suggest that the social relation
can be constituted by departing from the individual actor, who generally is taken
to be the element that constitutes the relation. Instead, throughout the analysis, the
focus is on joking and laughing as irreducible social, jointly produced phenomena.
Victoria, Terra, Prunella, Irisana, Krissy and others are but the individuals that
 SHOP FLOOR AND PLAYING FIELD 149

ters an evaluative term, beautiful (turn 10). In this case, although it might ap-
pear that she produces a mere repetition, pure repetition does not exist in lan-
guage and every repetition constitutes difference and has a function.6 In the pre-
sent instance, the student utterance is produced with descending pitch, which
marks it as an offer for a declarative statement. Victorias utterance also is
marked by a descending pitchthough it descends less strongly than that of the
students. It therefore follows the pitch contour, affirming the declarative nature
rather than having a questioning intonation. Up to this point, we observe a typi-
cal IRE sequence.7 The teacher initiates by offering a question, a student reifies
the question with a response offer, and the teacher produces an evaluative term.8

currently staff the phenomena described. Whether anything comes of as a joking or

laughing matter is not the result of an individual intention. It is part of a jointly
produced and jointly suffered situationVictoria, according to her own account,
has never intended to have her husband brought in at that place in the classroom
talk; and students could not have anticipated that there would be an offer of a
question in reply to which the reference to Victorias husband was possible.
6
At a minimum, saying a word a second, third, . . . or nth time makes it appear
against previous instantiations. It does work by the very fact that it appears again.
If we just analyze language in terms of meanings, then saying the same word
again and again would constitute something like mere repetition. There is, howev-
er, more to a word than its dictionary sense (Voloinov 1930; Vygotskij, 2005).
Both these authors use a short narrative from Dostoyevskys diary of a writer, in
which the novelist tells about overhearing six drunken workmen articulating an
obscenity six times in sequence. Not only does the word change in its prosody,
Dostoyevsky was attributing different evaluative content to these articulations. I
recently analyzed a classroom episode where physics students articulated the word
penis 10 times in sequence, each time with a change in prosodic clothing. Science
words, too, come to be articulated repeatedly, often without change, and without
other words added to make some complete statements (Roth 2014c). Of course, it
would make no sense to analyze such a sequence in terms of the dictionary sense of
the word penis. Instead, it is more useful to think about what happened as a game
of language, which, with Wittgenstein (1997), I understand as language together
with the practical activity to which language belongs. The two, practical activity
and its language, form an irreducible whole.
7
Here, the we evidently includes researchers and readers; it does not include,
as described above, the social actors themselves. They do whatever Victoria has
formulated, that is, refreshing their memory by talking about what they talked
about the previous week. This particular article, therefore, navigateswithout
stating so explicitlybetween the two accounts, one in which the social actors
would recognize themselves and their doing and the one directed towards a read-
ership interested in more theoretical descriptions. But in its particular construc-
tion, the authors describe what a researcher familiar with the discourse of IRE
would find as the actual practices in which people participate with the words and
language that they are familiar with. The paper therefore covers a middle ground,
between (a) a description of the work that could be read as an instruction for find-
ing or learning about the work in a science classroom and (b) a theoretical ac-
150 CHAPTER 9

Engineers are Incredibly Good Looking: A Side Comment? Victoria then so-
licits another answer by calling on Terra. The student begins to offer a response
even before Victoria has finished uttering his name, yea, they are incredibly
good looking (turn 13). It is clearly an utterance that completes a question
response sequence, as it completes the logic of the query for a type of people.
But it is a special type of response, as seen from the fact that it in turn is greeted
by and therefore responded to with laughter, both on the part of fellow students
(turn 15) as by Victoria herself. Up to this point, the teacher has had a serious
facial expression (turn 12). Right toward the end of Terras utterance, her eyes
move upward and she begins to grin with an expression that might be seen as
saying, Oh here he goes again (turn 13). She then breaks out in full laughter at
the end of uttering what can be heard as an acknowledgment that they did men-
tion that they [engineers] were incredibly good looking and while soliciting the
next student for making another offer to complete a questionresponse turn.

Fragment 9.2b
12 V: people who build stuff and design
stuff; beautiful
tER[ra ]? *

count common to research communities familiar with the classical formal analyt-
ic ways of doing research. To take up on the ethnomethodological articulation of
the issue at hand, formal analytic research has and continues to describe phenom-
ena of order (Garfinkel 1996), such as the IRE sequence: Phenomena made in-
structably observable in formal analytic details of concertedly recurrent achieve-
ments of practical actions are so provided for by [formal analysis] that a
phenomenon, whatever the phenomenon and whatever its scale, is made instructa-
bly observable as the work of a population that staffs its production (p. 5). This
paper is not interested in IRE, which is a term that it brackets, but in specifying
just what people are doing what researchers gloss as IRE.
8
This latter phrase is what Garfinkel (1996) refers to as a form of instruction that
researchers provide to each other, in addition to specifying any method for de-
scribing the analytic process, to find in a classroom precisely what the term is in-
tended to describe. Even though teacher-Initiation, student-Reply, and teacher
Evaluation might appear to be self-descriptive of what can be found, the difficul-
ties novice researchers experience to identify such sequences in actual classroom
talk should be sufficient as a hint that such self-evidence is only presupposed. I had
a number of graduate students interested in doing conversation analysis or in fol-
lowing the precepts of discursive psychology, who asked me whether this or that
fragment or videotape did indeed constitute what researchers call an IRE se-
quence.
 SHOP FLOOR AND PLAYING FIELD 151

> 13 T: [yea] they=re incrEDibly good

looking.
((* grin, then facial expression
gloss: oh, here he goes again
, gaze upward, slow lid clos-
ing))

14 (0.20)
> 15 S: .hh <<f>heh> ((others laugh,
too))
> 16 V: <<p, grinning>we did mention they
were incredibly good looking.>
<<laughing * >

pra(h)sa(.h)di(hh[h)> *

Here, Victoria begins to grin at what will be the very end of Terras response,
itself joined by a students in-breath followed by an out-breath after the fact can
be recognized as the beginning of laughter, when several other students join in.
The teachers grin and the first students sounds can then be heard as an invita-
tion to laugh, which, first, several students accept. This is an invitation to further
laughter, which Victoria accepts in her turn (turn 16).9 In this situation, Terra
has made an offer of a {question | answer} completion, but, whereas logically
correct, it has not led to the evaluative comment that the teacher utters in other
situations, here and in other lessons. Instead, the response has led to laughter,
which thereby has punctuated the otherwise serious nature of the lesson. In a
way, the offer is a challenge to the seriousness of the IRE sequence that is in the
process of unfolding. The structure is such that the challenge arises by being
both correct and incorrect simultaneously.10 It is logically correct as an answer
and yet clearly not anticipated as the answer, creating an effect that changes the
very nature of the lesson at this instance, as indicated in the emotional response
to which the utterance has given rise. One of the ways in which comedy, the

9
Readers clearly note the orientation of the text to turn pairs, taking a statement to
be an invitation because there is an acceptance, and taking a statement to be an
acceptance because there is an invitation.
10
The text uses dialectical reasoning, where something is both correct and incor-
rect simultaneouslynot because different participants can interpret something
differently but because it differs from itself. It differs because, as everything else
alive, it continuously changes.
152 CHAPTER 9

joke, functions is that it rides on the undecidable nature between the correct and
the incorrect nature of the answer to the response.11
There is another indeterminable aspect here, the one between the conceptual
and the bodily. On the one hand, we have discourse concerned with the concep-
tual issue. Victoria has offered a question to be completed by the naming of
types of engineers. Such naming would normally be analyzed in terms of its
cognitive aspects. Yet in the present situation, we hear and see responses on the
part of fellow students and the teacher that are not cognitive at all, the facial ex-
pression changes and there is a bodily production of laughter (rapid movement
of the diaphragm, cackle, punctuated taking in or releasing of breath [turn
15]).12 This {question | response} pair is both questioning (undermining) and re-
ifying the IRE sequence.13 There is a teacher evaluation, which here comes in
the form of laughter that follows, as previously, a repetition of what the student
has said and a formulation that they have mentioned this fact before. The very
structure of the IRE has been asserted all the while the inappropriate dimension
of the response has questioned the seriousness of the lesson to that point, and,
the seriousness with which science (engineering) tends to be presented general-
ly. We have both reversal and reaffirmation, an opening for life as a whole to
come through the crack momentarily provided, which immediately closes again

11
It is important to retain that the social order does not take its origin in the deci-
sion of individuals. Rather, there is an event that exceeds what individuals intend,
to which they are subject and subjected to. In the sociology of emotion, interaction
rituals, such as the one the course of which we are following in this analysis, are
understood as the emotion transforming mechanism: Interaction ritual theory
gives the most fine-grained picture of how emotions are transformed in the process
of interaction: rituals begin with emotional ingredients . . . and they produce other
sorts of emotions as outcomes (Collins 2004, p. 105). The ritual is joint work, the
unit that allows us to understand actions as integral parts in a whole-part relation.
What we observe is a social process, and our analysis has to attend to this social
nature if it intends to be rigorous. To abstract and make attributions to the indi-
vidual, we would have to show the conditions that make abstraction possible and
at what costs.
12
Much research in the field of science education, where the text has been pub-
lished in one of its iterations, is oriented towards the cognitive aspects of life irre-
spective of its affective side. But it is not so that there are cognitive and affective
sides that somehow add up or interact. Instead, every experience simultaneously is
practical, intellectual, and affective (Vygotskij 2005). These different aspects can-
not be disentangled from each other, though one or the other might appear more
visible at a particular point in time.
13
Here again we find a tension. The IRE is what researchers recognize. It is a non-
ethnomethodological characterization of the situation. If the participants were
interviewed afterwards, they might say that they are poking fun at the teacher,
and she might gloss the event as students being a bit giddy. But this fragment is
only part of the overall event, reviewing what had been talked (or lectured) about
during the preceding week.
 SHOP FLOOR AND PLAYING FIELD 153

as Victoria asks another student to contribute to the enactment of the IRE se-
quence.
In a way, there is a parallel to the analysis of Rabelaiss work by Bakhtin
(1984b), which shows how the official seriousness of life during the Middle
Ages could be questioned and relativized. But this questioning and relativization
had its placeduring feasts and carnival, in carefully framed poetry and writ-
ing, in the specific roles of the buffoons and harlequins. In some countries, the
practice of carnival continues; and in other countries, stand-up comedy and
comedy shows have taken over the role previously relegated to carnival. Nearly
anything can be said because critique and laughter are constituent parts of the
frame. Certain truths can only be articulated as joke and other derivatives of
humor (e.g., pun, satire, derision) because they are both correct and incorrect,
truths and non-truths, seriousness and its inversion. The joke relativizes itself in
asserting the truth of its own untruth, its undecidable nature, ambiguity, the
double take. There is an additional aspect. The joke seeks an audience14; it is
produced at the cost of a second party, but has to be witnessed by a third party.
Stand-up comedians and satirists deride, for example, politicians explicitly for
the audience present. In the present instance, the joke is at the expense of Victo-
ria, to whose husband the reference has been made, and we can hear that the au-
dience has attended through the laughter and cackles. Victoria also gazes in the
direction of the camera and the researcher recording the lesson, then to Prunella,
again at someone else in the classroom, and then returns to Prunella. That is, not
only does the author produce the joke with a sideward glance at and for an audi-
ence, but also the recipient of the joke, the victim is oriented to those third
parties who not only overhear what is said but also for whom it is said.

Engineers are Rich: A Stereotype? Fragment 9.2c has the same structure as
the previous one. At first, Prunella produces an utterance as a candidate for
completing a questionresponse pair to be evaluated in a third turn at talk. Pru-
nella suggests that they [engineers] . . . think of more efficient ways to do
something (turn 17). Victoria immediately responds with a positive evaluation,
Ooh, I like it and then, as previously, repeats what Prunella has said, the
more efficient ways of doing something (turn 18). A side sequence designed to

14
Here the text names precisely what is at stake: The joke seeks an audience after
already having found the comedian responsible for telling it. The joke is the
phenomenon; Victoria and her students are staffing it. The joke is the unit of anal-
ysis rather than the individuals. That this is so is evident as soon as we realize that
jokes are recognized whoever is involved and wherever we are. In the same way,
walking into a room, we recognize what is currently happening as a mathematics
lesson, because independent of the contextual particulars, independent of the staff,
the phenomenon exhibits itself so that practitioners easily recognize a mathematics
lesson rather than something else to be in the course (Roth and Thom 2009). This
is what allows the analyst in part B of this book to see and hear some social situa-
tion unfolding even though he does not know anything about the participants, their
histories or their institutional positions. It is the phenomenon that exhibits itself, in
and through the ways in which relations observably are produced and maintained.
154 CHAPTER 9

ascertain that other students understand the term efficient occurs prior to the
next solicitation of a response to the initial question.15 The next student called
upon, Irisana, suggests that they are rich (turn 22). Victoria grimaces and
grins, and then repeats the students utterance, they are rich, while emphasiz-
ing the adjective (turn 23). But even before she has completed her utterance16,
and in particular before having the time to provide an evaluative term, another
student comments with low volume, they are not that rich (turn 24). In the
next turn, Victoria utters what can be heard as taking on this student utterance
and simultaneously evaluating two previous comments, we did mention the
topic of stereotyping engineers followed by a negation of Irisanas and Terras
offers, they are not all incredibly good looking and rich (turn 25).
Fragment 9.2c
> 16 V: <<p, grinning>we did mention they were incredibly good
looking.> <<laughing > pru(h)ne(.h)lla(hh[hh])>
17 P: [uh] they
got um think of more efficient ways; to do something,
18 V: OOh i lIKe it (0.22) pruNELla
said that the more effICient ways
of DOing something. (0.40)
effICient mEAning?
19 (1.16)
20 P: bETta.
21 V: EAsia:: betta and? (0.43) quICka
(0.20) yes: irisANa
> 22 I: * they are RICH:

15
The side sequence is not designed by any individuals. It comes to be in and
through the joint actions of the participants. Once the Saying is over and some-
thing exists as the Said, then we may attribute the name side sequence. When
those involved do what they do and say what they say, they are not intentionally
producing a side sequence. We therefore see here the (perhaps dangerous) co-
existence of scholarly discourseinvolving terms such as IRE, side sequenceand
the description of the event in terms of the orientation and talk of the members to
the setting.
16
The term utterance is frequently employed to translate Bakhtins Russian
vyzkazyvanie; a better way of translating it might be statement, which uncouples
the Said from the act of speaking so that the statement can be attributed to the
hearer as well (Roth 2013g). The text says her utterance, when in fact if the
statement did not belong to the students, then there would not be a conversation.
Victoria produces the sound-words, but these sound-words resonate in the ears of
the students. Thus, the sound-words physically and psychologically belong to
speaker and audience. Taking this approach allows us to make serious headway
theoretically, because language and its content no longer is attributed to the indi-
vidual but is taken to be a social, transactional, jointly achieved phenomenon.
 SHOP FLOOR AND PLAYING FIELD 155

> 23 V: * they are RI[CH:

24 T: [<<p>they are not
that rich>]
25 V: we did [mention ]
[((a student laughs))]
the tOpic of stereotyping
engineers they are not All
incredibly good looking and rIch.
yes ((points))

Prunella provides an answer that receives a clear positive evaluation, ooh I

like it (turn 18), followed by a question concerning the sense of the term effi-
cient (turn 18). One student responds, but Victoria then provides the synonyms
for the term efficient, easier, better and quicker (turn 21). In the next re-
sponse, Irisana provides an answer according to which engineers are rich
(turn 22).17 Victoria repeats, a student utters in low voice, they are not that
rich and then Victoria limits the generality of this and a preceding response,
they are not all incredibly good looking and rich and classifies them as stereo-
types. Here she formulates18 the description as a stereotype, and that they have
already mentioned it. As previously indicated, the response is both appropriate
and inappropriate simultaneously. Victoria marks it as such when she says that
it is a stereotype. A stereotype takes a characteristic and over-generalizes it to
an entire population in a clearly inaccurate or inappropriate way. However, this
response may be correct as it is likely to be representative of engineers in par-
ticular positions at certain points during their careers but, as a stereotype, it is
also inappropriate. This particular response is realized in part as a humorous
one, as indicated by the teachers grin and students laughter. It thereby both
undermines and reaffirms the IRE structure. The teacher evaluates the comment
as an expression of a stereotype. More so, she articulates the two statements as

17
When using a phrase such as Irisana provides . . . we should not assume that
the girl reflected first in the way an agential, constructivist, or rationalistic dis-
course would make us believe. The give-and-take of verbal exchanges happens so
quickly that we find ourselves as having said something rather than saying some-
thing after more-or-less careful deliberation. Our entire discourse, lay as profes-
sional, is oriented to express agency and, therefore, is subject to a cause-and-effect
fallacy (Nietzsche 1954). Our difficulty exists in this historical condition of a lan-
guage that not only expresses content in terms of cause and effect but that uses this
same model to theorize itself: subject, verb (action), (direct, indirect) object. His-
torically, this is a way of orienting to the world particular to the Greco-Roman
tradition and the metaphysics it embodies (e.g. Heidegger 2006). How might we
writeand this, too, is a problem of research method, one focusing on the commu-
nication of the researchresearch findings without taking recourse to attributing
cause to individual human intentions and actions? How do we write research so
that the phenomenon, the jokewhich we tend to be unwittingly participants in
becomes the phenomenon rather than the individual human actor?
18
The text here refers to the technical term to formulate, that is, what members
themselves do to say what they are in the process of doing.
156 CHAPTER 9

having been the subject of a previous discussion, and of the topic of these char-
acteristics as stereotypes. She formulates the dual nature of the responses as ap-
propriate and inappropriate, as stereotypes; and she formulates the repetition.

Engineers are Smart and Are Called Wes. In Fragment 9.2d, we again find an
utterance on the part of an institutionally designated student followed by the
completion of the IRE sequence on the part of the teacher.19 The student says,
They [engineers] are smart (turn 27), followed by Victorias repetition,
theyre smart and the evaluative affirmative, Yea (turn 29). The next stu-
dent then offers, with a much louder speech volume, the name of Victorias
husband, Wes (turn 31). Victoria begins to repeat, Theyre called Wes,
when a boy in the front row claps his hand, Victoria, who preceding this point
has displayed a very serious face, breaks out in laughter, and several students in
class begin to laugh (turn 33).
Fragment 9.2d20
26 (0.40)
27 P: they are smart.
28 (0.32)
29 V: they=re sma:rt; yea:
30 (0.92) ((Fig. 3a))
31 B: <<f>wes>

32 (0.95)
> 33 V: <<len>they=re called [wes
(0.36)> <<smiling>
[yea(hh)<<ff>(.heh)>
> [((boy
claps))
> [* ((laughter, in class))
> 34 (0.68)]

Here, as before, Victoria both repeats the student statementthough in an

expanded form Theyre called Wesand provides the evaluative affirmative,

19
Student and teacher, here, are terms used as institutional designations rather
than as signs that point to the function the designated individuals have in the class-
room. These terms leave open who learns (i.e., is a student) and who teaches. We
may then see how students teach teachers, who learn while participating in rela-
tions with students (Roth and Radford 2010).
20
Video offprints were used in the original publication, in which Victoria was one
of the co-authors. She consented to the use of the photographs. Video offprints
showing students were neither needed in the text nor available without seeking
additional permission. In our research, we tend to have an additional ethics form
in which participants agree or not to the use of photographic and video images. If
researchers are interested to make available some of the visual aspects of the situ-
ation, they can always resort to the use of drawings.
 SHOP FLOOR AND PLAYING FIELD 157

Yea. That is, she marks the response as one that turns the utterance sequence
into a proper questionresponse pair; and, in and with her laughter, she simulta-
neously marks it as a joke, as something to be laughed about. It is both an ap-
propriate and inappropriate response simultaneously and the difference is unde-
cidable. Moreover, it is not just Victoria who laughs. There is a more general
enjoyment. Here, then, as before, the understanding about the dual nature of the
non/response is a sociological phenomenon, an infection, at least we know it for
those students who, as Victoria, laugh or provide other relevant signs. In this
situation we see the function of laughter to reproduce and transform a close rela-
tionship with a topic that is improper to science content. However, by laughing
with the students, Victoria actually marks a level of intimacy, which is enacted
as an interactive matter in and through the production of impropriety and the up-
take of it through the production of shared laughter in response to it (Pomer-
antz and Mandelbaum 2005, p. 166). In this instance of introducing the teach-
ers spouse into the classroom discourse, there is laughter. But in a similar
situation, the production of an answer making the same reference to the teach-
ers private life is not associated with laughter. Then, there is a possible initia-
tion (Fragment 9.2a), when, following Terras uttering of Wes, the name of Vic-
torias husband, there is a hint of a smile crossing her face. This smile comes
while she rearticulates the question and calls on Krissy to respond. Here, then,
what could have been a possible initiation of laughter is resituated by the con-
tinuation and by calling upon another student. An instantiation of intimacy is
averted as one of the possible interaction trajectories in that this sense is not
taken up and reasserted by other contributions.

The Social Actor and the Analyst-Theorist

The traditional research article contains a description of the (special) methods used
to produce, observe, and document some phenomenon and to interpret the docu-
ments (data) produced. This practice has historical origins in the natural scienc-
es, where researchers working in different parts of Europe and the world were in-
terested not only in communicating their findings but also in communicating how
to produce the particular phenomenon that were in some ways produced in scien-
tific laboratories. The methods scientists use are presupposed to take special train-
ing and no longer are those used in the everyday world of non-scientists. Even with
such descriptions, scientific experiments are not necessarily reproduced easily
somewhere else, as numerous investigations in the social studies of sciences show.
Scientist-analysts then become a special human kind, different from the normal
everyday just plain folk. The human sciencespsychology, sociology, and all oth-
ers modeled after themfollowed the natural sciences in the attempt to become
more scientific than they had historically been, given their origin in the humanities
and philosophy. A distinction therefore is produced between two types of individu-
als in society: researcher-scientist-analyst-theorist using special methods and other
folk thought to draw on everyday common sense and method.
158 CHAPTER 9

One consideration often overlooked is that scientists themselves develop on the

basis of the everyday common sense and method that initially characterizes their
way of being in the world. The scientific sense and method that they subsequently
develop, even if these overturn the everyday sense and method, are premised on
the latter, which provides the very material, resources, and tools for overcoming
itself. The everyday sense and methods are the condition for science, and, even
when overturned, come to remain its foundation (Husserl 1976b). This is even
more so important in the context of the human sciences, where the everyday world
not only constitutes the context in which the researcher-scientist-analyst-theorist
has developed and acquired the understanding of the social world that now has
become his/her topic of research. This everyday way of understanding the social
world not only is imported into the sciences but also (generally) presupposed with-
out interrogation. Husserl (1976a), aware of this effect, therefore calls on scientists
to begin by bracketing not only their scientific theories when they approach some
phenomenon but also their everyday ways of understanding. Bracketing here
means that these understandings are held at bay. We see this in operation in chapter
7, where the analyst is exhorted to attend to the relations to investigate the ways in
which knowledge-power are produced differentially rather than beginning with the
presupposition that institutional designation inherently means differential
knowledge-power. Bracketing, which neither negates the everyday world as the
sophist does, nor doubts its existence in the way the skeptic does, withholds judg-
ment:
That is to say, all sciences aiming at the natural world, as steadfast as these
are for me, as much as I admire them, as little as I am interested in saying
something against them, I exclude, making absolutely no use of their validity.
I do not adopt a single one of their statements, even if their evidence is per-
fect; none is taken on by me, none is giving me a foundation. (Husserl 1976b,
p. 65)
In the approach outlined here, social-analyst-theorists do not characterize their
work as requiring special methods and, in so doing, engage in the building of
boundaries with respect to non-social-analyst-theorists. Instead, researchers
acknowledge belonging to the very social world that is their topic. Moreover, re-
searchers do not (initially) use scientific or common sense to make judgments
about the social world of interest but bracket all explanations, descriptions, and
denotations. Husserl suggests accepting a statement only after it has been bracket-
ed. Whereas Husserl only uses the term bracketing, Garfinkel and Sacks (1986)
literally use brackets to enclose that which is to be researched. This would be ex-
emplified in the following fragment, involving an interviewer and a child. The first
bracketed (underlined) phrase formulates what is going to come in the same turn:
not only a question but also a question that is very simple. The second bracketed
and underlined phrase projects what is to come in the reply part of the turn pair: an
explanation. Garfinkel and Sacks articulate that form of talk as formulating, that is,
what conversationalists are doing is saying-in-so-many-words-what-we-are-doing
(or what we are talking about, or who is talking, or who we are, or where we are,
etc.) (p. 171).
 SHOP FLOOR AND PLAYING FIELD 159

Fragment 7.3
01 I: [so the question is very simple.] (0.24) .hhh [could you
explain me:?] why:: (0.55) why: we have day and why we
have night;
02 (1.28)
03 AJ: kay (2.06) ((licks lips with smack)) be (0.16) cause
.hhhh (0.84) we need day to pla:y: anweneed night to
sleep. (0.69) .hhh and then if we dont have day we dont
have the flash light or we can bump to <<dim>something
or something>; (0.47) .hh=yea; (1.04) we can bump to
something;

The bracketed part not only is simply formulating what is happening or sup-
posed to happen but, for the purpose of the analysis, needs to be held in abay as to
a judgment about what actually occurs or is to occur. This then allows us to inves-
tigate the work of doing whatever participants call asking a very simple question
or providing an explanation. To make clear that distinction, the authors denote
what happens by means of expressions such as doing [asking a very simple ques-
tion]. The purpose of the analysis is to investigate the work, the doing, how
very simple questions are asked and responded to, which may in fact turn out to be
much more complex than the everyday description lets on. What distinguishes the
everyday social actor and analyst-theorists is the latters orientation to the phenom-
enon and the rigorous ways in which they conduct their analysis to make them eth-
nographically adequate. The result of an ethnographically adequate description
would be something like a recipe that we could take to produce the phenomenon
(as in laboratory or cookbook instructions) or a recipe for finding some phenome-
non (as the descriptions in naturalist field guides).
Despite the foregoing, my article on laughter in science classrooms did have a
methods section. This was in response to the community for which the article was
intended, a community unfamiliar with the approach. The following text was used
to describe the analysis (Roth et al. 2011, pp. 440441) even though, as can be
seen, it argues that no special methods are required for doing that kind of analysis.
That is, in this description of method, the article clearly situates itself within the
practices of a community that differs from the practices of the community the for-
mer describes (Smith 1983). This is so because the analysis makes use of the same
social competencies as the participants themselves, which, in the particular context,
was producing and recognizing joking and laughing. That is, this text can also be
read and comprehended as an argument that the description of special analytic
methods is superfluous. The second part concerning the analytic method describes
the approach taken in applied conversation analysis. It makes the case by varying
the second part of a turn pair such that the first part can be seen to have something
different in that situation from which the transcription would have been extracted.
That is, this text constitutes something like a didactic text describing to its intended
audience its own superfluous nature.

Analyses

In our analyses we attempt to produce ethnographically adequate accounts of

the actions that we observe. This means that, as researchers we must articulate
the same hesitant and momentary contexts that the natives are displaying to
160 CHAPTER 9

each other and using to organize their concerted behavior (McDermott et al.
1978, p. 246). That is, an ethnographically adequate account aims at providing
descriptions of events at the level of the participants rather than, for example, at
the meso-level accessible only to temporally and spatially removed observers.
Such ethnography draws on the programmatic work of ethnomethodologists
and the empirical work of conversational and context analysts as well as on
supposedly diverse schools of anthropology as the interactionist, the cognitive,
the ethological, and the linguistic (p. 267268) and is very similar to the ap-
proach that others also advocate (e.g. Erickson 1982; Mehan 1979). Because so-
cial structure is both resource in/for and product of social interactions, therefore
visible in the give-and-take of each situation, no further additional meso- and
macro-analyses are required to access structure in institutional talk generally
(Boden 1994; Rawls 2002) and STEM [science-technology-engineering-
mathematics] classroom talk particularly (Roth 2010).21 Moreover, because we
investigate laughter as a resource in the interactions observed, a posteriori ac-
counts, for example by the classroom teacher, are of little interest. In fact, be-
cause a participants understanding of an event changes with hindsight and with
distance to the event (Husserl 1991), such a posteriori accounts may actually be
detrimental to an ethnographically adequate account of interaction as process.22
This approach has immediate consequences for professional analysts, includ-
ing the authors of this paper: to be able to make sense in the way of the actors,
the professional analysts have to have the equivalent social competencies. No
special (meso-, macro-) methods are required to do this research other than the
methods that the interaction participantsi.e., people, ethnothemselves use

21
This statement specifically rejects the very specification of a special methods
section generally and of the description of the special analytic methods particular-
ly that contextualize it. If we were to analyze this methods section, this statement
would provide us with some indication of the relationship between the article and
the readership of the journal. This readership could be hypothesized as being less
familiar with the approach common in ethnomethodology or conversation analysis,
which operate on the principle that no special methods are required or need to be
specified.
22
There are many reasons why we might be interested in bracketing the accounts
that social actors themselves provide after the fact. First, as Husserl states in the
cited reference, the way in which some original situation is experienced changes
with intervening experiences. Any account of the situation, therefore, also is a
function of all the situations that the social actors have participated in since. Se-
cond, social actors tend to articulate what they have done, providing yet another
gloss of what happened and perhaps why. All these glosses are to be bracketed on
the part of the analyst, who is encouraged to classify them as yet another alterna-
tive way of accounting for what has happened. Third, if analysts were to collect
these different ways of accounting for some phenomenon, they would do little more
than establishing a catalogue: a phenomenography of some phenomenon (e.g., the
different ways of asking a very simple question). The real novelty that research
can contribute is to derive the structure and movement of the work by means of
which is done what the gloss says that has been done.
 SHOP FLOOR AND PLAYING FIELD 161

(Garfinkel 2002). In talking, social actors produce both the context and content
of their talk.23 In trying to understand social action, we take the turn pair as the
smallest unit. That is, we are not concerned with individual utterances (turns at
talk), which cannot be understood in and of themselves. The value of an utter-
ance in and to a conversation is tied to its social evaluation, which the listeners
make available in their own immediately following turn or turns (Bakhtine [Vo-
lochinov] 1977).24 Thus, what the function of an utterance is from the perspec-
tive of the collective activity can be established only on how a second social ac-
tor interprets and acts upon the actions of the previous social actora
position consistent with the conversation analytic method employed in the tran-
scriptions and analyses of this paper.25

Conversation Analysis

To concretize our analytic approach, consider the example of This is a test

tube. Without further clues, we cannot say what is meant by this locution.
Although locution has the grammatical structure of a statement, it could be
heard as a question or as an insult. Consider the following three situations with
the same words in the first turn but with different intonations and different sub-
sequent turns.
A 01 V: THIS is a test tube?
02 S: no its a beaker.

B 01 V: THIS is a test tube?

23
Readers easily recognize that this phrase is an expression of the principles un-
derlying the production of ethnographically adequate accounts (see chapter 1 and
above). It is this very feature of talk that allows the instructor-analyst in part B of
this book to do what he does: derive the nature of the situation a specific instance
of which has led to the transcription in his hands. It is also that part that many or
all formal approaches (quantitative and qualitative) research simply omit. This
omission has another important implication: a misunderstanding of language,
which is taken to be analyzed for its content rather than for the fact that actors
shape their relation while they speak. This latter may be more important than the
actual semantic content of the speech.
24
This reference would not normally be found in the list of authors who had their
training in conversation analysis strictly speaking. Here the interest would be of
purely linguistic nature rather than focusing on the pragmatics of conducting so-
cial affairs. But long before there was conversation analysis the literary theorist
M. M. Bakhtin and the members of his circle, including V. N. Voloinov, had point-
ed to the irreducibly social nature of talk and text, which therefore can be under-
stood only in this conjunction of speaker and recipient or author and reader.
25
This statement, too, may be read as an indication of the differences between (a)
the types of analyses conducted in the article where the preceding analysis of
laughing and joking appeared and (b) the more common practices of the field to
speculate about the meanings or conceptions that individuals construct in
their minds.
162 CHAPTER 9

02 S: what else could it be.

C 01 V: this is a TEST tube.

02 S: so do you think i am dumb?

In situation A, the second speaker treats turn 01 as a question about the na-
ture of an object (perhaps also pointed to by the first speaker). In situation B, the
second speaker treats turn 01 as something like a rhetorical question, asserting
in fact that the object at hand is a test tube. In situation C, the second speaker
treats turn 01 as a statement about his/her mental capacities. That is, the content
(meaning) of turn 01 cannot be established on its own but only through the ef-
fect it has on others, which, in turn, is indicated by the subsequent locution or
action. We, the authors and analysts, therefore do not impute what the speakers
intend to say, because in most instances, we do not have access to it.26
Prosody is an important resource for interaction participants. That is, if the
intonation (pitch) moves upward in turn 01, then the utterance tends to be treat-
ed as a question. If the prosodic cues emphasize this, then the question tends
to be whether the indicated or another object is named test tube. To indicate
more precisely what speakers make available to each other as resources for
hearing, we also provide prosodic information, that is, information about pitch,
speech intensity, and speech rate.27
What ultimately matters in the analysis is the trajectory of the analyzed epi-
sode. This trajectory depends on how participants themselves hear one another.
Therefore, conversation analysts do not use the individual locution as the unit of
analysis but the turn pair, which is referred to as a speech act. The first turn en-
compasses the locution (the act of speaking) and the illocution (intend of speak-
ing), whereas the second part constitutes the perlocution (effect of speaking).
Therefore, it is the second turn that reveals whether the participants heard a
question, statement, admonishment, and so on. To understand what is happening
in an episode, therefore, analysts have to be able to follow the conversation that
they overhear in the recorded episodes.28 Rather than a special method, analysts
need to bring the same cultural competence that the interaction participants dis-
play to each other. (Interested readers find more on this form of analysis and
how to teach it in Roth and Hsu 2010, especially chapter 10.)

26
This is but another way of saying that what matters to the analysis is the ways in
which the participants themselves constitute order and make the order-constituting
nature of their exchange visible and accountable to each other.
27
Prosody may be understood to be part of the posture that social actors take, an
integral aspect of the social situation that can be documented in very objective
ways, for example, by means of the determination of pitch levels, pitch contours,
variations in speech intensity (volume, loudness), variations in speech rates, and
so forth.
28
The text should have said that the internal dynamic of the event is a function of
the turn-taking sequences, not a function of the researchers concepts. The internal
dynamic of the situation arises from the give-and-take that binds the actors and
their material world into a single unified and irreducible whole.
 SHOP FLOOR AND PLAYING FIELD 163

There is something different, though, between the possibilities of the social ana-
lyst and that of the social actor. What we get from the social actor are all those
ways in which social situations are organized, recognized, oriented to, and ac-
counted for. Whereas this form of analysis does get us at the ways in which the
corresponding work is done, it cannot get us at understanding how these social
lifeworlds came to be what they are. In chapter 1 I refer to the work of the Canadi-
an feminist sociologist D. E. Smith, who, being a single mother, found herself in
the same social spiral as many of her single-mother research participants. This was
so because of the use of a single-parent family discourse and its often-tacit referent
in the standard North American family. A relevant but little known description of
the same problem and the corresponding need for an analysis that transcends the
lifeworld was provided within a very different research tradition (Holzkamp 1983).
The author describes a situation where
a 1617-year-old boy had tremendous problems with his father and initially
made him responsible for everything. They almost killed each other, because
the boy thought, he is in my way, he is keeping me away from my possibili-
ties; until they realized eventually at some time in their conflict that the fact
that they are getting at each other is a deflection from the reasons for the en-
tire conflict situation, that their own situation is the result of their class. And
we could beautifully illustrate that the hopelessness existed in their making
each other responsible, which was made invisible by the real circumstances
of their misery. And then in the next stage, when they slowly grasped that
they had to try to change, and then the thing that touches us so frequently,
when the boy and his father first go to some office and try to find a new
school for the boy, after he was dismissed from school for a fourth time and
where the father previously only beat up on him. And all of a sudden they go
and overcome the fear of the institution and comprehend that it was not the
end, that the real enemy is not the father, that the son is not the actual enemy.
But these inimicous relations exist within the context whereby they are fixat-
ed upon each other, where one has to first of all comprehend the contextual
totality to overcome this immediateness. (Holzkamp 1983, p. 148)
In this description, the son, the father, and the unspecified psychologists work
together through the problem. This working through requires the son and father to
understand their previous relations not in the way in which they experienced these
but through the lens that allows them to understand the societal-historical situation
of working-class people. Here, the kind of work that the sociologist Smith engaged
in, that is, the historical analysis is required to grasp the very constitution of the
discourse that contrasts single-parent families to the model of the standard North
American family. This does not mean, the author emphasizes, that we must begin
with an examination of our subjective states or with a personally biographical ac-
count of our relation to the ethnographic situation (Smith 1981, p. 317). Instead,
our task is to transform members practices into our practices as members, or
rather to discover how to take up methods of inquiry in which the method itself is
explicated as an integral aspect of the inquiry (p. 317, original emphasis, under-
line added). It is because the methods of everyday society are self-explicated that
164 CHAPTER 9

we researchers learn about it, becoming familiar with the ongoing language-games
at play.
Smith uses the analogy with the solar system to articulate the relations between
the objects of description and the description. In the received ways of doing eth-
nography, the ethnographer is an observer at rest, like the early astronomers using
their position on the earth as the center of the universe. Everything else is moving,
and the observers own movements are attributed to the movements of the ob-
served (stars, social phenomena). Once observers realize that they are moving,
observations can be understood as relations. At that time, the ethnographic descrip-
tion can be located as an active relation between the reader and the text and
through the text to an object, at which point the character of that relation can be
made explicit (Smith 1981, p. 322).
When analysts read transcriptions, they do so in ways that are not idiosyncratic
(Livingston 1995; Smith 1983). Instead, reading itself is a practice that we learn
while participating in social relations; reading is a social relation before it becomes
something that can be attributed to individuals who read in the absence of others.
All we have to do is investigate when, where, and how we learn to readwhich is
never on our own, but always with others and at times that often precede our first
articulation of a recognizable word (e.g. Roth et al. 2013). Some of the reading
practices derive straight from the everyday world; the other are reading practices
cultured in communities of scientific practices. Understanding this double origin of
our reading practices, the product of two different but related forms of relations,
and locate them appropriately as part of our methodological reflection in the ac-
counts we provide allows us to come to grips with what we are doing in and
through doing research. Researchers must never forget this double form of belong-
ing, this double relation, and the ways in which it structures what they do.
10

The Documentary Method of Interpretation

In different chapters of this book, I refer to the IRE sequence of turns, where
someone who turns out to be in the know initiates an exchange generally by asking
a question, someone else replies, generally the person who is less in the know and
therefore turns out to be lower on the knowledge differential, and the first person
evaluates the foregoing reply. In chapter 2, the instructor-analyst detected, in the
exchange between Heidi and David, on the one hand, and the teenagers Amanda,
Ashley, and Michael, on the other hand, the forms of exchanges that go by the ac-
ronym IRE. The analyst took the concrete data, worked out the form in which the
relation played itself out, and then hypothesized possible societal situations. These
general situations concretize themselves in specific settings and events, where they
would have led to the concrete verbal exchange made present again in and through
the transcription. The general pattern underlying the analytic process is displayed
in Fig. B.2 (p. 33). I note that the instructor-analyst employed the documentary
method of interpretation, whereby he took the data as having an objective sense in
that the concrete words are available to any participant and analyst. The data also
have an expressive sense, as they articulate the specific (ruling) relations that par-
ticipants both make and are subject and subjected to. Finally, there was a documen-
tary sense: in the data, a type of situation, concretized in the particular setting,
manifested itself. In this chapter, I work out in greater detail the different uses that
have been made with the documentary method of interpretation: (a) as an analytic
approach in the human sciences interested in researching such phenomena as Welt-
anschauung (worldview) (Mannheim 2004); (b) the understanding work of lay and
professional sociologists alike, which consists of treating what has been said as a
document of, as pointing to, or as standing for something else that could have been
said (Garfinkel 1967); and (c) a way of learning some phenomenon, known only
through the ways in which it manifests itself, through experiencing such concrete
manifestations (Roth 2013d). Because researchers are learners, lay/professional
analysts engaging in understanding work, and professional analysts with respect to
some phenomenon in their field of interest, all three uses of the documentary
166 CHAPTER 10

Fig. 10.1 a The logic of deduction. b The logic of induction. c The logic of abduction

method of interpretation are relevant to understand what analysts do and how they
do it. Before elaborating on the different ways in which the documentary method
of interpretation has been put to theoretical and methodological use, I describe the
logic of abduction with which the documentary method of interpretation might be
confused upon cursory and initial encounter.

Abduction

There are three forms of reasoning that can be found in the sciences generally (Eco
1984) and in qualitative data analysis specifically (Reicherts 2013): deduction,
induction, and abduction (Fig. 10.1). In deduction, the reasoning begins with a
given generalizationa theory, rule, or lawand proceeds to the concrete case
considered and then to the actual observation. For example, the ethnographer might
have a general rule stating that in Francophone cultures, non-friend adults address
each other with the plural second-person pronoun vous [you] rather than with the
informal tu [you] that friends would use to address one another. The ethnographer
might decide to go to France (case) and make specific observations. She would
find that the rule is satisfied in every single observation. The ethnographer might
also decide to go to some other setting, such as the French-speaking province Que-
bec in Canada an enactment of this rule frequently would not be observed. Here,
the reasoning is deductive in that it goes from a rulevouvoiement, the fact that
others are addressed by means of the formal vousto the specific case (France,
Quebec), and to the observation (Fig. 10.1a).
Inductive reasoning takes the exact opposite direction, from the concrete case to
some generalization in the form of a rule or law (Fig. 10.1b). The ethnographer
 DOCUMENTARY METHOD OF INTERPRETATION 167

unfamiliar with the French culture begins with concrete observations in France
noting that the French sometimes address each other by means of tu [you, singular]
and at other times by means of vous [you, plural]. Looking more closely at the con-
texts (case), she finds that in the office, on the street, or in a store or market, people
generally would use vous whereas at home, at a party with family and friends, and
among school students, she would find people addressing one another by means of
tu. The unsuspecting researcher might find that among older people, even husband
and wife might use the formal vous, as shown in the case of the former president
Jacques Chirac and his wife Bernadette. The vouvoiement would be found more
frequently in middle and especially upper classes than in the working class. Based
on her observations, the ethnographer might generate rules such as whenever
adults do not know each other, they use the formal vous or working class people
always use the informal tu at home, with family and friends. These rules have been
derived inductively based on the concrete observation made. This inductive logic
frequently is associated with the grounded theory method (Glaser and Strauss
1999).
Abductive reasoning has elements of both deduction and induction, though
combined into a new form (Fig. 10.1c). Based on the observations of specific cas-
es, the ethnographer states a generalization and tests it out in a specific case, where
she makes new observations. Thus, for example, based on some initial observa-
tions in stores or on the street when people ask for time or directions, she might
generate a rule: In France, adults address each other by means of the informal vous.
She would then put this rule to the test in different kinds of situations (cases), find-
ing that in some the rule is never violated whereas in others it is. She would then
modify the original generalization and put it again to the test. Three forms of ab-
duction have been identified: overcoded, undercoded, and creative (Eco 1984). In
overcoded abduction, the rule or law is given (quasi-) automatically. For example,
someone very familiar with the French culturean ethnographer, a sociologist
overhearing a conversation in which the participants address each other by means
of tu will hear (i.e., without any interpretive effort) this as an instance of their fa-
miliarity with each other. In chapter 2, the instructor-analyst reads the name David
Suzuki and almost automatically posits the name to be that of the environmentalist,
broadcaster, and scientist. In the particular instance, information subsequently
available to the analyst, the concern for the environment and erosion, confirms the
rule. In undercoded abduction, the rule has to be selected among a set of possible
rules. We observe this form of reasoning throughout part B, where the instructor-
analyst produces hypotheses that possibly explain the situation while awaiting fur-
ther testing as the data become available in reading on. This form of abductive
reasoning has also been referred to as an abduction strictu sensu (p. 42). Finally,
in creative abduction the rule explaining the observation has to be invented ex
novo (p. 42). One classical and often used example for creative abduction is the
case of Keplers laws or the invention of a shift from the geocentric to the helio-
centric model in astronomy. In part B, the instructor-analyst is unfamiliar with the
origin of the data, which led me to the analogy with criminal cases. This would
therefore be an example of creative abduction, because
[w]e implement creative abduction when dealing with poetic texts, as well as
when solving criminal cases. Many interpretive decisions concerning sym-
168 CHAPTER 10

bols . . . involve creative abduction. Many cases in which language is used

not to confirm but to challenge a given world view or a scientific paradigm,
and to decide that certain properties cannot belong any longer to the meaning
of a given term . . . require an interpretive cooperation that displays many
characteristics of a creative abduction. (Eco 1984, p. 42)
The logic of inquiry that we observe in the case of the instructor-analyst in part
B is that of (creative, undercoded) abduction. This is in stark contrast to the deduc-
tive logic that one often observes in the research literature, not only the empirical-
quantitative research approaches but also in qualitative research. Thus, for exam-
ple, most studies in education proceed from a set of presuppositionse.g., teachers
are in power, teachers teach, teachers know, or teachers produce their identity in
teachingand then observe precisely what the rule states. The abductive logic has
to be distinguished from the documentary method, which refers to a way of know-
ing ones way around the world. In this method, there is no rule required that ex-
plains the concrete data at hand. Instead, some cultural-historical phenomenon
e.g., the Weltanschauung of a particular era or culture, or impressionism in artis
taken in the way it manifests itself. These manifestations may not share any prop-
erty and, therefore, have to be thought dialectically, in the partwhole manner of
family resemblances presented in chapter 1 (p. 22). Thus, listening to Claude De-
bussys Prlude l'aprs-midi d'un faune [Prelude to the afternoon of a faun], im-
mersing oneself in Claude Monets Impression soleil levant [Impression, sunrise],
or reading Arthur Rimbauds poem Le bateau ivre [The drunken boat] have noth-
ing in common at the objective levelauditory, visual, and poetic experience. The
artists intended expression was to render some mood rather than a reality objec-
tively given (as in classical painting, photo realism). Despite those differences, all
these pieces of art are documents of a period, impressionism, which manifests itself
differently depending on the particular art form or philosophy of the era.

A Method for Investigating Pretheoretically Given Cultural-

Historical Phenomena Scientifically

The documentary method of interpretation was first described by the sociologist K.

Mannheim, who was interested in identifying a method for getting at the ways in
which a particular era looked at and understood at the (social) world, its worldview
(Weltanschauung). The particular issue that he was wrestling with concerned the
fact that in the social sciences, some phenomenon was subject to a pre-scientific
understanding (Mannheim 2004). He contrasts the situation with physics, where
the pre-scientific understanding of an object does not matter to the scientific con-
sideration of the same object because the reigning physical laws can be explained
without drawing on common sense notions. In contrast, the pre-theoretical experi-
ence never ceases to be an issue in the scientific analysis of aesthetical objects
(phenomena). This is so because something like the worldview of an era, zeit-
geist, or a particular artistic approach such as impressionism or expressionism,
constitutes a sense totality. This sense totality manifests itself, and is encountered,
 DOCUMENTARY METHOD OF INTERPRETATION 169

Fig. 10.2 There are three levels of senses that can be distinguished in a (verbal, physical)
action.

concretely in the specific manifestations of whatever the object is. For example,
impressionism in painting is (comes to be) known in and through the many differ-
ent ways in which we encounter the paintings of P. Czanne, P.-A. Renoir, E. De-
gas, C. Monet, . Manet, or A. Sisley. Although the painters and their paintings
differ, we can make out, after considerable experience viewing paintings under the
heading impressionist painting, the paintings of that era and distinguish them
from expressionist painting or cubism, another era. With further experience, we
come to recognize style differences, and, with it, distinguish the paintings of Re-
noir or Manet from those of other impressionist painters. Style again is one of
those phenomena that are grasped through a holistic sense arising from the many
manifestations, which in turn become manifestations of the same phenomenon.
Getting at this totalitywhether it be style at the personal level, as an artistic
movement, zeitgeist, or worldviewis the task that Mannheim set himself. This
means coming to grips with the non-theoretical foundation of perception and ap-
perception. The documentary method of interpretation then can be understood as a
working out of and coming to grips with the pre-theoretical sense that underlies the
sense totalities of specific phenomena. This totality is not, and cannot be grasped
by means of, an addition of all its manifestations that we might consider to collect.
Instead, each manifestation, every cultural objectification, has to be viewed as
something in itself (Mannheim 2004), as an instance of the possible (Bourdieu
1992). That is, we have to move from considering a manifestation as a thing-in-
itself and think it as a manifestation of a possibility that also manifests itself in
very different ways. This new totality, which is the totality of all possible concrete
manifestations, constitutes a unity that lies beyond all realizations of sense for-
mations and yet somehow is given through these (Mannheim 2004, p. 112). In
response to the anticipated question, How can this be?, the author describes the
three ways in which something like worldview (Weltanschauung) is given in and
to our perception: objective sense, expressive sense, and documentary sense (Fig.
10.2).
The objective sense refers to the way in which some phenomenon is given un-
deniably, in an unmistakably material way. Thus, apart from what a speaker in-
tended to say and what a recipient actually heard, there is the objective fact of the
words that have actually been articulated. These words have been recorded and
transcribed in a way that we may have little qualms about that this is precisely
what has been articulated. Other aspects of communication that have an objective
sense are prosodic values such as pitch, speech rate, speech intensity, pitch con-
tours, and others parameters that can be determined using computer software (e.g.
PRAAT). The expressive sense refers to what the speaker intended to say, the ac-
tors intended to do, the painter intended to show, or the composer intended to be
170 CHAPTER 10

heard. When a person gives an alms to a beggar, the objective sense not only in-
cludes the actual gesture but also what we can see, understand, and describe as a
helping gesture. This help is independent of the intentions of the giver.
The gesture, however, may have another, an expressive sense (Fig. 10.2). That
is, it can be understood as an intended expression of his/her mercy, goodness, or
compassion. This second sense is layered upon the objective sense, from which it
cannot be detached. It is also a subjective sense, which Mannheim attributes to the
intention of the person expressing himself/herself, but which can also be viewed
from the viewpoint of the subjectivity of the interpreter confronted with the task of
understanding the intention underlying the act.
There is a third sense that in turn is layered above the two former types of sense:
the documentary sense (Fig. 10.2). Mannheim uses the example of the alms given
to the beggar:
For it is entirely possible that I, the understanding person, comprehend the in-
tended expressive sense together with the objective sense and at the same
time continue the interpretation in a different direction. Pursuing the given
context, I all of a sudden see that the act of charity was as an act of hypoc-
risy. In this case it no longer matters what the friend has objectively done
and accomplished, not even what he has intended to express by means of his
act, but what documents itself for me in his act, even if unintended by him.
(p. 116)
In this latter layer of sense, the documentary dimension, something about the
character of the person is manifested that is not available and given in the other
two forms of sense. Everything about the friends behavior now can be taken as a
document of his hypocritical nature, his facial expressions, gestures, gait, speech
rhythm, or intonations (pitch). That is, the documentary sense captures something
identical in the face of the apparently and radically different objective and expres-
sive moments. Mannheim concludes that
this orientation toward the documentary, this grasping of the homologous in
very different contexts is something particular that must not be confused ei-
ther with addition nor with synthesis or with pure abstraction of common
traits; it is something particular because the unity of differences and the pres-
ence of a unity in difference constitute conditions that are particular to the in-
tellectual/intelligible world that should be kept uncontaminated by allegories,
which at least in part owe their lives to a spatial-materially oriented phantasy.
(p. 127)
At this point, the difference between the documentary method of interpretation
and abduction should be clear. In the logic of abduction, the investigator derives an
abstract rule that explains or generates the particular cases at handsuch as when
we derive the rules of an unfamiliar game from watching the game on television. In
the documentary method of interpretation, the relationship is that between some
objectively given fact and the phenomenal totality that manifests itself only in and
through this and other facts. Whereas all the concrete instances gathered in abduc-
tion obey the same law, and therefore share in common something in the expres-
sion, the concrete manifestations of a phenomenon may actually differand even
 DOCUMENTARY METHOD OF INTERPRETATION 171

appear opposite such as in the case of use-value and exchange-value, both of which
are concrete manifestations of commodity value. The quotation suggests that we
cannot get at this whole by means of addition, synthesis, or abstraction. This is so
because any part (manifestation) is a part only because it is comprehended in the
context of a whole. There are no parts that are awaiting their constitutive whole,
like elements or building blocks that are put together to form a composite. Instead,
the partwhole relation requires a simultaneous apprehension of the whole and its
parts.
We may now see that this characterization of the documentary method describes
how the instructor-analyst in part B is going about his work. He takes the words
with an unmistakably given objective sense. Every this word has been said and is
irremediably a part of the objective protocol. However, rather than being con-
cerned with some meaning of this word, intended by the speaker or experienced
by the recipient, the analyst focuses on what the sequences of turns are an expres-
sion of. Just as the physical gesture giving alms is an expression of help, the in-
structor-analyst identifies a non-genuine question and its context of an IRE se-
quence. (Help is not something one-sided, but, as a social fact, involves joint
work that is in excess of all individual worke.g., it has to be offered/given and
accepted.) But in the verbal protocol (transcription), something else is manifested:
the nature of the situation as a whole. Analysis consists precisely in finding a total-
ity of which all the different sense data can be viewed as documents. The instruc-
tor-analyst works out those objective and expressive dimensions in which he rec-
ognizes those types of social situations that he is more-or-less familiar with but has
not thought about in terms of the relation between manifestation and phenomenon.
Mannheim then goes on to describe how this method is useful in the analysis of
images. It is in this sense that the documentary method of interpretation has been
taken up especially in the German-speaking research community (e.g. Bohnsack et
al. 2010). In this specific use, the documentary method of interpretation can be
used to analyze images generally and advertising (qua social phenomenon) specifi-
cally (e.g. Bohnsack 2010b). In an advertising image, the analyst identifies objec-
tive aspects, including composition, color, orientations, and relations. At a second
level, the analyst focuses on what the individuals featured are doing and how they
are related, for example, forming a closely seated group of individuals who, in
stark contrast, do not have eye contact with each other. This then is seen as an ex-
pression of community all the while underlying individuality. All of these expres-
sive forms are then understood as documentary evidence for a particular lifestyle of
clothing, precisely that which the company is in the process of defining by means
of their advertising campaign.
In the German scholarly circles, the documentary method of interpretation also
is used in the analyses of group discussions (Bohnsack 2010a). The author sug-
gests that following everyday conversations, such as these might occur during fo-
cus group meetings, there often appears to be a concatenation of topics of conver-
sation without real connections (the talk goes from topic to topic); and the
discussions themselves often have little common structure to how members to the
setting organize their situation. Surprisingly, the participants appear to comprehend
each others talk without engaging in acts of interpretation. They are unable to ex-
plicate some underlying message that underlies what they communicate and how
172 CHAPTER 10

they do it. The researchers task is to explicate what the participants are doing:
The task of the researchers as documentary interpreters, thus, is the theoretical
explication of the mutual implicit or intuitive understanding of the participants (p.
104). The author provides an example from his research where youths had been
asked to talk about smoking. The youths talk about not being able to smoke in the
presence of their fathers, and then talk about their behaviors in the peer group. This
is viewed as documentary evidence for a separation of two spheres, an inner sphere
of life in the family and an outer sphere that exists outside of it. Because the youths
respect their fathers, it becomes impossible to bring matters from the outer sphere
into the family discourse precisely because of the respect that the youths have for
the parents.
This same frame also characterizes talk about ethnic discrimination. In contrast
to the frequent practice in qualitative research to introduce concepts that are inac-
cessible to the participants themselves, because operating at an uncommon level of
abstraction, the social scientific interpreters employing the documentary method of
interpretation do not presume or presuppose that they know more than the actors
in the field, but that those actors themselves do not really know what exactly they
know (Bohnsack 2010a). It is precisely the task of the analyst to explicate the
forms of knowledgeability implicit in the actors actions. We may see this in anal-
ogy to unearthing grammar from the talk of people who do not know that they are
speaking in patterned ways that can be articulated in the form of grammatical rules.

The Documentary Method as Ethnomethod

Ethnomethodologyconcerned with describing the ways in which everyday peo-

ple, lay and professional, conduct their mundane everyday inquiries (e.g., into what
someone is talking about)recognizes that the documentary method of interpreta-
tion is one among a panoply of methods we use on a daily basis (Garfinkel 1967).
The approach is described following a problem that poses itself to sociologist gen-
erally and the sociological fieldworker specifically. How is a body of knowledge of
social phenomena and social structure produced when the investigator cannot pre-
suppose relevant (declarative) knowledge or even have the (procedural) knowledge
of knowing what she is doing while doing it (e.g., recognizing and participating in
the social phenomenon)?
The method consists of treating an actual appearance as the document of,
as pointing to, as standing on behalf of a presupposed underlying pattern.
Not only is the underlying pattern derived from its individual documentary
evidences, but the individual documentary evidences, in their tum, are inter-
preted on the basis of what is known about the underlying pattern. Each is
used to elaborate the other. (Garfinkel 1967, p. 78)
To exemplify, the author describes an experiment in which ten undergraduate
students were solicited to participate in a counseling session to explore alternative
approaches to psychotherapy to deal with personal problems. The counselor, sitting
in an adjacent room, would provide only yes-no replies to the approximately 10
 DOCUMENTARY METHOD OF INTERPRETATION 173

questions that each student posed. Researchers recorded students comments fol-
lowing a yes or no answeras part of the ploy, students were asked to unplug the
connection to the counselor. Unbeknownst to students, the counselor selected re-
sponses from a random table, each student receiving the same series of yes or now
answers. Garfinkel notes that the students fundamentally employed the documen-
tary method of interpretation, where each answer was taken as a manifestation of
what the adviser had in mind and in what precisely the advice consisted of. What
he had in mind was part of what their questions were seeking to reveal. Clearly, the
questions were not preprogrammed: not only because the student were introduced
to the protocol as they came to the session but also because each new question took
into account the retrospective-prospective possibilities of the situation at hand; and
this situation continuously changed in and as a result of the exchange with the
counselor. Garfinkel notes that there was a search for a pattern, which, however,
was presupposed from the beginning of the search. The pattern is presupposed
even in the face of contradictory forms of evidence. Take the following excerpt
from one of the sessions
SUBJECT: I thoroughly dislike the study of nuclear physics. Nuclear Physics
124 will be one of my required courses to get a degree in physics.
Do you think I could get a degree in physics on the basis of this
knowledge that I must take Physics 124?
EXPERIMENTER: My answer is yes.
SUBJECT: He says yes. I dont see how I can. I am not that good of a
theorist. My study habits are horrible. My reading speed is bad, and I dont
spend enough time in studying.
Do you think that I could successfully improve my study habits?
(Garfinkel 1967, p. 86)
Although the experimenter-counselors answer was inconsistent with the stu-
dents preferences (i.e., dislike of nuclear physics, nuclear physics as requirement
for degree) or with his study habits, the student took the answer (yes) as the
manifestation of what the counselor had in mind concerning his case. The next
question asked (about improving study habits) took everything that has been said
so far into account in search of finding out what the counselor had in mind and
what his advice really was.
In this example, the undergraduate students employ the documentary method to
find out what the counselor really had in mind advice concerning their cases. But
the documentary method is not only employed in the contrived situation where the
inquirers are deceived about the intentions of the activity and in a context where
the responses were reduced to yes/no alternatives. Instead, the documentary meth-
od is employed as part of everyday conversation (Garfinkel 1967). Thus, a verbal
exchange can be thought of in terms of a course of ongoing grasping work where
each articulation is taken as the document of, as pointing to, as standing on
behalf of an underlying pattern . . . that the person, by his speaking, could be tell-
ing the other about (p. 40). Importantly, in the course of this grasping work, each
piece of documentary evidence not only was used to derive some underlying pat-
tern, but also was interpreted in view of what was known about the continuously
emerging, inherently open pattern. After describing this grasping work, Garfinkel
174 CHAPTER 10

suggests that it is the same form of work that Mannheim referred to as the docu-
mentary method of interpretation.
The documentary method of interpretation has embedded in it a circularity,
which has been pointed out in the context of feminist approaches to sociology, as
an ideological circle (Smith 1990a). This circle goes from the appearances of
something to the underlying pattern that manifests itself in the appearance: texts of
femininity supply images, icons, or descriptions of behavior, etc., coupled directly
or indirectly to the doctrines of femininity that interpret, and are expressed by,
them (Smith 1990b, p. 132). This ideological circle clearly is exhibited in the
phenomenon of fashion and womens appearance, where very slender (skinny)
women feature in ads, which are seen in terms of the doctrines of femininity that
manifest themselves in the bodies of fashion models.
Garfinkel also provides some examples to exhibit the ways in which profession-
al sociologists employ the documentary method of interpretation. For example, a
sociologist reviewing her interview notes wondering what the interviewee had in
mind will take what has actually been said as a manifestation of an underlying pat-
tern. The same will be the case when a sociologist observes people in some setting
and attributes this observation to a pattern that by itself is not observed or is non-
observable. Complex scenes like industrial establishments, communities, or social
movements are frequently described with the aid of excerpts from protocols and
numerical tables which are used to epitomize the intended events (p. 95, emphasis
added). Here, the notion of epitome points to the use of excerpts or tables as proto-
typical examples of the underlying pattern.
It is quite evident that the instructor-analyst in part B was going about his work
in this way. He took pieces of concrete data, words that had actually been said, and
considered this in and as of what they were documents of. However, he was not
looking for an underlying pattern in the sense of some abstract system of rules un-
known to the participants themselves. Instead, he was taking each articulating as a
manifestation of a type of situation that concretized itself in and through the partic-
ular videotaped and subsequently transcribed event. At the outset, he did not know
what this event might have been. He worked his way through the transcription de-
veloping some sense of what the people were talking about and, most importantly,
how they did this talking. The latter allowed him to articulate forms of relations,
and it was based on these relations that he narrowed the type of event that might
have been at the origin of the concrete data at hand.
We might think about an investigation in terms of the claims or assertions that a
researcher will ultimately articulate. For example, in an early qualitative research
article concerning students development of skills for doing scientific research
(Roth and Roychoudhury 1993), I assert, among others: (a) the identification of
pertinent variables in a particular context increases with the familiarity of the stu-
dents in a specific physical and conceptual context; and (b) students interpreta-
tions of experimental results evolved from simplistic and insufficiently supported
statements to identification of complex relationships among the variables using
multiple representations of their experimental data. I wrote the findings section by
first stating each assertion and then providing supportive data. These supportive
data constitute in fact the forms of documentary evidence for what is stated in so
many words by the assertion. In fact, the assertion is the end result of engaging
 DOCUMENTARY METHOD OF INTERPRETATION 175

with the data; in the presentation, the logic is reversed: by first stating some gener-
alization and then mobilizing precisely those pieces of evidence (examples thereof)
that had led to the assertion in the first place. In the following, I present one asser-
tion and the way in which it was supported by the data. The particular data mobi-
lized, and all those data of the same nature, constitute documentary evidence of the
assertion. Finding such an assertion amidst what the researcher has in hand, in
terms of data, constitutes the work of research. At the end of the next section I de-
scribe why pragmatic philosophers suggest that this processfrom the sea of
data to the phenomenon that it is evidence ofcannot be taught. (The following is
the text of the slightly edited penultimate version that was part of the published
journal article.)

Assertion 1. The identification of pertinent variables in a particular context in-

creases with the familiarity of the students in a specific physical and conceptual
context.1
Students did not naturally identify pertinent variables when they began to do
experiments in a new and unfamiliar domain. The data from both the study of
grade 8 general science and the eleventh-grade physics students corroborate this
finding. The students usually began with very unfocused questions, looking at
phenomena rather than at factors/variables that could be related. Then, over the
course of researching in the same physical, conceptual, and social context for
several weeks, questions became focused and they addressed the relationship
between specific variables.2 Thus, one pair of eighth-grade students began their

1
This assertion has to be read in the context of an approach in science education
whereby students were exposed to some new phenomenon and were asked to iden-
tify variables. The significance of this assertion lies in the fact that it reports exact-
ly the opposite to be the case: the identification is a function of familiarity with the
context. The first paragraph that follows provides, in generic terms as well as by
means of specific examples, descriptions of what could be observed in the eighth-
grade ecology classroom that I had observed and that had been taught by my col-
league and my future graduate student G. M. Bowen.
There was an additional experience that might have contributed to my suspicion
that the reigning science education canon with respect to scientific process skills
had it wrong. Prior to my PhD, I already had a Masters degree in physics. As part
of my PhD program I took a course in teaching methods where the instructor
asked us to identify variables. Without knowing anything about the phenomenon, I
found it hard to specify what might be a pertinent variable. As a personal experi-
ence, however, this is not good enough for stating claims that are useful for under-
standing the experience of other individuals as well.
2
This statement is generic, articulating some general sense of what could be ob-
served. It therefore stands in the same kind of relationship discussed here: phe-
nomenon to the specific observations that serve as the documentary evidence. In
the next statement, then, we find a series of specific observations that I had made,
and which in fact constituted the starting point for the analysis. The purpose of the
analysis was to derive statements that constituted some phenomenon; the method
176 CHAPTER 10

investigations with a question such as What different types of animals and

plants live in different amounts of light? In this question the variables type of
animals and type of plants are too vague, and the students never specified the
criteria for the classification of the animals and the plants. In the report, the only
distinction they made related to the size of the plants, large and small. The vari-
able amount of light was only vaguely defined and measured qualitatively in
terms of lots of light and little light. Two weeks later, the same two students
were investigating the relationship between specific variables such as in Is
there a relation on our slope between soil moisture and the air temperature and
what kind of plants fit into this relationship? Although the second part of the
question still contains a fuzzy variable, kinds of plants, the other two varia-
bles are very specific and could be measured directly. This development contin-
ued, and toward the end of the eight-week study of the biome, the students were
investigating questions such as What is the relationship between percent soil
moisture and organic content in three different parts of our area? or What is
the relationship between the amount of light and leaf growth in one week? At
this point, the variables were specific and the students measured both the
amount of light and the leaf growth quantitatively.
A similar pattern of skills development was observed in the older physics
students.3 They too began their research programs with start-up question that
did not seek causal relationships, but later on investigated relationships between
specific variables. Thus, a research program that began with the question Do
light bulbs act like resistors? would investigate such questions as What is the
relationship between the voltage across and the current through a light bulb?
and What is the relationship between power used by a light bulb and the light

by means of which this occurs is precisely the documentary method of interpreta-

tion.
3
Here, another piece of documentary evidence is mobilized in support of the asser-
tion. During my analysis, I had the sense that there were similarities between what
I could observe in the eighth-grade ecology unit and what I observed in my three
physics classes that I had videotaped over the course of a three-month period.
Whatever the data, the trick in the analysis is to see in them documentary evi-
dence for some phenomenon; and this phenomenon also manifests itself in other
sections of the data, involving different students of different age and in a different
disciplinary context.
From a methodical point of view, the problem is to derive a description of the
phenomenon when the researcher does not know what the phenomenon is that s/he
ultimately comes to describe, one that adds to the literature. That is, it is not just
any description that will do or any phenomenon but whatever we come to describe
on the basis of the documentary evidence we have has to contribute something new
to the research literature. If it does not, then we do not need to engage in the re-
search in the first place.
From the communicative point of view, having a second, third, . . . nth example
will allow readers to be much more confident about the strengths of the results,
that is, that these report something that is not just a singular case but something
that can be observed across a range of cases.
 DOCUMENTARY METHOD OF INTERPRETATION 177

intensity? The later questions stated relationships between two variables that
could be tested experimentally, whereas the first question lacked such clarity
about the variables to be studied.
My analysis of the videotaped planning sessions also showed the initial diffi-
culties students had with identifying variables.4 Jonathan, David, and Jeff had
investigated the deceleration of a frictionless cart by a spring that generated the
idea to study the oscillations of a spring. The following transcript resulted from
their subsequent planning and design session. This excerpt depicts how the stu-
dents came to formulate a question for research and how they came to identify
the variables.5

Jonathan: You just mentioned the timing . . . to the full stop? ((He has a spring
in his left hand, simulating oscillations by pulling the other end of
spring with his right hand.))
Teacher: No, the time for a full cycle, like . . . clock . . . clock . . . clock
((with a pen in the right hand simulating the movement of an oscil-
lating object through a photo gate-timer))
Jonathan: Does the period decrease? ((Stretching a spring between his hands,
small oscillations around an imaginary equilibrium.))

4
Before that I write that the students began with unfocused questions when they
were unfamiliar with the context of their inquiry. After having provided descrip-
tions of the documentary evidence in support of the assertion generally, I now se-
gue into the description of the difficulties identifying relevant variables that stu-
dents face initially, during a phase when they are still unfamiliar. In fact, the
students are in a similar situation as the researcher: from the contextual particu-
lars, which serve as documentary evidence of some phenomenon that can be re-
searched in terms of specific variables, they have to identify such variables. Just as
the researcher needs to read the data to become sufficiently familiar with these to
identify both documentary evidence and the phenomenon that it is a manifestation
of, these eighth- and eleventh-grade students had pains to get started on identify-
ing a phenomenon and ways of structuring it from the contextual particulars that
in themselves are not parsed into variables or documents of something else.
5
The transcription that follows is an actual document. But when the analysis be-
gins, it is not evident that this excerpt is document of something; it may actually be
read as documentary evidence for a number of things. The beginning researcher
may experience difficulties because the excerpt does not have a sign saying, Cut
here! There was also no sign for me that stated the phenomenon, Students have
initial difficulties with identifying variables, which I could have taken as a di-
rective for identifying cases in the database. The situation described here is similar
to the one faced by the instructor-analyst in part B, who had some transcription
arbitrarily chosen from somewhere in a recorded video without an indication of
what the people involved were doing. In the research on scientific process skills, I
had at my hand the videotapes and transcriptions from the eighth- and eleventh-
grade classes and had to identify segments that would be documentary evidence
for something that I did not know what it was at the outset.
178 CHAPTER 10

David: What would it be to measure . . . to measure. . . . An example of it

would be if you take a spring and another on the bottom if the spring
is really stretched out, if it goes really quickly . . . ((shows a
stretched spring by holding his two hands about 1 meter apart, the
lower one oscillating quickly)) . . . if it wasnt stretched out very
much? ((Hands about 20 centimeters apart, the lower one showing a
lower frequency than the first time.))
Jonathan: ((Jonathan takes spring, suspends the cover plate of his calculator
on the spring and lets it oscillate. The others watch. Jonathan is
about to change cover plate with a pen.))
Teacher: Now, if you take . . .
Jonathan: ((Suspends the pen from the spring. Makes it oscillate.))
Teacher: . . . less mass, . . . how is the period?
Jonathan: But it doesnt go that far.
Teacher: No, it extends the spring farther. It has the same distance to oscillate,
but the spring is less extended. . . . ((Showing oscillations with right
hand.))
Jonathan: When the mass goes up, it goes slower.
Jeff: Lets do it, . . . look at mass and period.

The excerpt shows that in unfamiliar contexts, experiments were planned by

describing situations instead of through the relation of two or more variables.6
The students took a spring (a small version of the slinky toy) and suspended
various objects from it to compare the oscillations. They did not start with a def-
inite idea about the variables; but through their conversation the students nego-
tiated the isolation of the two variables, mass of the object hanging from the
spring (independent variable) and period of the resulting oscillator (dependent
variable).
As the familiarity with the context increased, the students identified more
variables and defined these more precisely.7 In their biology experiment, eighth-

6
This sentence states what can be seen in the transcription; it can also be read as
an instruction to see in the transcription what the sentence itself asserts: students
do not identify variables but describe situations that they want to investigate. The
remainder of the paragraph then states how the students, in and through talking
about the situation that they wanted to investigate, identify one or the other aspect
as something to be varied while holding everything else constant.
7
In the preceding paragraph, I describe what could be observed early during a
student investigation, the ways in which the variables to be investigated emerged
from general descriptions of phenomena out of which aspects were identified for
investigation. In this paragraph, observations towards the end of an inquiry are
reported. The two paragraphs, therefore, provide contrasting observations. It is
this contrast that constitutes documentary evidence for the phenomenon denoted by
and described in the assertion. We may read the first sentence as stating a rule.
The remainder of the paragraph provides a description how in the case of one
group of eighth-grade students this rule comes to play itself out. The rule, however,
is an observer construct and does not elaborate the sense of the student-participant
 DOCUMENTARY METHOD OF INTERPRETATION 179

grade students investigated selected biomes on the school grounds. For exam-
ple, one group of students focused on the moisture levels of the soil. In their ini-
tial experiment, they examined if a relationship existed between the slope and
the moisture level of the soil. Gradually, the number of variables affecting the
moisture levels in the ground increased. By the end of the study unit, this group
of students had investigated the effects of surface temperature, ground cover,
soil composition, percentage of organic material, filtration rates of the soil, and
amount of light received by the soil on the moisture levels.
Any problems with this approach to student research occurred mainly during
the first two weeks. Most students were very enthusiastic about the freedom to
choose their own focus questions and were bubbling with ideas from the start.
However, there was also a considerable number of students (about 25% of each
class in both the eighth-grade science and the eleventh-grade physics programs)
who were anxious, wondering how to go about the work. Many of these stu-
dents approached the respective teacher asking what he wanted them to do. The
problem was to get these hesitant students ask their first question. Once the stu-
dents had worked through their first problem on their own, with encouragement
from the teacher, and by conversing with other students, they developed their
second and further question with less hesitancy. In another study to be reported
elsewhere, I found that only 3 out of 46 grade 11 physics students wanted more
direction and guidance for each lab.8
(May 24, 1991, Process skills 2)

So how did I do it? How did I get from having at my hands a lot of videotapes
and many pages of transcriptions to an articulation of an assertion for which parts
of the database constitutes the evidence in which some assumed phenomenon doc-
uments itself. Given that the study was written up for and published by the premier
journal in its field (science education), the method of arriving at the assertions had
to be at least sketched in some way. This description needed to be sufficient for the
reviewers to guarantee the quality of the study according to the perceived standards
of this field of researchwhich was just opening up to qualitative research after
decades of quantitative orientation. Although the excerpt from the methods section
correctly describes at some more abstract level what I had actually done, and, in
the peer-review process had been accepted as an adequate description, it leaves
open a lot with respect to the how (exactly) did this occur:

Data Analysis
The videotapes of the laboratory activities, the teachers course of studies, and
their reflective notes were the main sources of the data for the study. The re-
searcher viewed and re-viewed the videotapes as soon as possible after they

in the research. Thus, I had derived this rule through a process of creative abduc-
tion.
8
The final paragraph of the section pertaining to Assertion 1 does not actually
constitute documentary evidence for the phenomenon. Instead, it describes the
ways in which students were going about their work in a context where the teach-
ers had asked them to design their own investigations.
180 CHAPTER 10

were recorded by the laboratory and teaching assistant. Tentative assertions

were formulated to direct further data collection.9 After the first three weeks of
the study, both researchers met to jointly view the tapes. The initial assertions
were refined, modified, or discarded on the basis of the data.10 From this session
emerged new data collection strategies with the intent of gathering additional in-
formation to support or refute assertions.11 The researchers independently ana-
lyzed the written artifacts and then constructed assertions following a process
similar to that for the tapes.12 In addition to the meetings, the researchers com-
municated amply throughout the study by using an electronic mail facility. For
the present study we decided to examine students development of five higher
order process skills: identifying variables, identifying hypotheses, operationally

9
The first sentence states what I was reading at the time. There is then a jump in
the description, because the second statement states that tentative assertions were
formed. How did the researcher (I) get from the data to the assertions? How did
the reading and re-reading the transcription, the viewing and re-viewing of vide-
otapes, lead to the formation of a tentative assertion?
10
How were the assertions refined and on what basis? New and seasoned re-
searchers alike may want to at least record in their field notes / laboratory note-
books when, how, and on what basis they are changing an assertion. In this way,
there is a trail of decision-making that could at least conceptually serve for making
an audit. I date all my entries and then label the field notebooks in terms of the
time period that each covers (Fig. 10.3), thereby providing an audit trail that can
be used to trace what I have done and how I have done it while conducting re-
search and writing articles.

Fig. 10.3 Dated field notebooks constitute an audit trail that could be used to trace re-
search findings to their origins.
11
The process appears to be, in essence, not so different than what the instructor-
analyst is doing in part B. However, this similarity may only be at the descriptive
level.
12
At the time, I was concerned with the legitimacy of the kind of work I was doing:
a high school teacher doing research, among others, in his own classroom. I there-
fore conducted and wrote up the study and then asked a university-based science
educator to read and evaluate what I had done and to ascertain the appropriate-
ness of the assertions given the data.
 DOCUMENTARY METHOD OF INTERPRETATION 181

defining, designing experiments, and graphing and interpreting data. However,

our initial viewing of the tapes showed little evidence of definitions in terms of
operations. We decided to look at evidences for the use of definitions of objects,
events, or actions.
(Dec 29, 1990, Teaching process skills)

While doing the analysis, I did in fact go through cycles of writing about pieces
of data, seeking further evidence for what I had described, and modified what I
was writing based on subsequent readings. For example, an early research note
pertained to the inappropriateness of the science process skills as categories of
analysis. The example used in the research note would again constitute the doc-
umentary evidence that said something about what is named in the title of the
research note: process skills as categories of analysis. The first paragraph states the
general sense that emerged while I was reading through my materials, the second
paragraph constitutes an example in support of the general description, and the
third paragraph provides a reading of the example. This example was not so much
intended for the future readers of the article I might be writing but more for myself
as a way of documenting what I was seeing, how I was seeing it, and what to make
of some aspect of the data sources. I use words or short phrases to characterize
what the little analysis seems to be saying or be about, and, in so doing, establish
first categorization attempts.

Process skills as categories of analysis

Structuring the description of activity in terms of the process skills proved to
be less advantageous. Process skills did not appear in isolation and sequentially,
but often in a complementary relationship. This pointed to the unsuitability of
using the process skills as categories of analysis. The following example illus-
trates the point:
A group of students measured a velocity-time graph for a rubber stopper fall-
ing through water. The graph indicated that after the stopper had reached
terminal velocity asymptotically, the value suddenly jumped to a lower value
of terminal velocity. The students repeated the measurement twice and saw
that the pattern was consistent. They then went to change the timing appa-
ratus, but achieved a similar result. Finally, one of the group realized that the
stopper tipped over from tapered end first to the butt end. When they fixed
the stopper so that it could not turn over during its descend, the terminal ve-
locity did not change.13
In this part of their activity, the students obviously interpreted, formulated
models, recognized variables, formulated hypotheses, used space-time relations,
measured, communicated, and observed without that a specific skill was neces-
sarily shown in an overt manner. The students, because they did not have to

13
In observations such as the one described here, I felt that what I knew to be
common sense in the community was not bearing out: a sequential use of scientific
process skills as it had been described in the science education literature prior to
my own research.
182 CHAPTER 10

graph by hand had the option of discarding a set of data, or saving them on a
disc.
(Nov 18, 1990, Assertion material)

Readers may note that the source file had assertion material as its name. To-
day I write research notes. These provide me with a space where I can write any-
thing without having to worry about someone else reading it. These notes become
the materials from which I compose an article once I have the sense that I have
something to be reported back to the research community. It is out of such research
notes that I eventually arrive at what is going to be published. Prior to that I do a
lot of writing, analyzing, and re/naming sections of text.
It seems apparent that there is a dialectical process at work between the data
sources I have at hand and what is already known in the research community
where I ultimately report some results not yet known during the initial stages of the
research. What I can write about and contribute to the community is a function of
that community and of the data I have. Without knowing who I am doing this re-
search for and who my readers are, I do not know what to look for in the data
sources. Without an intimate knowledge of the data sources, I do not know what
there is to be reported to the community.

The Documentary Sense: How We Know and Learn Cultural-

Historical Things

Garfinkel refers to sociologists, who, at one point, do not know about social struc-
ture but then, at another point, do know about it. The author refers to the documen-
tary method of interpretation by means of which this change comes about. He does
not explicitly refer to knowing and learning, concerns that we might find in the
learning sciences. We may take a broader approach than Garfinkel and view the
documentary method as the manifestation of a particular epistemology that de-
scribes learning, and therefore knowing, in general (Roth 2013d). In Garfinkels
use, the document and what it documents (stands in for) may be related as the con-
cept and its example in a Kantian approach. The notions of epitome and underlying
pattern both allow such a reading. In my own work, I pushed the idea of knowing
existing only in the form of manifestations so that the underlying pattern or con-
cept is known only concretely such that there is a partwhole relation between the
presently considered manifestation and all the manifestations that have been expe-
rienced previously, including the changes in the manifestations upon manipulation
of the object or event. This has empirical consequences for thinking researcher and
researched. I elaborate the history and nature of this idea in the following para-
graphs.
The analysis of how we experience geometrical objectse.g. a cubeshows
that we never perceive the geometrical object, such as the six-sidedness of a cube,
the eightness of its vertices, and the twelfth-ness of its edges (Merleau-Ponty
1945). Instead, we perceive a cube always from a point of view, which reveals
particular aspects. Changing the point of view either by turning the object in our
 DOCUMENTARY METHOD OF INTERPRETATION 183

Fig. 10.4 Network growth implies not only accretion of new nodes but also changes in the
older nodes and the relations these have with other nodes.

hands or walking about it brings about a change in perception. Knowing an object

as a cube implies knowing what will happen if the point of view on the object is
changed. As a result, we do not know the object in some abstract way but in the
form of correspondences between our body movement and the associated changes
in perception. The unit of all the different perceptions is tied to the unity of the
body, grounded in a primordial I can: action and perception are inherently corre-
lated rather than the result of a mental construction. We know the cube not because
of its abstract geometrical properties but in and through the ways in which it mani-
fests itself to us: our touching and gazing.
In the preceding object, we considered the way in which we know some materi-
al thing. Taking into account the suggestion that we ought to approach social facts
in the same way as we do material facts (Durkheim 1919), I suggest thinking
(about) and theorizing social phenomena such as a queue (line-up) through the way
in which we actually encounter it: through the manifestations of what people call a
queue. This has consequences for thinking about knowing and learning of social
scientists in the course of research specifically and of persons at all stages of their
lives generally. The proposal is to think about knowing some object or phenome-
non as a whole, which always only manifests itself in a particular way. If we have
experienced only one manifestation, then the whole is shaped by this one experi-
ence. As we experience a different manifestation of the same phenomenon, the
whole changes because not only has a new manifestation been added but also be-
cause of the relations to the other manifestation(s). We can think of this in terms of
a network in a continuous process of formation (Fig. 10.4). Because of the rela-
tions to other manifestations (nodes), not only does the whole (network) change
but also all the other manifestations that are understood in their relations to the
whole as well as in their relations to all other parts.
Let us concretize with the example of a queue. At some point in time, we expe-
rience a queue for a first time, for example, while accompanying our parents to the
supermarket. Each time we line up, the phenomenon manifests itself in a new but
not very different way. At some later point, we may want to take a form of public
transport, such as a bus. Here, queuing manifests itself very differently then in the
supermarket, where the arrangements at the check out force people into lines:
depending on the culture, we may find very orderly queues (UK) or a sort of piling
up and a lot of elbowing to make it onto the bus. Later we may queue in an office
or shop where tickets have to be taken, and, in this way, experience a radically
different manifestation of a queue, which may no longer be visible in the way actu-
al line-ups are.
184 CHAPTER 10

Our participation in making a queueour knowledgeability with respect to

queuingis the cumulative result of experiencing a queue in all of these concrete
manifestations. It is the cumulative result of experiencing manifestations of queues
always and only in concrete form. More importantly, perhaps, this does not require
inferring explicit rules for queuing, just as we learn to competently speak a lan-
guage without any explicit grammatical rules. In fact, speaking more-or-less com-
petently is a prerequisite for learning grammatical rules, because without this pre-
requisite, there is nothing we can grammaticalize, that is, form rules about. We
may take an artificial neural network in the process of learning a language as an
interesting analogy. Such a neural network learns language simply by being ex-
posed to it (Elman 1993). Without being told (by a programmer who embeds rele-
vant procedures) any grammar and without extracting any rules from the sample
sentences it is exposed to, the network eventually treats certain input words as
nouns, others as verbs, yet others as adjectives. Moreover, it makes differences that
we think about in terms of the subject of a sentence and an object, direct or indi-
rect. That is, the artificial neural network learns a language without having any
explicit grammatical rules encoded in it: it distinguishes the words as it is exposed
to them.
Another way of exemplifying this is by means of the idea of learning by analo-
gies. The following serves not only to suggest how we learn and know but also as
an analogy for doing research. Some time ago, I was doing an empirical investiga-
tion of learning chaos theory in a tenth-grade high school physics course (Duit et
al. 1999, 2001; Roth and Duit 2003). In this unit, students come to interact with a
range of situations all of which were, from an instructional point of view, analo-
gous to each other in that they exhibited some essential feature of a chaotic system:
a magnetic pendulum (Fig. 10.5a), a (two-dimensional) computer simulation of a
system with three attractors (Fig. 10.5b), a chaos bowl in which a steel ball exhibits
trajectories similar to the pendulum (Fig. 10.5c), and the drawings of a ball moving
along or over a ridge (Fig. 10.5d). The classical way of thinking about students
encountering and discussing these situations is in terms of the analogy, which ex-
ists between the deep features of the four situations. That is, learning is thought
in terms of the extraction of a commonality that all four situations sharejust as
the Kantian definition of a concept/category that collects all situations that share a
feature in common. The documentary method makes us think about the situation
differently. Here, the learner knows only the manifestations, and the (practical)
knowledgeability always and only exists in the form of familiarity with these con-
crete manifestations. This is so even if the learner were eventually to advance to
such an extent (e.g. by studying physics at the university) that s/he could write the
equations for such a system. The equations would be just another manifestation of
the phenomenon rather than an abstraction thereof.
Readers may want to ask: Does it matter whether we view learning one or the
other way? It actually does. In the classical way of viewing, learners (investigators)
have learned when they abstracted the common deep feature and, therefore,
know the four situations in Fig. 10.5 to be analogous to each other. Once under-
stood in this way, learners (knowers) extend this knowledge to new situations (un-
less they are incapable of transferring this knowledge to new situations). From
the documentary method point of view, the new situations would extend the phe-
 DOCUMENTARY METHOD OF INTERPRETATION 185

Fig. 10.5 Four situations with the same underlying structure. a A magnetic pendulum. b
The traces of a computer simulation of the magnetic pendulum. c A chaos bowl. d Draw-
ings of a ball rolling along or over a ridge.

nomenon, which now exists in even more manifestations, and the relationship be-
tween the different manifestations actually changes the ways in which each con-
tributes to the overall phenomenon. That is, the way in which the phenomenon is
known exists only in and through its many different manifestations. In our re-
search, we had used a number of new situations during interviews with the stu-
dents, including the dice and the Galton board14, and talked to students about other
situations (including lottery and roulette) (Fig. 10.6). In general, the students did
not automatically extend what they had learned using the original situations (Fig.
10.5) for talking about the new situations. My classically trained and thinking col-
leagues tended to say that the students failed to see the analogy or failed to
transfer, whereas the documentary approach leads us to understand that the stu-
dents are incrementally building up the phenomenon from ever new manifestations
what they are told are comparable situations.
We can use this way of theorizing knowing and learning to think about what the
instructor-analyst was doing and how he has done it. Each statement someone
makes in the transcription, each turn pair, is to be thought of as a manifestation in
the way that the tenth-grade physics students encounter a new situation as a mani-

14
The Galton board is a board with nails in the arrangement shown in Fig. 10.6. Steel balls
falling from above onto the nails from a reservoir will undergo trajectories that cannot be
predicted for any single ball. However, in the end, after a large number of balls have fallen,
a predictable distribution will be observed.
186 CHAPTER 10

Fig. 10.6 Three new situations that may be viewed as sharing an underlying commonality
with the situations in Fig. 10.5: dice, Galton board, and lottery or roulette games.

festation of something that they do not come to know other than in and through
these manifestations. Even though these manifestations differ substantially, and
precisely for this reason, they determine the phenomenon in its fullest. I suggest
that new and experienced analysts can be thought of in a way analogical to the
students, who do not know the phenomenon. The instructor-analyst does not know
the phenomenon as he works from line to line through the transcriptions. His anal-
ysis is a form of learning from ground zero to the point where he names or de-
scribes the type of situation that might have been at the origin of the talk. Whereas
Garfinkel is interested in how people find out about something, we may think
about the instructor-analyst specifically and all social analysts more generally as
being in a process of learning. Through the process of learning they come to know,
come to be knowledgeable about, whatever it is they will know once everything
has been said and done. At the beginning, the analyst does not know what s/he will
know in the end and, therefore, will not be able to actively orient towards their
ultimate finding. They have to take their current knowledge/understanding as a
project*-in-the-making the nature of which is itself shaped by the process.
We may use the documentary method as a learning process, which leads to
knowledge defined in terms of all the concrete manifestations that we have experi-
enced with respect to a particular phenomenon, those individuals doing research
appear to go about such inquiries more persistently or in the context of problems
that subsequently lead to what are recognized scientific findings. How do we
teach rigorous data analysis designed to arrive at articulating something that had
not articulated before? How do we, educational, sociological, or anthropological
researchers learn analyzing data rigorously by means of the documentary method
of interpretation? How does an interviewer find out about the reigning attitudes
among the members of a particular group? For each of these to happen, the investi-
gator has to select among alternative courses of interpretation and inquiry to the
end of deciding matters of fact, hypothesis, conjecture, fancy, and the rest (Gar-
finkel 1967, p. 77); s/he has to make such selections even though s/he cannot know
the end and, therefore, cannot know what s/he does or has to do to get there. At any
point, therefore, the investigator has a developing sense, which pragmatist philoso-
phers have come to refer to as a passing theory (Davidson 1986). The questions
then include: How do we arrive at passing theories? and How do we teach arriving
at passing theories? The pragmatists response is: There are no rules for arriving
 DOCUMENTARY METHOD OF INTERPRETATION 187

at passing theories, no rules in any strict sense, as opposed to rough maxims and
methodological generalities (p. 446). Here Davidson uses passing theory in the
sense of the momentary status of the inquiry about what another person is thinking
about. Such a passing theory, like any other theory, is derived by wit, luck, and
wisdom from a private vocabulary and grammar, knowledge of the ways people
get their point across, and rules of thumb for figuring out what deviations from the
dictionary are most likely (p. 446). The difficulty then is that the procedure cannot
be taught in the same way that it is impossible to teach coming up with new theo-
ries in the sciences. There appears to be however a sense that the more we engage
in such inquiries, the better we get at seeing things. In my experience, those indi-
viduals who have done a lot of (interaction) analysis are better at doing analysis,
attending to and highlighting more detail, generating more possibilities.
11

Getting Time Back into the Analysis

Time and temporality are at the heart of what it means to be humancaptured not
in the least by the conjunction of Being and time in the title of a philosophical
work that has influenced scores of social scientists, including individuals from
cognitive science and artificial intelligence (Heidegger 1977). Yet time and tempo-
rality are the forgotten dimension in much of social science research (Roth et al.
2008). Two areas where this neglect of time is quite apparent is in (a) research on
(mis-) conceptions, conceptual frameworks, abilities, beliefs, motiva-
tions, and the like and (b) the approach to analyzing classroom or meeting talk
with hindsight. Thus, when analysts take an interview transcription to distill from it
what a person thinks and believes or what his/her motivations and abilities are,
they flatten something that has unfolded in time and treat it as if the entire inter-
view had been simultaneous. This conflation is associated with a separation of talk,
attributing the words, phrases, and language to individuals and treating the text as
the sum of independent components (i.e., there is no relation between interviewers
questions and the interviewees responses). Because time has been eliminated, the
interview talk now is attributed to some underlying belief, knowledge state, con-
ceptual framework, conception, or motivation. Moreover, such research eliminates
any temporal relation that we experience when actually participating in events
where we do not know what will be happening minutes and even seconds hence.
To grasp the inner dynamic of social eventse.g. conversations, classroom talk,
business meetings, or grant funding decision-makingwe have to get the temporal
relations back into the analysis. In any case, even if we wanted to continue doing
research of the classical type, the honest researcher would explicitly articulate the
assumptions required and made when time is eliminated from analysis.
190 CHAPTER 11

Where the Future Meets the Past

In a strong sense, there is no present as an extended something. There is a past that

extends from the now backward; and there is future extending forward from the
now. The present lies at the intersection of the two. Because it is the limit point of
past and future simultaneously, like a point on the border between two countries, it
belongs to both. This requires us thinking the present syncopically, always follow-
ing what has been said and done up to a point and always preceding what is to
come thereafter (Roth 2014b). This has consequences for the ways in which we
analyze data, as I show with an excerpt from an article about how students and
their teacher talk about scientific phenomena in terms of models. The idea about
the present as a syncopic point belonging to two different orders simultaneously is
depicted in a diagram re-presenting turns 216218 from the article (Fig. 11.1).
When the teachers Saying has come to an end, the phrase Whose variable will be
this? will have been heard. Before that end, we cannot say what the teacher is
saying, for what he has said, the Said, is available as something that can be grasped
only when the Saying actually has ended. The nodal point on the boundary be-
tween past and future is virtual, for it already has passed by the time we have
grasped it as a present.

[S]tudents presented and described the models they developed to their class-
mates, usually prompted by the teacher as his response to students ideas (con-
tent) or their reasoning, with the explicit purpose of evaluating the models pre-
sented. The teacher refrained from evaluating their ideas or reasoning himself.
When students identified problems in their models, they tried to come up with
solutions that would resolve that problem. Due to the nature of their models at
this stage, the solution was usually the development and use of an algorithm that
defined a physical entity or a physical process. When students provided such a
solution, the teacher asked them to clarify their ideas (prompt) by describing
these step-by-step. . . .
For example excerpt 6 below begins with the teacher summarizing a long
conversation during which students talked about a model in which they wanted
to show a boat trying to move in the opposite direction to the water flow in a
river, but the model they were viewing at the time did not have that effect.
Achilleas then provided a possible solution that included the use of a variable
(that would be the speed of the boat due to the paddling in the opposite direction
to the water flow), and that would be subtracted from the speed of the boat due
to the water flow, thus describing an algorithm that defines relative speed. . . .

214 Teacher: So, you said that on the one hand we need the boat to move back-
wards, but in this model it moves forward? What we shall do about this?
215 Achilleas: Maybe we can create another variable, and have this variable affect
the motion of the boat in some way, so that its speed will be twice more than
the speed of the water.
216 Teacher: Ok. Lets take it one part at a time. You said create another variable,
right? Whose variable will be this?
217 Achilleas: The boats.
 GETTING TIME BACK 191

Fig. 11.1 Once the teacher has come to what will have been the end of his turn, there is an
open horizon of who will begin to speak, what about, to whom, etc. There are many possi-
bilities for what will be next, whereas the past has factual and definitive nature.

218 Teacher: Ok. The boats. And what exactly will this variable be for?
219 Vikki: [We will use it] to subtract from the water speed.
220 Achilleas: . . . from the canoes power.
(Louca et al. 2011, p. 941)

When the teacher (or anyone else for that matter) has come to what will have
been the end of the turn, we do not know what comes next. The teacher, too, does
not know at the end of turn 214 who will talk next and what will have been said at
the end of his/her turn. The teacher, therefore, cannot know whether he will have
had a turn 216 and what he will have said when the turn has come to an end be-
cause there is a pause and someone else will have spoken next. But from the per-
spective of the event that we observe, what the speaker has done is available only
when we know the effect. But this effect lies in the future, for example, when we
place ourselves right at the end of What we shall do about this? (turn 214). At
that point, right after the ringing of this, we do not know what will be next. In
fact, in the materials made available by the article, we do not have a precise tran-
scription and do not know about the temporal unfolding. The authors make it ap-
pear as if Achilleas was talking immediately after the teacher. At a first level, the
removal of all temporal information within and between speakers constitutes a
reduction that no longer allows us to see important dimensions of the relationship
itself: how it is organized. Consider the more precise description of an exchange
involving the teacher Jeannie and her student Mario in a mathematics class. Many
analysts would claim that a question had been asked at the end of turn 056. Yet
what we see unfolding thereafter is a pause. A pause is a truly collective phenome-
non, because it takes both Jeannie and Mario to produce it. It turns out that Jeannie
is not at the end of her talking, for she is actually the next person to speak after
what was possibly the end of her turn. In Fig. 11.1, a pause, a student talking, or
the teacher herself continuing all count among the possibilities that the future holds
in store. Whatever happens, only one of those possible futures comes to be real-
ized, not only annihilating this possibility but also creating new possibilities.

Fragment 11.1
056 J: kay (.) first week (0.84) wHY (0.16) wOUld (0.75) there be? (0.91) why
would there be:::sIX (0.61) dollars in the piggybank.
192 CHAPTER 11

Fig. 11.2 Following the teacher statement Whose variable will be this?, there are many
possibilities for what will happen next. One is realized, thereby annihilating all other possi-
bilities and creating new ones.

057 (1.04)
058 for the first week. (.) what did you ge::t (0.37) to do ((she takes the goblet of
week 1))
059 (0.75)
060 M: becau::se
(Roth and Radford 2011, p. 51)

Jeannie speaks again, or continues to speak (turn 058) until there is another
pause unfolding (turn 059), which comes to an end with Marios beginning to
speak (turn 060). Here, the pausing turned out to be shorter than in the preceding
case. In fact, had it come to be longer, at any point thereafter, it might have be-
come more probable for Jeannie to take yet another turnan anticipation not so
unlikely when we investigate classroom talk, especially in situations characterized
by a lot of teacher talk, brief student replies, and evaluations.
The authors describe turn 214 as summarizing a long conversation in which a
model was presented, and then suggest that Achilleas provided a possible solution.
The authors also claim that the teacher did not evaluate what students were saying.
To begin with the last claim, in saying What we shall do about it? after stating
that the boat was to move against the current (backward) but did not do so in the
model. We have here not only an invitation to do something about the discrepancy
but also an evaluation that something should be done about it. Achilleas, in begin-
ning to speak, responds first of all to the invitation; when he starts to speak, no-
body knows what will have been said when he ends. That is, the inner dynamic of
the classroom talk, its temporal unfolding that arises from the back-and-forth has
been eliminated in the quoted analysis that sets side-by-side (or in sequence) the
content of the teachers statement and the content of Achilleass statement. The
teacher also produces an evaluative term when he replies to Achilleass turn by
saying, Ok. Lets take it one part at a time. He not only states projectively to take
one part at a time but also that whatever has preceded did not take the issue one
part at a time and, therefore, that it ought to be taken one at a time. Similarly, in
stating What will be this variable for? (turn 218), the teacher connotes that the
 GETTING TIME BACK 193

Fig. 11.3 The teachers statement Okay. The boats accepts and positively values the
preceding turn, which was offered in reply to an invitation.

purpose or nature of the variable had not previously been stated. Let us return to
the issue of the temporal unfolding.
In turn 216, after having stated to take one part at a time, restating to Achilleas
what he has said, the teacher articulates what will have been an invitation to state
the owner of the variable. We do not know whether there was a pause and how
long it was. But, as suggested above, one of the possibilities that exist at this point
is that the teacher will be speaking again; another possibility is that someone other
than Achilleas will take a turn (Fig. 11.2). There might be a counter-question of-
fered: What do you mean by . . .? In Achilleas taking the turn an stating the
boats, one of these possibilities is realized so that it no longer exists as possibility
but now is a matter of historical fact. As Fig. 11.2 shows, the conversation begins
to map out a definite trajectory while allowing the virtually unlimited number of
other possibilities for the unfolding of the conversation to disappear.
Again, we do not know whether there are pausesfor example, the research on
wait time II suggested teachers may allow more than one student to respond prior
to taking another turn. Here, it is the teacher to whom falls the next turn, which
will be an evaluate one: There is an acknowledging interjection Okay, followed
by an (approving) restatement of the preceding reply the boats to the earlier
invitation (Fig. 11.3). The representation shows how the classroom talk is tracing
out a trajectory in a field of possibilities that open up and disappear again. Now we
are at a stage where the problem of traditional research approaches in the analysis
of verbal exchanges has become literally visible.
In traditional analyses, what has actually been said and done is read through the
lens of the final achievement. That is, researchers position themselves at the right-
most dark circle in Fig. 11.3 and then read the transcription as if it arrived at this
point by some inherent necessity. For example, the reading will then suggest that
the teacher asked a question in turn 216, which led to the statement of the boat
and some variable that would be subtracted from the speed of the boat to account
for the water speed. Readers certainly will note that Vikki has actually said the
opposite: to subtract from the water speed rather than the boat speed and Achilleas
makes a statementwhich we do not know how it is taken upconcerning the
194 CHAPTER 11

boats power, which allows a researcher familiar with the history of physics to
think about an Aristotelian language for describing relative motion rather than a
Galilean one. By reading the unfolding classroom talk in terms of what it ultimate-
ly led to, by flattening the Saying to the Said, all historicity and contingency has
disappeared. In school settings, students are evaluated on the basis of the out-
comes, which represent but an infinitesimal number of possible outcomes. Thus, in
one research project I showed how many different designs of an earthquake-proof
building students considered in the course of their design conversation, only one of
these was actually realized (Roth 2001). But that history of a contingent evolution,
which is so important to designers themselves, is lost when the transcription is read
through the lens of the final result. All development has been eliminated from the
analytical possibilities and the represented classroom talk is a mere tracing out of
conceptual ground. As a result, the temporality and open-endedness of human life
has disappeared.
As we see in the Roth and Radford transcription, pauses, too, matter; and these
pauses need to be grasped and theorized in terms of joint social action as any other
contribution. In presenting the transcription of a lesson in the way that the Louca et
al. (2011) paper exemplifies, there is another reduction observable. It makes it ap-
pear as if conversation are all about content and little to nothing about the ways in
which we make the relations that allow any content-related dimension to be. At
this point then we are doing the data analytical equivalent to Monday morning
quarterbacking, a practice in which football games are analyzed with hindsight
(There are a lot of s/he should have . . . ). Monday morning quarterbacking and
its equivalent in qualitative research has the problem that it does not take into ac-
count the nature of everyday practical actionwhich does not have time out,
brooks no delay (Bourdieu 1980). When the timing is not right, then there might be
troublee.g. an offense may be noted if a gift is not accompanied by a counter-gift
in some traditional culture or if a teacher does picks on a student (e.g. Roth et al.
2001). In one and the other situation, an escalation may follow, where a situation
literally is experienced as out of hand. Any form of action, even a pause, is and
becomes a resource for current and future actions so that a teacher might be found
to talk even though a student may have been anticipated to take the next turn (as in
the Roth and Radford transcription above). Once we grasp escalation as a phenom-
enon, we also understand in a better way all those situations that do not escalate
because the actors collude so that the situations unfold in more rather than less
predictable ways and, assessable after the fact, consistent with this or that (social)
norm or (school) regulation.
The issue of reading conversations specifically and life generally from an a pos-
teriori perspective with 20/20 hindsight affects research in another way as well, for
example, the ways in which we have to take what research participants say during
interviews. Participants, too, talk to researchers knowing where a course of action
specifically and life generally has taken them. They now look back and not only
talk about the actual events (conversations) in terms of the outcome but also about
what they should have said or done in its course to come closer to the desired out-
come (Fig. 10.4). In the present example, the reply to the question starting Whose
. . . ? denoted a thing (the boat), whereas a similar situation in chapter 9 led to the
statement of a persons name (the teachers husband). In either case, teachers
 GETTING TIME BACK 195

Fig. 11.4 In a post-event interview, the interviewee, knowing the actual outcome of a
conversation may be tempted to say that at one point she should have said/done something
different, anticipating that this might have achieved the desired outcome.

might subsequently talk about what they should have asked to arrive at whatever
the desired outcome is stated or presupposed to be at this point in the unfolding
interview. They would do so even though there is no guarantee that a different in-
vitation (question) would have come any closer to that desired final statebecause
of the inherent open-endedness of such talk that makes it impossible even for the
teacher to anticipate what s/he will be saying only seconds hence, codetermined by
the actual replies and the contingencies of the unfolding event.

The Making and the Taking of Time

Time and temporality are integral to the ways in which humans relatebut not in
the way these are often conceived. The general conception of time is that outlined
by Kant (1962), who conceptualized it, next to space, as one of the conditions of
knowing. As in relativity theory, time and space constitute something like a four-
dimensional box within which events are thought, marked by where and when they
occur. Yet when we participate in verbal exchanges with others, we experience
time and temporal relations differently. We experience them or ourselves as taking
time, but also as making time allowing the speaker to go on. In some situations,
two or more speakers speak simultaneously, whereas in others one can observe
long pauses. It has been observed, for example, that Hawaiian children tended to
underperform in reading achievement when attending regular school programs
i.e. programs in which the standards of taking turns at talk typical for the American
(white) middle class were enacted (Au 1980). These same children, scoring in the
regular program in the bottom 11% on standardized tests performed above the
mean when learning to read in a context that allowed them to participate in ways
typical for the Hawaiian culturelots of overlap and co-talk. That is, the timing
196 CHAPTER 11

of speaking during reading lessons was crucial to learning and knowing to read.
The converse may be the case in other cultural contexts. Thus, for example, long
pauses in conversation are common as Indian people have learned to be comforta-
ble with silence (DuBray and Sanders 1999, p. 73), which needs to be taken into
account not only in schooling but also during exchanges between health profes-
sionals and members of the various North American First Nations. Members of
these nations often do not participate in conversations when there is not sufficient
(temporal) space that allows them to enter. Attending to the timing of talk and to
the temporality of speaking, therefore, constitutes an in-road to grasping the ways
in which social relations are constituted and, therefore, the ways in which
knowledgepower comes to be co-produced and transformed. Let us consider a
few situations.
In Fragment 11.1, we observe longer pauses in the course of the teacher Jean-
nies speaking. As a reference, consider that in some studies teachers have been
reported to wait an average of seconds for an answer before going on to talk, often
providing the answer themselves (Tobin 1987); and in (telephone) conversations,
the standard maximum silence tends to be approximately 1 second. In the frag-
ment, we observe pauses that are of the same order of the maximum silences re-
ported in the literature: 0.84, 0.75, 0.91, and 0.61 (turn 056). Some readers may
want to argue that Jeannie had not finished her question forgetting that in many
situations, speakers do not (get to) finish a phrase: the ending of a phrase, a gram-
matical feature, itself is the result of the enacted relation (Roth 2010). Furthermore,
consider the rise in the pitch, normally heard as a question, followed by a 0.91-
second pause. When we continue reading/hearing, we in fact observe a repetition
of the why would there be that is followed by a completion, six dollars in the
piggybank (turn 056). In fact, we may hear it as the teacher taking the time to
formulate the nature of the question. The transcription features, as the next turn, a
pause that has the length of the standard maximum silence. Here, Jeannie is taking
time to formulate the question. But because pauses constitute possibilities for an-
other person to speak, it is not only Jeannie taking time but also Mario giving time
so that Jeannie may produce what grammatically is a complete phrase. Jeannie has
time because Mario is giving it by not beginning to speak. He could have started
talking about what he had already contributed by saying, beginning in any one of
the four pauses, I already told you . . ..

Fragment 11.2 (extract from Fragment 11.1)

056 J: kay (.) first week (0.84) wHY (0.16) wOUld (0.75) there be? (0.91) why
would there be:::sIX (0.61) dollars in the piggybank.
057 (1.04)

Grammatically, the phrase in turn 056 is a complete question. Interactionally,

there now is space for someone else to speak, Mario or one of the two girls at the
table, and, in so doing, constituting the nature of Jeannies phrase: question, state-
ment, invitation, solicitation, and so forth. As the event unfolds, we notice that
after 1.04 seconds, it is Jeannie who speaks. Thus, even though there is a possibil-
ity of hearing turn 056 as a question soliciting an answer, Jeannie speaks again.
How might a participant or onlooker gloss what has been happening here? This
 GETTING TIME BACK 197

requires us to consider what Jeannie is saying now (turn 058) and how it relates to
what she has said before (turn 056).

Fragment 11.3
057 (1.04)
058 for the first week. (.) what did you ge::t (0.37) to do ((she takes the goblet of
week 1))
059 (0.75)
060 M: becau::se (0.30) ((points to goblet 1))
[but there is nine ] <<dim> [the first week]>
061 J: [there is, you know ] <<crsc> [is it really ] six dollars?> ((points to and
looks closely at Marios first cell))

The contents of turn 058 may be glossed as an elaboration or clarification of the

question. It is as if the pause was taken to mean that Mario could not reply. In the
elaboration and clarification of the question, turn 058 thereby also attributes the
possible problem to the nature of the question-invitation offered. The content spec-
ifies that the preceding phrase pertains to the first week: Why would there by six
dollars in the piggybank in the first week? It can also be heard as an invitation to
retell or recall what Mario had done earlier, namely place six chips into a goblet
corresponding to putting six dollars into the piggybank as a starting amount.
Again, there is a pause and then Mario speaks, drawing out the articulation of the
latter part of the adverbial conjunction because. There is a brief pause and then
Mario and Jeannie speak at the same time (turn 60, 61). We observe that the se-
cond part of Jeannies turn picks up and refers to the first part of Marios turn,
where he articulates the number word nine, which now contrasts the six that were
part of Jeannies earlier phrase (turn 056).
In turns 060 and 061, two individuals speak at the same time. In some cultures,
this would be heard and experienced as vying/competing for the floor. In North
American schools, it would be deemed inappropriate, for speakers tend to be at-
tributed turns one individual at a time. Such turn taking clearly is enacted on
speaker panels in Canadian television newse.g. At Issue, airing as part of the
CBC The National nightly news programwhere one individual speaks, clearly
selected by the news anchor, who also intervenes should there ever be two people
speaking at the same time. The anchor also delimits the time other speakers have.
In French television news, on the other hand, anchor and invitee may be speaking
simultaneously, or, when there are several invitees, they may be found speaking
simultaneously. Moreover, some politicians manage to get extended turns at talk
despite apparent efforts on the part of the anchor to return to the speaking floor.
That is, there are substantial differences in the ways that conversations between
anchors and invitees come to play themselves out in different cultural contexts.
Time may be given or not given; and time may be taken or not. We may there-
fore gloss turn 057 as time that is given to Mario to respond, and also as being tak-
en on the part of Jeannie prior to what might be glossed as making another attempt
at offering a reply-inviting question. It is part of the invitation to provide a reply to
a question. Turn 059 constitutes time given on the part of Jeannie to Mario, who,
simultaneously, takes his time to reply. It is in these pauses that we may recognize
an important feature of the social relation. Depending on the culture, the time given
198 CHAPTER 11

Fig. 11.5 The dynamic of a conversation depends on its temporality, which arises from
the documentary sense associated with the time following a speech act.

and taken will be associated with respect and disrespect, with proper or improper
cultural protocol (e.g. Bourdieu 1980). This association derives from the docu-
mentary method of interpretation where something like an extended and develop-
ing pause is taken to be treated as, or a document of something else. We can for-
mulate this diagrammatically (Fig. 11.5).
Following what can be heard as a question, there is a pause unfolding. It pro-
duces possibilities for next turns, which themselves change as the pause is unfold-
ing. If there is a phrase that constitutes a second turn in a {question | reply} (or
{invitation | acceptance} or {initiation | continuation}) pair, then the verbal ex-
change continues (Fig. 11.5). If, however, the time continues to unfold without the
intended / designated speaker to take a turn, this is then may be treated as a non-
reply (non-acceptance / rejection or initiation / non-continuation). That is, the ex-
pressive sense layered above the objectively experienced lengthening period is that
of non-reply (non-acceptance / rejection, non-continuation). There is also a docu-
mentary sense, which then has consequences for the nature of the subsequent ac-
tion taken. Thus, a non-reply may be taken as the incapability (lack of ability) to
comprehend what is intended / offered as a question; or the non-reply might be
taken as a document that the question itself was not clear (Fig 11.5). In the former
situation, the teacher may engage in one form of action that addresses capability,
whereas in the latter situation the question itself might be clarifiedwhich is the
case that we observe in turn 058 (Fragment 11.3) or in turn 07 of Fragment 8.3 (p.
133). Because a reply is beginning in turn 060, the practical logic follows the lower
path in Fig. 11.5.
In the preceding example, we observe that both expressive and documentary
sense depend on the temporality of the event, the timing in the back-and-forth of
the verbal exchange. However, there are no fixed numbers that can be assigned to
the absolute amount of time, as measured by a clock. It is therefore deceiving to
use temporal measures for saying something about classroom interaction. It is
more important for an investigation to work out the pragmatic properties of tem-
poral relations and temporality.
Taking into account these considerations of time, we may observe the consider-
able reduction that has occurred in the lesson from a Cypriot classroom. The effect
of the reduction is a complete focus on conceptual issues that no longer depend on
 GETTING TIME BACK 199

real human beings that relate to each other to make possible any form of verbal
exchange. Even though Vygotsky complained 80 years ago that psychologists pre-
sent thought as if it was thinking itself, current research continues to present it in
this way. This presenting is associated with the methodological reduction that no
longer includes the concrete pragmatics of verbal exchanges, to which rigorous
forms of analysis closely attend.

Time, Ethics, and the Subject

Bringing time back into analysis is important because it not only allows us to com-
prehend the dehiscence of intended (planned) action and situated, practical action
but also the inherent ethical nature of being human (Roth 2013e). The representa-
tions used above (Fig. 11.1 to 11.3) make salient the continuously moving syncop-
ic intersection where past and future meet. Because of this movement, the effect of
an action or sayingthe what an action has done or a saying has saidcannot be
known until after the movement (acting, saying) has come to a close and therefore
can be apprehended as this or that action or saying. Moreover, to understand the
inner dynamic of the event, we need to seek evidence for this effect within the sit-
uation rather than outside it (e.g. in our interpretation). That is, rigorous analysis
seeks the effect in the articulations of those actually affected by the action or say-
ing. Once we attend to the temporality of unfolding events generally and unfolding
verbal exchanges more specifically, the fragility of human life becomes more ap-
parent, including the fact that we are not only agents but also, and simultaneously,
patients.
In chapter 7 I suggest that to understand the joint nature of (social) action, we
need to investigate turn pairs, which allow us to exhibit the co-determination of
action (cause) and effect. In this way, cause and effect come to be temporally dis-
tributed. In this chapter, we note the open-endedness of the future whereby one of
the possible futures is annihilated in its realization and others open up as a conse-
quence (Fig. 11.2, 11.3). If we take this into account, then we immediately under-
stand that actors / speakers do not know with any precision what the effect of their
actions / talk is until after the response can be grasped, that is, for example, until
after someone else has replied. Thus, in the locution the boats (turn 217), the
statement Whose variable will be this? comes to be reified as a the first part of a
{question | reply} pair and the locution itself as its second part. It is not the teacher
who determined that turn 216 will have been the cause effecting a reply. Instead,
the next turn made it such.
There are immediate and radical consequences with respect to thinking ethical
relations in verbal exchanges that explode all classical (Kantian) forms of ethics.
First, although in classical approaches the actor / speaker bears the responsibility
for the act / speech, our attention to the temporal delay and reverse relation be-
tween cause and effectthe latter allowing us to identify the formeran actor /
speaker inherently cannot know just what she is saying / doing. Because the effect
is in excess of what the actor / speaker intends, there is an inherent irresponsibility.
200 CHAPTER 11

Second, the recipient of the action / speech has to open up to receive (listen)
without knowing what is coming at him/her. In attending and listening to the other,
recipients expose themselves to the unknownwhich, in many situations, include
hurt. All we have to do is think about the verbal exchanges in the family circle,
school classrooms, at work, or online where someone expresses being bullied or
hurt in other ways by what someone else has said. To be hurt, or to feel bullied,
however, the person has to open up and, thereby, expose himself/herself. But per-
sons declaring themselves to be a victim assaulted by the words of the other also is
actively hearing these words in one rather than another way.
Recipients are patients. But as part of their response, when they speak or act in
turn, they also determine the nature of the preceding (speech) act. This makes pre-
vious speakers / actors patients, because someone else determines what they have
done. In saying something, I am exposing myself to someone else who determines
what I will be determined to have said and done. That is, interlocutors simultane-
ously are agents, acting upon and affecting the conditions, and patients, being sub-
ject and subjected to the conditions.
With this perspective, we have all of a sudden exploded what did not appear to
be but in fact is an ideologya whole system of ideas that shapes the ways in
which we approach verbal exchanges without taking into account operating sets of
relations, positions, and dispositions. When we rigorously take the temporal un-
folding of life generally and of talk more specifically, and attend to what we can
actually observe, the narratives of the agent and about cause-and-effect relations
come to be exposed as an ideological system to which we are subject.

Regaining Time = Regaining Learning and Development

Regaining time in and through the analyses that rigorously accounts for the timing
of social events will allow us making inroads to better theories of learning and de-
velopment. When classroom talk is flattened to conceptual content, which appar-
ently exists outside and independent of time, learning and development are lost as
phenomena. They have to be introduced afterwards. When we attend to time as a
constitutive feature of relations, however, learning and development are phenome-
na inherent to the very event. This is so because the analysis accounts for the in-
terweaving of micro-time, where we find the individual psychological experience
of temporality, and meso-time, where we find the communicated and intersubjec-
tive experience of time. In this way, the analytic approach presented here brings
together what have been called the different constitutional dimensions of time
(Tateo and Marsico 2013). In the analysis of Mario and Jeannie, we see the possi-
bilities of learning and development, which arise when in and through their ex-
changes, Jeannie comes to understand and comes to address what Mario does not
grasp; and in so doing, she experiences, and learns about, new ways of asking her
fourth-grade students questions such that these do rather than do not respond. Sim-
ultaneously, the micro-analysis reveals an unfolding understanding, accounted for
in the different appreciations that Mario articulated early in the exchange (Me, I
dont understand) and sometime after the exchange, after having completed the
 GETTING TIME BACK 201

task on his own (Me, I now understand). In this approach, time and temporality
no longer are external to the event but are constituted there from within in and
through the rhythmically changing ebb-and-flow and give-and-take of the ex-
change. The intersubjective time, pauses that belong to both of them and the pauses
they produce and use as resources, become resources such that attributions are
made to the otherbased on the expressive and documentary senseas being
slow or deliberate and thoughtful student, a sensitive teacher considerate of stu-
dents needs (Fig. 11.5). As the verbal exchange unfolds, in its characteristic of
time and temporality, a societal event is produced with all its recognizable middle-
class institutional relations where individuals take teacher and student positions. It
is also the place where language changes at the cultural-historical level, as postu-
lated in sociological approaches to language (e.g., Voloinov 1930) and in careful
studies relating instantaneous, ontogenetic, and cultural changes of language (Roth
2013f). Thus, it is out of such micro-genetic changes in very local contexts that
new words and grammatical forms emerge. More importantly, time is irreversible,
for the realization of possibilities produces a directionality that cannot be turned
around. The reversal of any two turns would have produced a different form of
joint action rather than the same action in the reverse.
PART D

Epilogue
We start with premises that are not arbitrary, not dogmas, but real premises, from
which we can abstract only in imagination. These are real individuals, their actions
and their material conditions, those that pre-exist and those that are produced by
their own actions. These premises therefore can be documented in a purely empiri-
cal manner. (Marx/Engels 1958, p. 20, emphasis added)
Philosophers only interpreted the world differently; the point is to change it. (p.
7)
12

Socially Responsible Data Analysis

In this book, I present and describe data analysis that aims at working out how
people make their social world an orderly phenomenon in quite ordered and order-
ly ways. People can do so because in whatever they do, they make visible every-
thing required for the continuous production of the situation as orderly phenome-
non. As Marx writes in the quotation that appears in the exergue to this last part of
the book, the premises of such analyses can be documented by purely empirical
means and, therefore, have the status of objective data. The words that appear in
verbal exchanges do constitute that can be represented in many situations without
too much contestation. There is therefore a material basis for the societal relations
that humans produce while doing whatever else they do (intend doing). This con-
tinuous work, as seen in part B, also is available to the analyst, who, without hav-
ing any background information and by merely studying the transcriptions, recov-
ers what kind of social situation the participants are in the course of producing.
There are many scholars writing about intersubjectivity as a problem. But this
problem is not apparent here at all, just as it is not in the everyday verbal exchang-
es we have with others. Intersubjectivity is a problem for the constructivist mind,
which is informationally closed, as some (constructivists) say, and engaged in a
near solipsistic endeavor of constructing some systematic ways of predicting what
is happening in the world. If this were to be the case, then the analyst in part B, or
any social analyst in what s/he analyzes, would be challenged to the limit, even to
the point of doing the impossible.
There is a second dimension as well: we never produce the social world on our
own, and what happens in the world is not (merely) the sum of, synthesis of, or
abstraction from all individual actions. Instead, the social world is a unit on its
own, which is captured in such descriptions as meetings and gatherings having
their own personality. Socially responsive data analysis attends to that dimension,
that is, to the ways in which the social world is shaped by and shapes the social
agents. Doing socially responsive analysis is equivalent to saying that we attend to
the social rather than to the individual as part of a project to reconstitute the social
by addition, synthesis, or abstraction. Doing such analysis means rigorously at-
tending to pragmatic aspects of the situations in which we are interested and at-
tending to the ways in which people actually perform rather than presumably think.
208 CHAPTER 12

In this way, socially responsive data analysis also is socially responsible analysis,
responding and accountable to society and its salient issues.
The classroom situations from which the chapters in part B are culled, though
absolutely real in their own right, do not represent what investigators normally
donot in the sense of having to reconstruct unknown social situations from a
very small set of indices. They are similar to the kinds of data analysis social scien-
tists conduct in the sense that some initially unknown phenomenon is taken to
stand in a wholepart relation to the concrete data at hand: the data constitute the
documentary evidence that both provides us access to the phenomenon, which in
turn is used to make sense of the data. Initially, this phenomenonwhereas pre-
sumed to exist in the fact that we engage in the study (in data interpretation, in the
search)is unknown. It is through the documentary method of interpretation that it
can be articulated and by means of abductive processes of reasoning.

Learning to Do Socially Responsive Data Analysis

The scientific habitus is a rule made man, an embodied rule or, better, a
scientific modus operandi that functions in a practical state according to the
norms of science without having these norms as its explicit principle: it is
this sort scientific feel for the game (sens du jeu) that causes us to do what
we do at the right moment without needing to thematize what had to be done
and still less the knowledge of the explicit rule that allows us to generate this
conformable practice. Thus the sociologist who seeks to transmit a scientific
habitus has more in common with a high-level sports coach than with a Pro-
fesseur at the Sorbonne. He or she says very little by way of first principles
and general precepts. Of course, she may set those forth as I did in Le mtier
de sociologue, but only if she knows that she cannot stop at that point: there
is nothing worse, in a sense, than epistemology when it becomes a topic for
society conversation and essays and a substitute for research. She proceeds
by way of practical suggestions, and in this she looks very much like a coach
who mimics a move (if I were you I would do this . . .) or by correcting
practices as they are executed, in the spirit of practice itself (I could not ask
this question, at least not in this form). (Bourdieu 1992, pp. 223224)
Learning to do rigorous, socially responsive data analysis as proposed here is not a
simple acquisition of rules. This is so because the process is similar to arriving at a
new theory, even though we are at some level engaged in producing such theories
on a daily basis. That we do so has been shown in the experiment described above,
where undergraduate students attempt to find out, by producing questions that
would seek to find the sense for the answers that preceded them, just precisely
what the counselor had in mind (Garfinkel 1967). This study allows us to bridge to
the pragmatic philosophical position on what is involved when participants in a
verbal exchange make sense of the ongoing talk (Davidson 1986). The language
philosopher approaches the problem as being one of interpretation, where recipi-
ents use passing theoriesi.e. theories adapted to the contingencies of the situa-
 SOCIALLY RESPONSIBLE ANALYSIS 209

tionfor hearing phrases in particular ways, adapted to other information that lets
them know how to hear it: literally, as joke, malapropism, stumble, momentary but
egregious stupidity, and so on. But such passing theories are not like other theories,
valid across time and space. Instead, they are always adapted to the always-
singular demands of any current situation. What kind of (linguistic) ability is re-
quired to communicate by means of speech? It is the ability that permits [an indi-
vidual] to construct a correct, that is, convergent, passing theory for speech trans-
actions with that person (p. 445). The author notes the circularity in this
statement, because it is equivalent to saying that communicative ability is both the
ability to speak, to make oneself understood, and the ability to hear, to understand.
This definition in fact erases the boundary between knowing language and know-
ing ones way around the world. The description is appropriate, as we can see es-
pecially in part B but also in part C, which exhibit the fact that analysts do indeed
work on the basis of knowing their way around the world, a form of knowledgea-
bility without which it would be impossible to perceive the ordered and orderly
ways of exhibiting the (social) order of the situation. Davidson notes that there are
no rules for arriving at passing theories, which means that there are no rules that
could be taught by stating and exemplifying them. Instead, there is the documen-
tary method.
The documentary sense goes together with the objective and expressive sense.
Whereas lay and professional social analysts alike use the documentary method,
the latter are likely more experienced and more methodical in working the rela-
tion between the different forms of sense involved. More importantly, perhaps, the
professional analyst would guard against importing unquestioned assumptions,
descriptions, concepts, and theories. The professional analyst brackets both every-
day and professional theories, concepts, descriptions, and assumptions rigorously
attending to the joint, ordered and orderly ordering work that underlies the ordered
appearance of the social world. This can be used to realize another sense of so-
cially responsible analysis, that is, for doing analyses that serve social critique the
intention of which is to exhibit societal-historical injustices that are part of the rul-
ing relations in which we participateboth as agents and as patientsproducing
and reproducing. Such responsibleand responsivesocial analysis, while not the
goal of ethnomethodology has been at the heart of critical feminist sociology that
takes (some of) its tools and theory from that field (e.g. Smith 1990a, 1990b).
Learning what is required by rigorous data analysis cannot be done by means of
abstract precepts about method. Mastery in rigorous data analysis, which is a social
practice exhibited as an event in part B, requires learning it alongside a kind of
guide or coach who provides assurance and reassurance, who sets an example and
who corrects you by putting forth, in situation, precepts applied directly to the par-
ticular case at hand (Bourdieu 1992, p. 221). We may in fact consider what the
instructor-analyst does in part B as a first step, a demonstrationsome call it
modelingof a practice so that students experience what it is in a concrete situa-
tion that they themselves contributed to setting up by choosing a relevant and per-
sonally interesting event and transcribing it. Some first analysis could then be the
starting point of a discussion, such as the one that followed the analysis of the Hei-
di transcription. Here, the classroom talk picks up on the hearing of a contrastive
conjunction but preceding erosion is natural (Transcription 1, turn 20).
210 CHAPTER 12

Fragment 12.1
Jean: But, is it that. It just doesnt say erosion is natural. He said, but, he
makes a contrast, that it is something that he is not mentioning, but
there is something
Instructor: [yea, but he hes [with the but, with the but
Jean: [more
Instructor: [so what is it?
Reza: [maybe is not is. Eh eh the emphasis is on the fact
[that everybody should know that. Maybe,
Instructor: [that there is a concern
Reza: [maybe] but everybody knows about the erosion of the cap rock.
Instructor: [yeah ]
Reza: Maybe this but means the height is very obvious. It depends on the
way that he he he structures the sentence. We cannot say just from
the [order.]
Instructor: [no no ] no, but
Stuart: Are you saying he is trying to emphasize the fact that its [natural.]
Reza: [Maybe.]
Lilian: Yes
Stuart: There is a possibility.
Instructor: But. So, what is the. What is the But. I mean, the question he didnt
even ask: what is the but set up to be a but to?
Leanna: Uh um.
Instructor: So, eh. It, is. Would it be. Is it, is it reasonable to hear it as a as a fol-
low up to and if thats fall off, the hump could erode way very
quickly.
Pei-Ling: And it is because Heidi said we are really concerned.
Student: Yes.
Pei-Ling: So they say but its natural [so dont have any concern in there.
Instructor: [Yea, yea. So so, we, if we hear it, then
we hear the but to the concern. That, were concerned. But its all
natural.
Leanna: So we shouldnt be as concerned.
Student: [Yes.]
Instructor: [Yea.]
Reza: Or he may go back to his first place question. Has Fred changed that
much whil- while ah. [Here,
Instructor: [whi- which one.
Reza: I mean, that the the previous part that that he just mentioned. Has
Fred changed that much while you are been here? So, hes looking
for the change.
Marines: An-han.
Reza: The the second person mentioned just erosion, and he he again talk
about that ok erosion is natural. What else?

In subsequent cases of student-produced transcriptions, all participants in the

seminar could then participate in a joint analysis of the transcription, which, there-
 SOCIALLY RESPONSIBLE ANALYSIS 211

fore, would constitute the basis for experiencing interaction analysis while doing
the analysis. Here, because it is an instructional situation, the more experienced
instructor-analyst may choose to follow Bourdieus (1992) advice and provide as-
surance and reassurance, set examples, and correct by putting forth, in the concrete
situation at hand, methodological precepts applied directly to the particular case at
hand. This would then be a form of pedagogy, where those learning to do analysis
would come to understand something they have (not) done in terms of the precepts
that the instructor may offer. That is, the students first participate in analysis and
then find in their own action the pertinence of this or that instruction.
We see such interactions between instructor on students occurring in the follow-
ing example over a transcription that yet another student has provided in which, as
the analysis would bring out, is an excerpt from some talk show in which the hosts
and guests speak about the infidelities of Tiger Woods. At one point, the seminar
participants, engaging in interaction analysis, are in the process of working out
whether the talk show excerpt has one or multiple topics, whether there is one con-
versation or whether there are parallel conversations / monologues. A graduate
student, Brian, then raises the question whether this would actually be called a
conversation or parallel monologues. The instructor then points to the need to
show, through the analysis, whether there is evidence for parallel talk or joint talk.
He then immediately and without transition moves into a corresponding reading:
who talks about what and when. He proposes using a table format for presenting
transcriptions as one of the strategies for the identification of underlying patterns
(Table 12.1). The table in which the transcription is placed into columns (Table
12.1) clearly exhibits what Brian expresses as his sense that there are two mono-
logues, with Joy and El each talking. The problematic nature of their talk is raised
in turn 08, when El offers up as a question Where are you coming from?, Joy
formulates being interrupted (let me finish the freaking sentence), and Joy even-
tually saying never mind.

Transcription 6
01 Joy: Ok, the thing about this guy . . .
02 Shelly: I dont think its right.
03 Joy: Let me say this about Tiger on his behalf. He has never held himself up
as one of these, like, pro marriage right wing, ur . . . kind a guys who is
anti gay and . . .
04 El: where is he goin
05 Joy: in other words, in other words the guy is not a hypocrite in his personal
life.
06 El: so only right wingers can be hypocrites? Is this the lesson we are learning
today? What are you talking about?
07 Joy: not in that sense, its like the Larry Craig syndrome, yknow where the
guy is tapping in the bathroom, meanwhile he votes against gay legisla-
tion.
08 El: where are you coming from?
09 Joy: wait a minute, wait a minute let me finish the freekin sentence.
10 El: well it was . . .
11 El: so what if he were . . .
12 Joy: never mind . . .
13 Woo: see, Id do you here what, youre here too
 212

Table 12.1 A different form of representation of a transcription may afford more easy detection of an underlying pattern.

Joy Shelley El Woo

1 Ok, the thing about this guy
2 I dont think its right
3 Let me say this about Tiger on his behalf. He has never held himself up as one
of these, like, pro marriage right wing, ur kind a guys who is anti gay and
4 where is he goin
5 in other words, in other words the guy is not a hypocrite in his personal life.
6 so only right wingers can be hypocrites?
Is this the lesson we are learning today?
What are you talking about?
7 not in that sense, its like the Larry Craig syndrome, yknow where the guy is
tapping in the bathroom, mean while he votes against gay legislation.
8 where are you coming from?
9 wait a minute, wait a minute let me finish the freekin sentence.
10 well it was
11 so what if he were
CHAPTER 12

12 never mind
13 see, Id do you here
what, youre here
too
14 (everyone is talking, . ((Inaudible))
15 la, la ,la ,la, la, la, la, la.
16 I care
17 Ok it is not Ok, it doesnt matter what
your politics are what you stand for,
duh, its not OK its not O.K. to do that.
18 Im talking about hypocrisy . Thats duh what
19 integrity..
20 how about I vow to you, I am married to
you, how about the hypocrisy in your
wedding vows? Now about that?
21 thats personal hypocrisy. Im talking about societal hypocrisy
 SOCIALLY RESPONSIBLE ANALYSIS 213

14 (everyone is talking, ((inaudible))

15 Joy: la, la ,la ,la, la, la, la, la.
16 Shelly: I care

Brian: Would you actually call this a conversation or would you call this
parallel monologues because nobodys listening?
Instructor: Um
Matt: Nobody listening?
Instructor: Well in part there, there may be multiple topics, this is the question.
I dont have a formal take on it yet, but what I need to do is I need to
analyze to see if they are talking parallel to each other or if they are
talking to each other, and for the purpose of each other, if they are
addressing one another. And language is only part of it. If you know
that in part, that they only talk to one another: at least El and Joy.
Because El talks about, here we have Joy talking about hypocrisy.
And El coming in and talking about hypocrisy, and then Joy down
here ((turn 20)) continues to talk about hypocrisy. So at least El and
Joy seem to have a conversation, whatever the others do and think,
and the question is, our analysis would have to work them in. But if
we had the five women lined up in a vertical system, then we would
see that El and Joy uh
Bonnie: They are not talking to each other, yea.
Matt: And others kind of talk, they are talking about something else and
theres a very interesting part there, turn fifteen. Because there was a
lot of, they were talking all together and kind of the temper kind of
came out and Joy would say LALALA and everything calmed down
and then Shelly would say, okay I care and uh L is but jump in and
say no its not okay, this doesnt matter what you BLALALA. And
the conversation keeps going.
Instructor: And so here ((after turn 16)) its back, whatever happens in be-
tween.
Laura: So Shelleys I care could have been to the whole group, as a way
to get them back on topic?

Here we have an example where rather than talking about precepts for qualita-
tive data analysis, the instructor and graduate students engage in the analysis to-
gether, providing opportunities to point out in the practical case, the role of rele-
vant methods, concepts, and theories. One of the suggestions concerns the tabular
format for presenting transcriptions, which the instructor would also demonstrate
in other parts of the course. Thus, for example, the transcription involving four
speakers in the fragment presented here would present itself as shown in Table
12.1. We also see that instead of following Brian in characterizing what is happen-
ing as parallel monologues engages in an analysis, which shows that El and Joy
both are talking about hypocrisy and that they at least appear to have a conversa-
tion. He then points out that the analysisperhaps implying a more careful analy-
sis than that can be done in the time slot that the seminar has availablehas to
work these [take-ups] in, that is, has to address this fact that there is take up of
214 CHAPTER 12

the words of the other, an aspect that plays a central role in a sociological approach
to the philosophy of language (Voloinov 1930). The instructor furthermore points
out the need to see, through his analysis, whether the participants are talking to
each other, and for the purpose of each other, if they are addressing one another.
He thereby exhibits an orientation to the nature of language as articulated in recent
philosophical works (Derrida 1996), which, like the pragmatic approach, takes into
account the fact that speaking is not a solipsistic affair seeking to establish inter-
subjectivity but rather that the language, which has come to the speaker from the
other (culture generally), always remains with the other, is intended for the other,
and, in speaking, returns to the other. This orientation to the other is itself a form
of social responsiveness, a responsibility for the (generalized) other, which is the
social. The fundamental unit of the social is not the socius, the individual, but the
societal (ruling) relation.
Some readers may want to argue that taking the social as the unit is a matter of
choice, a choice that differs from taking the individual as the unit. But this is not
so. We always find people in social relations with others, and individuals without
social relations would not act in ways that we characterize as typically human. The
individual human (hunter, fisherman) is an unimaginative fantasy of the 18th cen-
tury, just as is . Rousseaus contrat social [social contract], which binds by nature
independent subjects into relations (Marx/Engels 1983). The fantasy was expressed
in the novel Robinson Crusoe (published at the beginning of the 18th century),
who apparently, but only apparently, establishes a life all on his own when strand-
ed on an island. Epistemological Robinsonade refers to the attempt to explain
knowing and learning as if it were possible that children developed the competen-
cies required to navigate the (social) world all on their own (Davydov 1988). It is
an abstract assumption, detached from all social reality. What we observe in fact is
that in reality, a child cannot live or develop without practical and linguistic rela-
tions with adults (Leontjew 1964, p. 370). Observations conducted among deaf-
blind children show that when left mostly on their own, they do not innately devel-
op human competenciesthey are often compared to or described as vegetables
imprisoned in their physical sensations (Meshcheryakov 1979). Instead, they de-
velop characteristically human practices when their (initially physical) needs are
satisfied with human objects and tools, by means of human methods, and in human
relations. This, therefore, is a strong reason why we need to engage in relational
analysis in the manner that reflexive sociology recommends (Bourdieu 1992;
Smith 1999a).

Doing Socially Responsive Data Analysis

On the one hand, through their praxis humans therefore are the origin of the
active creation and conscious control of their conditions of existence, that is,
subject to and subjects of their societal life processes; on the other hand, they
are determined in their activity and consciousness by the objective living
conditionsbecause of the natural and societal necessities that underlie their
existence and thus material reproductionand therefore by the societal con-
 SOCIALLY RESPONSIBLE ANALYSIS 215

ditions, which they create and change through their collective objective la-
bor. (Holzkamp 1977, p. 324)
If there is to be a science which goes beyond systematic description and
analysis, it must lie in the dynamism of the historical development of these
social relations. The fact that the analysis of social relations in this sense has
not done much at the micro-level should not deter us. (Smith 1981, p. 316)
In the first introductory quotation, the author suggests that human beings not only
are subject and subjected to material and societal conditions but also are the sub-
jects creating these conditions. In the second introductory quotation, the author
articulates a position that (social) science must not stop short and be satisfied with
describing and analyzing social situation but in fact deal with the dynamism of the
social relations that are at the source of societal conditions. She favors, as the re-
mainder of the article shows, the use of ethnomethodology- and conversation-
analysis-based analyses for the purpose of transforming those ruling relations that
we are subject and subjected to. Such analysis is required for understanding how
the ruling relations play themselves out, out of hands and out of the control of any
particular individual, so that new practices may be considered to transform the
relations in ways that the general interests are met, that is, interests that are com-
mon to all members of society. I do not remember whether Smith somewhere
writes about transforming her own situation as a single mother understanding the
relations with the school as governed by the invisible force of the mothering dis-
course and the associated sociological discourse about the standard North Ameri-
can family. I re-read part of her Writing the Social (Smith 1999) but could not find
to what extent there were changes not only in her research objects and methods but
also in her personal life. I certainly would hope. It had done in the life of others
and in mine, once they and I had come to recognize ideology at work in our own
actions.
Socially responsive data analysis is not a practice that is the purview of profes-
sional researchers. In chapter 9 I refer to the case of a father and son relation, ini-
tially blaming each other for their conditions, which, in for the son, meant being
expelled from different schools. Once they engage in a critical analysis of their
situation, father and son not only come to understand how their practices are
shaped by societal conditions but also transform these conditions such that the son
re-entered and completed school. That is, socially responsive data analysis may de
facto be and become a means for lay folk to organize and transform their everyday
collective workand this works even when the participants still are school stu-
dents (e.g. Stith and Roth 2008). An example of this derives from a study in urban
Philadelphia, where a science teacher (Jen Beers), with the aid of a doctoral student
(Sarah-Kate LaVan), introduced critical analysis with videotapes as data into her
science classroom.
On this day Jen had chosen an incident that occurred while she had been absent
from the school the day before. Her students had disrespected the substitute
teacher who was doing her internship at that time. The students were resisting to
being in the classroom with the replacement teacher to the point that one particu-
larly reticent student was removed from the class. The planned work for the day
was not completed. In that context, Sarah-Kate also showed some of the video
clips students had selected showing them displaying resistance. Jen provided some
216 CHAPTER 12

concrete examples of resistance, referring to the relations between students and the
replacement teacher. The students told what had happened from their perspective
and one student (Ace) in particular, who was also part of the research team, spoke
up. Ace noted that some students and he diss-ed the replacement teacher by act-
ing out. In presenting his position, Ace initially provided a detailed description of
the situation. He provided evidence supporting his claims. The discussion moved
from a description and analysis of their practices to articulate new, future practices
that would prevent students, including Ace, from being removed from the class-
room and others from being seen as acting out. LaVan and Beers (2005) provide
the following fragment from the session in which the high school students first
describe acting differently in different settings and then use a sociological concept
code switching to articulate how Ace may acting in conformance with the class-
room rules without having to feel as selling out.

Fragment 12.2
Terrell: You do it again.
Ace: ((Loud voice)) =I guess I aint gonna be nuttin in life if I cant do
that=
Terrell: ((Loud voice, body positioned toward Ace)) But you do it and you
dont know you do it. You dont act the same
Shania: [But dont have a negative attitude about it like, look at it as some-
thing positive like
Terrell: [You dont act the same. No go ahead ((nodding to Shania)).
Shania: Like when somebody realizes that they bad then they change their
behavior.
Ace: I aint gonna be nuthin in life then.
Terrell: Do you go to church?
Ace: No.
Terrell: You dont go to church?
Ace: Yeah, I go to church sometimes.
Terrell: Well, you dont act that way when you go to church do you? You
code switch. You dont realize you do it, you just do.
Ace: Yeah but this is different. ((Voice gets louder and more charged
with emotion)) Im not gonna give up who I am for her. I wont sell
out.
Terrell: You dont have to (0.5) you just gotta code switch to get by.
Jen: (1.0) Think of it as playin the game.
(LaVan and Beers 2005, p. 197)

The students not only identify and analyze behavior shown in the video clip,
but, as seen here in the opening turn, they identify particular forms of relations
when these occur during the discussions. Here, the issue analyzed was Aces be-
havior on the day before. Terrell points out that Ace is acting in ways that are criti-
cally analyzed and negatively valued as part of the meeting. Terrell points out that
Ace not only is acting in a particular way but also that he does not even know that
Ace is doing so. As the exchange unfolds, Terrell uses code switch to describe
 SOCIALLY RESPONSIBLE ANALYSIS 217

what Ace has to do, and he provides an example of code switching in terms of the
different practices that Ace enacts when he enters church.
Another example of critical data analysis conducted in a reflexive, rigorous
manner occurred early in my own work as a high school teacher (Roth 1993). At
the time, there was a lot of research on how teachers used metaphors to orient
themselves in their classrooms and how the choice of new metaphors may involve
major changes in the classroom life. I was interested in investigating what my own
metaphorcognitive apprenticeshipwas covering up about the social relations
between students and myself. I engaged in two forms of analysis of videotaped
classroom events: one based on the cognitive apprenticeship metaphor and the oth-
er one employing what I had learned from reading about conversation analysis. As
a result of the study, I learned that the use of metaphors for understanding teaching
practice introduces some real dangers. Whereas the use of metaphors may have
been beneficial in facilitating changes in teacher classroom actions, I came to un-
derstand that there was a real danger: teachers may merely appropriate a discourse
without changing their praxis once the metaphor had been legitimated as a desira-
ble norm. If, so I reasoned, we classroom teachers are interested in changing prac-
tice, we can find help in new conceptual tools that reject the use of a priori descrip-
tions and frameworks, and seek understanding that emerges from the phenomena
of interest. In this search for new understandings, even our descriptions should not
be sacred and constrained by preconceptions of what descriptions ought to look
like, but should arise from the need of a best possible understanding of a particular
phenomenon. Because of my interest in the relation between teacher and students,
I chose to take a close look at our conversations at the level of individual utteranc-
es in order to construct an understanding how the participants managed these. Such
an understanding was exactly what my previous analyses could not provide.
My second type of analysis highlighted several features of the discourse strate-
gies employed when I interacted with the students: there were interruptions, expan-
sion and reflection questions, questions that problematized concepts, and repair
sequences in a triadic interaction pattern. It may be possible to integrate these in-
teractional patterns into the cognitive apprenticeship framework, where they would
be understood as providing fine structure to my scaffolding moves. However, there
are some potential problems. In the apprenticeship metaphor, scaffolding is slowly
faded out until students are independent. I suggested that for some, this process
might imply an almost linear transition from teacher to student control of the prob-
lem. Yet the analysis I offered provided evidence that contradicted such a linear or
nearly linear transformation in teacher-student relations: my conversational moves
were contingent on the students discursive patterns. Thus, a group of students
working largely independent and producing elaborated answers to occasional
teacher questions that problematized some issue at one instance, was engaged in an
IRE-type repair sequence in the immediately following situation. A further prob-
lematic arises from the fact that some conversational moves can be understood
both as a scaffolding device and as an instance of modeling scientific inquiry.
Thus, the concepts of modeling, scaffolding, and fading are like tools that are not
sharp enough to do the job they are supposed to.
218 CHAPTER 12

To exemplify, consider the following materials that led to the published article
(Roth 1993). Concerning the way in which I organized classroom life and in which
I saw and understood what was happening, I wrote the following research note:

Throughout the students inquiry, I conceptualized my role as that of an advisor

and resource person. The role of an expert who scaffolds student performance
was most apparent during the interpretation of data and construction of
knowledge claims. Here, I had provided the scaffolding support students needed
to coordinate isolated items of their prior knowledge and construct new, more
integrated frameworks (to students I had become, in students words, the phys-
ics coach). This construction occurred first in the collaborative effort between
students and teacher, from where each individual could appropriate, that is indi-
vidually construct, his own representation. I monitored students emergent
meanings throughout each lesson.1
Interacting with the students throughout the focus finding sessions, I suggest-
ed alternative research questions, coached students as they evaluated their ideas
in terms of instruments and materials, encouraged students to frame new exper-
iments in terms of the findings of previous ones, and encouraged students to fo-
cus on details of their plans. These interactions can be understood as instantia-
tions of the scaffolding metaphor.2 However, I also emphasized that it was the
students responsibility and privilege to make decisions with regard to both the
focus question and the plan for the experiment (If you think that this is worth-
while investigating, I would like for you to look into that or I want to leave it
to you to decide what question you will investigate). Such a shift in responsi-
bility to the students corresponds to the process of fading, and thus overlaps
with scaffolding. During the scaffolding and fading phases, I served as a re-
source in questions of equipment and materials. Unavailable materials or in-
struments sometimes precluded an experiment, although students had framed a

1
The circularity is quite apparent in this first paragraph. I was both organizing
what I was doing in the classroom and what I was providing in terms of materials
and support in terms of the stated metaphors and looking at and analyzing vide-
otaped classroom events through the lens of these metaphors. This is not singular
to my own, teacher-based efforts of doing research. Instead, we find this always to
be the case in experimental research where the researchers design the intervention
and study it through the same lens. Little surprise, then, that I found great con-
sistency without, initially, understanding the function of ideology. But this is pre-
cisely why ideology is so dangerous.
2
Here again, I had picked up the concept of scaffolding from the research liter-
ature current at the time and was thinking about my teaching in terms of this met-
aphor. I found confirmatory evidence when I subsequently looked at the investiga-
tions that there was in fact scaffolding going on. Others working with me at the
time and using the same metaphor also confirmed, including those teacher-
colleagues who were doing the research with me in the school at the time (e.g.
Roth and Bowen 1995) and those outside colleagues who participated in publish-
ing our research (e.g. Roychoudhury and Roth 1996).
 SOCIALLY RESPONSIBLE ANALYSIS 219

suitable question for a high school laboratory. In such cases, I helped students
shift their focus and do a related experiment.
(November 1992)

Following this note there was an instruction to myself: Analyze the following
two episodes (a) through open coding and (b) by using the metaphor of apprentice-
ship. I subsequently provided some fragments from the data that were consistent
with the ways in which I described and explained classroom life. However, there is
a circular way in which the ideology operates. I strived organizing classroom life
according to the cognitive apprenticeship metaphor, and I saw it occurringat
least initiallywhen I looked at what was happening. I began to have doubts, in
part because of the very ways in which metaphors functioned: organizing our part
in classroom relations without having to think about the relations in explicit ways.
In part, my doubts about the usefulness of metaphor as an analytical tool were
fueled by the critical, fine-grained analyses of teachers management of classroom
discourse that I had read about (Lemke 1990). On this basis, I began exploring new
ways of analyzing data that would allow me to bracket macro-level descriptions
and engage in rigorous (microlevel) analyses of studentteacher (ruling) relations.
The following derives from the materials I was subsequently using as part of my
publication. I re-presented the transcriptions that I had used to depict the metaphor
analysis, now enhanced by transactional details that are typical for conversation
analytic studies. I do not present these materials here as an ideal case or form of
rigorous analysis but rather as the beginning effort of a practitioner to engage in
critical analysis by rigorously attending to the relational details of his everyday life
situation: a physics classroom.

1.1 Rob: You heat up a gas and insert a test tube and see how many bubbles and=
1.2 T: =But this time you want something quantitatively.
1.3 Atif:How could you do the thermal expansion of a liquid?
1.4 Rob: How do you do it with a gas?
1.5 (1.6)
1.6 T: How could you do it?
1.7 (.)
1.8 Atif: You take a flask with a stopper and a tube going to another bottle full of
water upside down. When you heat the gas is going to go through the
tube and in that other flask. And then the water comes out=
1.9 T: =But quantitatively (.)
1.10 And could you do it with different types of gases?
1.11 Atif: A tube filled with a bubble, and as the gas expands, the bubble moves
along the glass tube.
1.12 T: And very similar with liquids? could you?
1.13 Rob: Yeah, just as the water expands it goes up.

The episode began with Robs suggestion to conduct an experiment on the

thermal expansion of gases (1.1).3 As the latch-on (identified by the sign =)

3
In these analyses, the unit of analysis still is the individual so that each turn is
interpreted as if it existed on its own. The result is an analysis that does look at
interactions, where the participants are thought of as coming into the relation with
220 CHAPTER 12

indicated, Rob was interrupted by my comment that they were to measure ther-
mal expansion quantitatively rather than qualitatively (1.2). Atif apparently pur-
sued a different idea and, on the following turn, asked how the thermal expan-
sion of a liquid could be measured (1.3). Rather than answering Atifs question,
Rob wondered how the thermal expansion of a gas could be measured quantita-
tively (1.4). Although he looked at me to indicate the direction of his question, I
remained silent for a conversationally long pause before reflecting the question
back to the students (1.51.6). This pause provided students with an opportunity
to consider their own question, an argument supported by the research on the
importance of wait time to students elaboration of their own ideas (Tobin
1987).4 However in the present case, the question came from one student, tenta-
tively directed towards the teacher. Not answering in a conversationally appro-
priate time, this pause may have also indicated to students that the teacher want-
ed to signal something about the question.
Following the pause, I turned the question back to the students by repeating it
(1.6). Whereas Robs question seemed to ask for the method of conducting such
an experiment, the stress on could in the reflected question suggested that there
are several possibilities. From the point of a conversational analyst, this is a
non-sequitur, a break in the ordinary conversational rule that a question be fol-
lowed by an answer.5 As a result of this break in convention, it became the stu-
dents task to answer their own question. By that time, however, Atif was ready
to answer Robs question. He elaborated on an experiment that would allow him
to show the thermal expansion of a gas (1.8). Rob also enacted his explanations

intact intentionalities and subjectivities. The focus of the analysis is agency at the
expense of the radical passivity that is actually observed and experienced when we
participate in societal events, such as lessons, meetings, shopping, lectures, and
the likes.
4
A relational approach, which attends to the social sphere in more rigorous ways,
recognizes the relational nature of pauses, which provide opportunities for taking
the floor to both teachers and students when the research context are classrooms,
and any participant in other societal activities. Those participating in verbal ex-
changes both produce and are subject/ed to pauses, which are transactional fea-
tures of societal relations.
5
My analytic reasoning here seems to put the cart before the horse: a phrase is not
a question but becomes a question at the instant that the response becomes a reply;
the same phrase is an insult when the response names or treats it in that manner.
We may hear this comment in terms of H. Sackss maxims, whereby something can
be heard in a particular way it should be heard in this way. Relational thinking
pertains not only to the social relation as a whole but also to the parts, for exam-
ple, the turn sequences (chapter 7) that assign question and reply status to two
sequentially ordered turn phrases at the same time that their whole as {question |
reply} comes to be. At that time, I had not yet understood the idea of dialectical,
mutually constitutive relations, dialectical reasoning, or relational thinking. Per-
haps as a document of the times and the cultural context (Anglo-Saxon scholar-
ship), my social constructivism was simply constructivism now placed in a social
context.
 SOCIALLY RESPONSIBLE ANALYSIS 221

with gestures that outlined how the glass tube was to be connected from the
flask to the inverted bottle so that the gas escaping from the flask could be cap-
tured (and measured) by displacement of water in the inverted bottle. Interest-
ingly enough, Atif took the initiative in designing an experiment, although he
indicated in an interview that he did not like to do that. He preferred to be given
instructions regarding the equipment to be used, the procedures to be completed,
and the questions to be answered.6
By reiterating his request for quantitative experiments, I interrupted Atifs
explanation (1.9). In this, the interruptions in (1.2) and (1.9) were conversation-
ally and semantically equivalent. Both called on the student to think about quan-
titative experiments rather than qualitative ones by interrupting the previous
speaker, and by taking the turn away from him.7 The emphasis on quantitatively
highlighted my intention to move students to increasingly mathematize their la-
boratory experiences by using mathematical representations for their observa-
tions.8 However, both statements can also be read as value judgments about the
content of students utterances. Any one of three aspects(a) the use of but, (b)
the interruption before the end of the terms, and (c) the stress on quantitatively
(1.9)could indicate disagreement with the proposed experimental design. The
interruption even signaled that this disagreement was so important that it war-
ranted an interruption, which constituted a break in conversational conventions.
As such, it might also signal or establish differences between the interlocutors.
Some participants are assumed to have the right to interrupt others discursive
contributions, which points to an asymmetrical relationship between partici-

6
Here, we see how practical action and accounts thereof provided a priori or a
posteriori may fall apart. What we say we do, for example, in an interview, and
what we actually may have little to do with each other. This is not so devious and
not so particular to everyday folk practice. Instead, research has shown that even
experienced scientists, after having done a kind of dissection for nearly 30 years,
may have intend to do one in the morning only to find, after having collected data
for five hours, that they have not done the dissection as they had planned doing it
(Roth 2009).
7
The interruption may not have been intended as interruption. The intent, if there
had been one in the conscious awareness of the teacher, may well have been the
desire to assist students in getting on with their task. But the intent does not matter
in our attempt to understand what is being done here. The effect of beginning to
speak at that time was a possible interruptionpossible because we do not know
whether the student was going to speak longer than he turned out to do. There is
no struggle observable, which should not be taken as evidence that there is no
struggle. What the participants do bring about together is the differential access to
the speaking floor and, therefore, the institutional differences between designated
teacher and designated students. As emphasized throughout this book, rigorous
data analysis is a relational analysis, attending to fields, forces, and irreducibly
joint social work and ruling relations, instead on attending to things.
8
Here we observe a problematic aspect of the analysis: the analyst introduces his
own intentions when these have no place in relational analysis because what mat-
ters is the effect of social action.
222 CHAPTER 12

pants.9 Such an asymmetry underlies the traditional master-apprentice relation-

ship. Here, rather than letting apprentices engage in activities which lead to
blind alleys and failure, masters and their equivalents in apprenticeship-like sit-
uations interact with the apprentice in an immediate mode.
Following my call for quantitative measures, I asked a question. It was de-
signed to encourage students to construct patterns not only for the thermal ex-
pansion of a particular gas, but for patterns across a number of gases (1.10).
The stress on different emphasizes a desirable variation in the choice of sub-
stancesfrom a scientists perspectiveso that the thermal expansion of one
substance could be evaluated in the context of the behavior of other gases.10 Atif
responded to my first request by elaborating on his basic design in (1.8) so that
he could now obtain a quantitative measure for the thermal expansion (1.11).
By itself, utterance (1.11) did not make much sense. But in the context of the
earlier design communicated through utterance (1.8) and the accompanying ges-
tures, the experimental design Atif envisioned took shape. He further helped lis-
teners construct an understanding of his design by indicating the moving liquid
bubble in the imaginary glass tube connecting the flask and the inverted glass
bottle submerged in water. My question in (1.12) may have encouraged students
to think whether their experimental design was transferable to the measurement
of the thermal expansion of liquids.11 As in (1.10), I stressed one concept, here
liquids. This stress on liquids contrasted the current conversational focus on
gases. In this, (1.10) and (1.12) were very similar. Both of my utterances raised
the question of expanding the experiment to different substances in the same
phase or across phases. In a sense, these questions provided new opportunities
in the students search for lateral extensions of their research. These questions
also modeled scientific approaches to expanding research programs into new ar-
eas. Rob immediately responded by sketching an experiment in which water ex-
pands into an upright glass tube that emerges from the stopper in the flask
(1.13).
The line-by-line analysis of the present excerpt revealed three types of con-
versational action including (a) interruptions, intended to encourage students to

9
We can see this analysis in terms of having the right to in terms of a realization
that social actors projectively orient towards behavior such that what they will
have done more or less conforms to some stated or statable rule. But this projec-
tive orientation should not be taken as the cause of the practical action that follows
(e.g. Suchman 2007). Instead this projective orientation, like a plan, becomes re-
source for a posteriori accounts of what has happened, whether a plan (intended
action) has been realized or whether a plan (intended action) has failed.
10
Whether something was taken as emphasized has to be shown through a rigorous
data analysis, which focuses on how other participants have taken up a phrase or
part thereof.
11
Here there is an explicit statement that allows us to identify the speculative na-
ture of the analysisconsistent with the constructivist spirit of the timeswhere
possible mental contents and effects are part of the analysis rather than sticking
with what the participants make available to each other in the social relation that
is constituted by means of this action that makes available.
 SOCIALLY RESPONSIBLE ANALYSIS 223

immediately repair a developing idea (1.2, 1.9); (b) expansion questions, model-
ing scientific problem searching or providing new leads for students to consider
expansions of previous experiments into new areas (1.10, 1.12); and (c) reflect-
ed questions, which put the next conversational turn and the responsibility for
finding an answer to their own question back to the students (1.6). The effect of
the interruptions was to constrain or redirect students thinking in specific
ways.12 However, both interruptions could also be read as value judgments. By
contrasting the students utterance with but, I signaled problems with the stu-
dents design. The immediately following quantitatively assumed that the stu-
dent recognized its contrast with the absent qualitatively. That this interjection
was not clear became apparent in (1.8) when Atif presented another qualitative
experiment to study thermal expansion. However, in spite of the fact that I did
not contrast quantitative by uttering qualitative, Atif seemed to understand the
significance of the utterance (1.9), and continued by designing a quantitative
experiment.
The effect of expansion and reflected questions, on the other hand, opened up
possibilities for students to engage in further reflection and inquiry. However,
while the reflected question left further inquiry entirely to the students, the ex-
pansion questions provided direction for such an inquiry. Could you do it with
different types of gases? and very similar with liquids? both encouraged spe-
cific expansions of the original focus question.13 At the same time, these expan-
sion questions can be viewed as a part of my modeling of scientific inquiry. In
the present context, and in the context of the students inquiry, both of these
meanings are possible. Such ambiguities are constitutive of many learning situa-
tions (Jordan 1989). Jordan cited the example of a potter masters rejection of a
piece of clay brought by the apprentice. This rejection could have two equally
reasonable interpretations. On the one hand, it might be a lesson to the appren-
tice how clay feels when it will not do the job at hand. On the other hand, it
might have simply been an attempt to save herself the effort of searching for the
clay.

Although I recognize the shortcomings of the data analysis I had conducted and
the article I had written at the time, it did take me to an important realization: the
need to bracket everyday and scientific categories and, therefore, undermine their
ideological functioning. This ideological functioning is not immediately evident in
situation but rigorous data analysis can show us the contribution that these catego-
ries make to work that produces the societal relations in and on the basis of which
the social order is built.

12
There is a realization, or rather, the result of a relational analysis, that whatever
the intention might have been for talking when I was talking at that time, the effect
was an interruption, which itself constrained or gave a new direction to whatever
students were talking about.
13
The statement that this encouraged something to happen is overly optimistic and
goes in the direction of a causal explanation. What happens here is that this was
observed as having happened, but whether this was the effect of the preceding ac-
tions cannot be ascertained.
224 CHAPTER 12

Coda: Real People and Real Activities

The question whether there is objective reality in/to human thinking is not a
question of theory but a practical question. In praxis, man has to prove the
truth, that is, the reality and power, this-sidedness of his thinking.
(Marx/Engels 1958, p. 5)
All societal life is essentially practical. All mysteries, which drive theory
to mysticism, have their rational solution in human praxis and in the compre-
hension of this praxis. (p. 7)
In the German Ideology, Marx and Engels propose grounding social science in the
praxis of human relations. These relations are not based on mystical and mythical,
metaphysical foundations but on the reality of human (material) praxis. Theirs is a
critique of a method of reasoning about society and history that treats concepts as
if they were agents (Smith 2005, p. 54). This, as the feminist sociologist argues, is
the crux and problem of present-day sociology, which we may extend to most oth-
er social sciences as well. The institutional ethnography that she recommends,
which has intentions similar to those that are outlined in the present book, does not
take up the project of Marx, which belongs to political economy. Instead, it has
learned, as I have, that the concepts we use have themselves histories in the con-
crete actuality of the doings of lay and professionals alike. Rigorous data analysis
attends to the praxis and practicalities of social relations, their constitutions and
their products (effects), including the differentiation into institutional positions and
the privileges that come to be associated with them (as orientational or explanatory
resources). Here, any abstract properties of social relationsknowledge, power,
positionare themselves product of the praxis of social relations. That is, the so-
cial relations are not things but concrete, ever-changing forms of human praxis.
Rigorous data analysis attends to the fine details of the continuous stream of
societal life that we produce as much as being subject to. It is by carefully attend-
ing to the details of the relations that human beings entertainby means of words,
gestures, body orientation, gaze direction and the likesthat we can recover a
sense for what it has meant to be a participant in that situation. It is precisely such
attention to the details of relations, by focusing on talk and on what talk leaves
unsaid, that the analyst in the second part of this book was able to recover the na-
ture of the situation in which the unknown transcripts were produced. This book is
something like an invitation to educators and other social analysts to attend more
closely to what people actually do and undergo in the recognizable re/production
of societal life rather than using mentalist concepts to explain, in nearly causal
ways, social events after they have come to a conclusion. Analyzing how real peo-
ple experience real activities needs to retain the openness that we experience with
respect to the outcome of situations, where we never quite know what will happen
next and where we will be situated only a short periods hence.
Appendix A: Transcriptions for Part B

The transcriptions analyzed in part B of this book are provided in their entirety as
provided to the instructor analyst with the sole exception that consist turn number-
ing has been added. Punctuation and spelling is identical to what the graduate stu-
dents had brought to the data sessions.

Transcript 1 (Heidi, David Suzuki)

01 Heidi: Its amazing how the landforms here really look like other things
other than rocks. Do you notice that sometimes?
02 Amanda: Yeah
03 Ashley: Yeah
04 Michael: Yeah
05 Heidi: Its funny
06 David: As long as youve got a good imagination
07 Heidi: Well, weve noticed that maybe, yeah, maybe lets go down this
way. We noticed that the longer youre out in the badlands, and the
hotter it is, the more things look like things
08 David: Yeah-ha-ha-ha. Your imagination gets looser, huh?
09 Heidi: Well, it does. Some people claim it might be a kind of hallucina-
tionnow watch out for the cactus. When we get around here, you
want to take a look around and see if theres any landforms that look
like something that would be familiar to you, not just like a rock. So
what do you think that landform over there is? Does this one look,
look like anything to you?
10 Amanda: Hmm. Oh! That rock right there looks like a camel
11 Heidi: Oh . . . no . . . thats it! We actually have a name for this guy. We
call him Fred the camel . . . see the hump . . . see the big droopy lips
226 APPENDIX A

pointing to the left. And if you look off in the back can you see any-
thing else?
12 Amanda: ?
13 Michael: No.
14 David: Ill give you a clue. Where are camels found?
15 Ashley: Egypt.
16 David: Very good!
17 Amanda: A pyramid. A camel and then a pyramid
18 David: Very good! Has Fred changed that much while youve been here
Heidi?
19 Heidi: Not that much although he did get a bit of a facelift, hes lost his
double-chin. But, uh, were really concerned that the cap rock on the
hump of Fred may fall off. That ironstone. And if that falls off the
hump could erode away very quickly.
20 David: But erosion is natural, its going to fade away over time.
21 Heidi: Fred is naturally going to erode away. But if he ever lost his hump,
all wed do is change his name to Humphrey the camel.
22 David: Awwwwwwooooooo
23 Heidi: Ha ha Ha Ha HA Ha
24 Amanda: [Ha ha
25 Michael: [Ha ha
26 Ashley: [Huh ha
27 Heidi: Sometimes it takes a while. Hmmm Hmmm Hmmm

Transcript 2 (Vicky)

01 Vicky: So you may be on red you may be on amber, like some of those
questions you can answer . . . you may be on green. Lets just look at
first impression. Would everybody just show me . . .
02 Vicky: If you show green traffic light are you on green means yes I under-
stand it all, I can do the question, Im confident, Im happy. Um an
amber traffic light means that you . . . have grasped it a little . . . you
you think . . . you think you can sort of do it . . . but youre not really
shhh, youre not absolutely sure and a red traffic light means Ive
got no idea, I dont get it.
03 Vicky: Weve got a whole range of answers here . . . look at this see part
one . . . k . . . Right . . . 25.73 to correct to one decimal placccce . . .
you look at the next number and you say; is it five or bigger? Does
anyone know the relevance of that question? Hands down.
04 Newscaster: Vickys using traffic lights together with the controversial no
hands up, policy.
05 Vicky: (whispers) hands down, hands down
06 Vicky: Theres some very enthusiastic students there who were dying to
speak out and . . . they find it boring if they cant . . . um so they . . .
I suspect they do get frustrated (break to class) . . . as a teacher you
 APPENDIX A 227

sometimes resort to picking them because you want the lesson to

move forward . . . and . . . it often can be that five or six students
will have a dialogue with the . . . teacher all the way through the les-
son and that the other 25 will sit . . . dormant. If you have no hands
policy . . . then it . . . changes everything because they suddenly real-
ize that they might have to answer a question whereas before, they
knew they could go to maths and just sit there and do nothing. They
dont necessarily like it . . . but at that point when youve asked a
question youve got every single student engaged and thinking
whats the answer, I might have to answer it.

Transcript 3 (Bullrush)

01 I:And the girls are putting their hand up too.

02 P:Were tough out west. [girl tackling boy in background]
03 I:And those girls, more than hold their own.
04 I:Do you ever worry about getting hurt? [kids playing in background]
05 M1:
No
06 I:Have you ever been hurt?
07 M2:
Guys
08 M1:
Ya
09 M2:
Yeh you guys, were supposed to be playing bullrush (yelled from be-
hind)
10 M1: Ya, actually, really . . . ya, I have really been hurt.
11 I: And what happened then?
12 M1: I think I got up and dealt with it.
13 P: I think were wrapping kids up in cotton wool a little too much . . . um .
. . years ago a good parent was somebody who just let kids play. Nowa-
days a good parent might be considered to be somebody [who] takes
them to dancing lessons . . . and rugby practice.
14 I: there may be some people, some parents who think, hmm, I dont want
my children, my little girls, playing bullrush! What do you say to those
people?
15 M1: Well, if you think you cant handle it, well then, dont play . . . thats
just pretty much it.
16 I: So what do the parents think?
17 P: Some parents have come and asked about it. Theyve wanted to be reas-
sured . . . but I think generally, Ive had really good support from them.
[kids yelling and waving there arms in front of the camera]. I think our
understanding of what is safe really means, is changing. And actually,
kids are safe doing things, that, maybe we have thought, werent safe . .
. for quite a few years.
18 I: For these educationalists, the risks involved with a bit of rough and
tumble, are far less than the risks associated with an activity.
228 APPENDIX A

19 P: If you get a kid to test themselves when hes 7 years old on a scooter or
. . . a tree . . . climbing a tree. He is not going to have to test himself
when he is 17 . . . behind the wheel of a car.
20 I: And they also say it makes for better students.
21 P: The only time they get into trouble is when theyre bored . . . and they
really dont get a chance to be . . . [laughter]
22 I: And yes, before you ask, the kids do go back in the class after playing
bullrush, with a bit of mud . . . but the full on mud sliding, well, thats
before they head home . . . to the washing machine I presume.

Transcript 4 (Mikela)

01 P: Come on inside. Does she look like a little girl?

[looking at little girl, left hand pointing at little girl, brings hand towards
herself, looks down at page]
02 P: Heres Mikelas story
LG: [gets up and enters center of carpet stands and swings arms]
03 P: A little girl.
LG: [holds fingers]
04 P: Does she look like a little girl?
05 C: Yeah
LG: [turns head towards talking child]
06 P: Okay. Would you be a little boy?
[brings left hand down towards lg2 and looks down at lg2]
LG2: [looks up at P and nods]
07 P: Okay . . . A little boy
[gestures hand towards center of carpet]
LG2: [gets up] [stands next to LG, hands by side]
08 P: Would you be the little house?
[looks down at lg3, left hand pointing towards lg3, nods]
LG: [walks towards LG2 and stands beside LG2]
LG3: [stands up, body swinging, hand to mouth]
LG2: [grabs dress, looks at lg3, walks away from lg3 behind lg to stand beside
lg]
09 P: Okay. A little house. How should Mikela since its your story how
should the house be how should she make herself?
10 LG: [looks down, then stretches both arms out to the side, then spreads legs
apart]
LG2: [hops once]
11 P: Alright would you do that? Like that.
LG3: [swings side to side puts arms partially out to the side]
12 P: And do you go inside the house?
[points at lg]
LG: [nods]
 APPENDIX A 229

Transcript 5 (Kiana)

01 M: Kiana, what did you do today?

02 K: Er . . . cooking. [She avoids the camera and looks at the other side
03 M: Cooking! What elsa did you do?
04 K: [She turns back to the camera]
05 Er [She avoids the camera again and spends about 3 seconds to say it]
06 [She turns back to the camera again]
07 [She clenches her fist and waves her right hand]
08 M: Whats that?
09 K: Crayoning.
10 M: Crayoning. So drawing.
11 K: Yeah! [She nods and smiles]
12 M: Yeah!
13 [She turns her head away]
14 And . . . [She turns back to the camera]
15 [She turns her head away touching her mouth]
16 [She continues touching her mouth and starts touching her tongue]
17 ???? [Japanese? She starts talking again and turns back to the camera
right after that]
18 M: Thats all?
19 K: [Kiana nods and stares at the camera]
20 M: I saw maybe you went into pool today, did you?
21 K: [She nods and leans forward]
22 M: Yeah? Is it a big pool or small pool?
23 K: [She picks her nose with the left hand and she uses a pinch gesture with
her right thumb and index finger]
24 M: Whats that?
25 K: Small pool. [She still uses the pinch gesture and smiles]
26 M: A small pool! Did you have fun?
27 K: Yeah! Yeah! Yeah! Yeah! Yeah! Yeah! Yeah! Yeah! Yeah! [She starts
waving her hands in the air]
28 M: And what did you eat for lunch today?
29 K: Er . . . [She spent about 3 seconds on it]
30 [She turns her head away]
31 Lots. [She turns her head back but avoids looking at the camera]
32 M: Hmm . . . lots.
33 K: [She looks at the camera again]
34 M: Like what?
35 K: Er . . . (mom?) . . . Kianas home. [She gazes away from the camera]
36 [She looks at the camera]
37 M: You ate lots of food at Kianas home?
38 K: Yeah! [She keeps staring at the camera]
39 M: What did you eat in kinder today?
40 K: Lots. [She still stares at the camera]
41 M: Lots . . . And what are you gonna do now?
230 APPENDIX A

42 K: Er . . . I dont know. [She looks towards the other direction for a little
while and turned back]
43 M: Are you tired?
44 K: No. [She shakes her head and smiles]
45 M: I think so. What do you do after you have a shower and bath?
46 K: [She looks towards the other direction]
47 M: Whats next?
48 K: [She looks back to the camera and lifts her fist to her mouth and then
opens her mouth]
49 M: Whats that?
50 K: Brush your teeth. [She still looks at the camera]
51 M: Yes. You gonna brush your teeth. And what we gonna do before bed
tonight?
52 K: Er . . . lots play. [She looks down and does not look at the camera until
she says play]
53 M: What? No! I dont think so. How about we. . .?
54 K: Hmm. [She shows a shy smile and looks down]
55 Little bit play. [She looks at the camera and bites her bottom lip]
56 M: Hmm . . . little bit play and maybe read some books?
57 K: Yeah! [She smiles and raises her left hand]
58 M: Sing a song?
59 K: Yeah! [She raises her left hand again]
60 M: Alright! And then (well go to bed?).
61 K: Hmm. [She nods]
62 M: Ok. So you have fun today?
63 K: Yeah. [She nods slightly]
64 M: Alright. Bye!
65 K: [She leaves the sofa and comes close to the camera staring at it]
66 Bye!
Appendix B: Transcription Conventions

The following conventions are used in the presentations of transcriptions other than
those listed in appendix A. The transcription conventions are those of standard
conversation analysis enhanced for the transcription of prosodic features (Selting et
al. 1998). Unless modified, all words are written with small letters.

Notation Description Example

(0.14) Time without talk, in seconds okay. (0.24) OH
((turns)) Verbs and descriptions in double ((head sidward))
parentheses are transcribers
comments
(.) Period in parentheses marks a NO. (.) <<assertive>you
hearable pause less than 0.1
seconds long
:: Colons indicate lengthening of U::M:::
phoneme, about 1/10 of a second
per colon
[ ] Square brackets in consecutive J: wa[xes.]
lines indicate overlap C: [wAX?] (0.93)
<p> > Piano, lower than normal speech <<p>like we get hot air
volume from.>
<<pp> > Pianissimo, a lot lower than <<pp>kay.>
normal speech volume, almost
inaudible
<<all> > Allegro, words are uttered with <<all>okay?>
faster than normal speed
<<len> > Lento, slower than normal <<len>it has (0.22)
speech bin=decided that it is too
dIFficult.>
<<plaintive> Transcribers glosses are provid- <<plaintive>i=just=found=
> ed for ways of speaking out=a=way=to=dO=`it.>
JAne Capital letters indicate empha- change of plA:Ns
sized sounds.
232 APPENDIX B

.hh Noticeable in-breath ..HHH UUm (0.42)

hh Noticeable out-breadth
,?;. Punctuation is used to mark T: so can we tell a shape
movement of pitch (intonation) by its color?
T: does it belong to anoth-
toward end of utterance, flat, er group (0.67) O:r.
slightly and strongly upward,
and slightly and strongly down-
ward, respectively
Downward and upward arrows (0.36) O::N::; IT (0.30)
indicate pitch jump downward
and upward
= Equal sign indicates that the loo::ks=similar
phonemes of different words are
not clearly separated
` Diacritic indicates movement of buthh
pitch within the word that fol-
lows down-up, up, down
References

Antaki, C., Biazzi, M., Nissen, A., & Wagner, J. (2008). Accounting for moral judgments in academic
talk: The case of a conversation analysis session. Text & Talk, 28, 130.
Ashmore, M., & Reed, D. (2000). Innocence and nostalgia in conversation analysis: The dynamic rela-
tions of tape and transcript. FQS: Forum Qualitative Social Research / Forum Qualitative Sozi-
alforschung, 1(3). Accessed at http://www.qualitative-research.net/index.php/fqs/article/view/1020/
2199
Au, K. (1980). Participation structures in a reading lesson with Hawaiian children: Analysis of a cultur-
ally appropriate instructional event. Anthropology & Education Quarterly, 11, 91115.
Bakhtine, M. [Volochinov, V. N.] (1977). Le marxisme et la philosophie du langage: essai
dapplication de la mthode sociologique en linguistique [Marxism and philosophy of language: An
essay on the application of the sociological method in linguistics]. Paris: Les ditions de Minuit.
Boden, D. (1994). The business of talk: Organization in action. Cambridge: Polity Press.
Bohnsack, R. (2010a). Documentary method and group discussion. In R. Bohnsack, N. Pfaff, & W.
Weller (Eds.), Qualitative analysis and documentary method in international educational research
(pp. 99124). Opladen, Germany: Barbara Budrich Publishers.
Bohnsack, R. (2010). The interpretation of pictures and the documentary method. In R. Bohnsack, N.
Pfaff, & W. Weller (Eds.), Qualitative analysis and documentary method in international educa-
tional research (pp. 267292). Opladen, Germany: Barbara Budrich Publishers.
Bohnsack, R., Pfaff, N., & Weller, W. (Eds.). (2010). Qualitative analysis and documentary method in
international educational research. Opladen, Germany: Barbara Budrich Publishers.
Bourdieu, P. (1980). Le sens pratique [The practical sense]. Paris, France: Les ditions de Minuit.
Bourdieu, P. (1992). The practice of reflexive sociology (The Paris workshop). In P. Bourdieu and
L.J.D. Wacquant, An invitation to reflexive sociology (pp. 216260). Chicago, IL: University of
Chicago Press.
Collins, R. (2004). Interaction ritual chains. Princeton, NJ: Princeton University Press.
Couzin-Frankel, J. (2013). The power of negative thinking. Science, 342, 6869.
Davydov, V. V. (1988). The concept of theoretical generalization and problems of educational psychol-
ogy. Studies in Soviet Thought, 36, 169202.
Davidson, D. (1986). A nice derangement of epitaphs. In E. Lepore (Ed.), Truth and interpretation (pp.
433446). Oxford, UK: Blackwell.
Derrida, J. (1996). Le monolinguisme de lautre ou la prothse dorigine. Paris, France: Galile.
DeWitt, J., & Osborne, J. (2010). Recollections of exhibits: Stimulated-recall interviews with primary
school children about science centre visits. International Journal of Science Education, 32, 1365
1388.
Dostoevsky, F. (1994). A writers diary. Evanston, IL: Northwestern University Press.
DuBray, W., & Sanders, A. (1999). Interactions between American Indian ethnocity and health care.
Journal of Health & Social Policy, 10, 6784.
234 REFERENCES

Durkheim, . (1919). Les rgles de la mthode sociologique (7ime ed) [Rules of sociological method,
7th ed.]. Paris, France: Flix Alcan.
Eco, U. (1984). Semiotics and the philosophy of language. Bloomington, IN: Indiana University Press.
Edwards, D., & Potter, J. (1992). Discursive psychology. London, UK: Sage.
Elman, J. L. (1993). Learning and development in neural networks: The importance of starting small.
Cognition, 48, 7199.
Erickson, F. (1982). Money tree, lasagna bush, salt and pepper: Social construction of topical cohesion
in a conversation among Italian Americans. In D. Tannen (Ed.), Analyzing discourse: Text and talk
(pp. 4370). Washington, DC: Georgetown University Press.
Feyerabend, P. (1993). Against method (3rd ed.). New York, NY: Verso Books.
Foucault, M. (1969). Larchologie du savoir [Archeology of knowledge]. Paris, France: Gallimard.
Foucault, M. (1975). Surveiller et punir: Naissance de la prison [Survey and punish: Birth of the pris-
on]. Paris, France: Gallimard.
Garfinkel, H. (1967). Studies in ethnomethodology. Englewood Cliffs, NJ: Prentice-Hall.
Garfinkel, H. (1996). Ethnomethodologys program. Social Psychology Quarterly, 59, 521.
Garfinkel, H. (2002). Ethnomethodologys program: Working out Durkheims aphorism. Lanham, MD:
Rowman and Littlefield.
Garfinkel, H. (2007). Lebenswelt origins of the sciences: Working out Durkheims aphorism. Book
Two. Workplace and documentary diversity of ethnomethodological studies of work and sciences
by ethnomethodologys authors: What did we do? What did we learn? Human Studies, 30, 956.
Garfinkel, H., & Sacks, H. (1986). On formal structures of practical action. In H. Garfinkel (Ed.), Eth-
nomethodological studies of work (pp. 160193). London, UK: Routledge & Kegan Paul.
Glaser, B. G., & Strauss, A. L. (1999). The discovery of grounded theory: Strategies for qualitative
research. New York, NY: Aldine de Gruyter.
Goulart, M.I.M., & Roth, W.-M. (2006). Margin|center: Toward a dialectic view of participation. Jour-
nal of Curriculum Studies, 38, 679700.
Heidegger, M. (1977). Sein und Zeit [Being and time]. Tbingen, Germany: Max Niemeyer.
Heidegger, M. (1985). Gesamtausgabe. I. Abteilung: Verffentlichte Schriften 1910-1976. Band 12:
Unterwegs zur Sprache [Complete works I. Part: Published writings 19101976. Vol 12: On the
way to language]. Frankfurt/M, Germany: Vittorio Klostermann.
Heidegger, M. (2006). Gesamtausgabe I. Abteilung: Verffentlichte Schriften 19141976 Band 11:
Identitt und Differenz [Collected works, Division I: Published writings 19101976 vol 11. Identity
and difference]. Frankfurt/M, Germany: Vittorio Klostermann.
Holzkamp, K. (1977). Kann es im Rahmen der marxistischen Theorie eine Kritische Psychologie
geben? [Can there be a critical psychology in the framework of Marxist theory?]. Das Argument,
103, 316336.
Holzkamp, K. (1983). Grundlegung der Psychologie [Founding/Foundation of psychology]. Frank-
furt/M, Germany: Campus.
Husserl, E. (1976a). Husserliana Band III/1. Ideen zu einer reinen Phnomeologie und phnomenolo-
gischen Philosophie. Erstes Buch. Text der 1.3. Auflage [Husserliana vol. III/1. Ideas to a pure
phenomenology and phenomenological philosophy. First volume. Text of the 1st3rd ed.]. The
Hague, The Netherlands: Martinus Nijhoff.
Husserl, E. (1976b). Husserliana Band VI: Die Krisis der europischen Wissenschaften und die
transzendentale Phnomenologie. Eine Einleitung in die phnomenologische Philosophie [Husserli-
ana vol. 4. The crisis of the European sciences and transcendental philosophy. An introduction to
phenomenological philosophy]. The Hague, The Netherlands: Martinus Nijhoff.
Husserl, E. (1991). On the phenomenology of the consciousness of internal time (19831917). Dor-
drecht, The Netherlands: Kluwer Academic Publishers.
Husserl, E. (2008). Die Lebenswelt: Auslegungen der vorgegebenen Welt und ihrer Konstitution. Texte
aus dem Nachlass (19161937) [The lifeworld: Interpretations of the pregiven world and its consti-
tution. Texts from the estate (19161937)]. Dordrecht, The Netherlands: Springer.
Ilenkov, E. V. (1977). Dialectical logic: Essays on its history and theory. Moscow, USSR: Progress
Publishers.
Ingold, T. (2011). Being alive: Essays on movement, knowledge and description. London, UK:
Routledge.
Jefferson, G. (1989). Preliminary notes on a possible metric which provides for a standard maximum
silence of approximately one second in conversation. In D. Roger & P. Bull (Eds.), Conversation:
An interdisciplinary perspective (pp. 166196). Clevedon, UK: Multilingual Matters.
 REFERENCES 235

Jordan, B. (1989). Cosmopolitical obstetrics: Some insights form the training of traditional midwives.
Social Science in Medicine, 28, 925944.
Jordan, B., & Henderson, A. (1995). Interaction analysis: Foundations and practice. Journal of the
Learning Sciences, 4, 39103.
Kant, I. (1968). Werke IV: Kritik der reinen Vernunft [Works vol. 4: Critique of pure reason]. Frank-
furt/M, Germany: Suhrkamp.
Langemeyer, I., & Nissen, M. (2011). Activity theory. In B. Somekh & C. Lewin (Eds.), Theory and
methods in social research (pp. 182-189). London, UK: Sage.
LaVan, S.-K., & Beers, J. (2005). Videos als Katalysatoren fr kogenerative Dialoge im Naturwissen-
schaftsunterricht [Videos as catalyzers for cogenerative dialogue in school science]. In M. Welzel &
H. Stadler (Eds.), Nimm doch mal die Kamera!: Zur Nutzung von Videos in der Lehrerbildung
Beispiele und Empfehlungen aus den Naturwissenschaften (pp. 191205). Mnster, Germany:
Waxmann.
Lave, J., & Wenger, E. (1991). Situated learning: Legitimate peripheral participation. Cambridge, UK:
Cambridge University Press.
Lemke, J. L. (1990). Talking science: Language learning and values. Norwood, NJ: Ablex.
Leontev, A. N. (1983). Dejatelnost. Soznanie. Linost. [Activity, consciousness, personality]. In
Izbrannye psixhologieskie proizvedenija vol. 2 (pp. 94231). Moscow, Russia: Pedagogika.
Leontjew, A. N. (1964). Probleme der Entwicklung des Psychischen [Problems in the development of
the psyche]. Berlin, Germany: Volk und Wissen.
Livingston, E. (1995). An anthropology of reading. Bloomington, IN: Indiana University Press.
Louca, L. T., Zacharia, Z. C., & Constantinou, C. P. (2011). In quest of productive modeling-based
learning discourse in elementary school science. Journal of Research in Science Teaching, 48, 919
951.
MacBeth, D. (2003). Hugh Mehans Learning Lessons reconsidered: On the differences between the
naturalistic and critical analysis of classroom discourse. American Educational Research Journal,
40, 239280.
Mannheim, K. (1952). Essays on the sociology of knowledge. London, UK: Routledge & Kegan Paul.
Mannheim, K. (2004). Beitrge zur Theorie der Weltanschauungs-Interpretation [Contributions to the
theory of worldview interpretation]. In J. Strbing & B. Schnettler (Eds.), Methodologie interpreta-
tiver Sozialforschung: Klassische Grundlagentexte (pp. 103153). Konstanz, Germany: UVK.
Markaki, V., Merlino, S., Mondada, L., & Olof, F. (2010). Laughter in professional meetings: The
organization of an emergent ethnic joke. Journal of Pragmatics, 42, 15261542.
Marx, K./Engels, F. (1958). Werke 3 [Works vol. 3]. Berlin, Germany: Dietz.
Marx, K./Engels, F. (1962). Werke 23 [Works vol. 23]. Berlin, Germany: Dietz.
Marx, K./Engels, F. (1983). Werke 42 [Works vol. 42]. Berlin, Germany: Dietz.
Mavin, T. J., & Roth, W.-M. (2014). A holistic view of cockpit performance: An analysis of the as-
sessment discourse of flight examiners. International Journal of Aviation Psychology, 24, 210227.
McDermott, R.P., Gospodinoff, K., & Aron, J. (1978). Criteria for an ethnographically adequate de-
scription of concerted activities and their contexts. Semiotica, 24, 245275.
Mehan, H. (1979). Learning lessons: Social organization in the classroom. Cambridge, MA: Harvard
University Press.
Meshcheryakov, A. (1979). Awakening to life: On the education of deaf-blind children in the Soviet
Union. Moscow, USSR: Progress Publishers.
Nickerson, R. S. (1998). Confirmation bias: A ubiquitous phenomenon in many guises. Review of Gen-
eral Psychology, 2, 175220.
Nietzsche, F. (1954) Werke in drei Bnden [Works in three volumes]. Munich, Germany: Hanser.
Popper, K. R. (1962). Conjectures and refutations: The growth of scientific knowledge. New York, NY:
Basic Books.
Rawls, A. W. (2002). Editors introduction. In H. Garfinkel, Ethnomethodologys program: Working
out Durkheims aphorism (A. W. Rawls, ed.) (pp. 164). Lanham, MD: Roman & Littlefield
Reichertz, J. (2013). Induction, deduction, abduction. In U. Flick (Ed.), The Sage handbook of qualita-
tive data analysis (pp. 123135). London, UK: Sage.
Rorty, R. (1989). Contingency, irony, and solidarity. Cambridge, UK: Cambridge University Press.
Roth, W.-M. (1993). Metaphors and conversational analysis as tools in reflection on teaching practice:
Two perspectives on teacher-student interactions in open-inquiry science. Science Education, 77,
351373.
Roth, W.-M. (2001). Designing as distributed process. Learning and Instruction, 11, 211239.
Roth, W.-M. (2006). A dialectical materialist reading of the sign. Semiotica, 160, 141171.
236 REFERENCES

Roth, W.-M. (2008). The nature of scientific conceptions: A discursive psychological perspective.
Educational Research Review, 3, 3050.
Roth, W.-M. (2009). Radical uncertainty in scientific discovery work. Science, Technology & Human
Values, 34, 313336.
Roth, W.-M. (2010). Language learning context: Talking the talk. London, UK: Routledge.
Roth, W.-M (2012). Limits to general expertise: A study of in- and out-of-field graph interpretation. In
C. A. Wilhelm (Ed.), Encyclopedia of cognitive psychology (pp. 311348). Hauppauge, NY: Nova
Science.
Roth, W.-M. (2013a). What more? in/for science education: An ethnomethodological perspective. Rot-
terdam, The Netherlands: Sense Publishers.
Roth, W.-M. (2013b). To event: Towards a post-constructivist approach to theorizing and researching
curriculum as event*-in-the-making. Curriculum Inquiry, 43, 388417.
Roth, W.-M. (2013c). Undoing decontextualization or how scientists come to understand their own
data/graphs. Science Education, 97, 80112.
Roth, W.-M. (2013d). Meaning and mental representation: A pragmatic approach. Rotterdam, The
Netherlands: Sense Publishers.
Roth, W.-M. (2013e). Toward a post-constructivist ethics in/of teaching and learning. Pedagogies: An
International Journal, 8, 103125.
Roth, W.-M. (2013f). Technology and science in classroom and interview talk with Swiss lower sec-
ondary school students: a Marxist sociological approach. Cultural Studies of Science Education, 8,
433465.
Roth, W.-M. (2013g). An integrated theory of thinking and speaking that draws on Vygotsky and Bakh-
tin/Voloinov. Dialogical Pedagogy, 1, 3253.
Roth, W.-M. (2014a). Curriculum*-in-the-making: A post-constructivist perspective. New York, NY:
Peter Lang.
Roth, W.-M. (2014b). Working out the interstitial and syncopic nature of the human psyche: On the
analysis of verbal data. Integrative Psychological and Behavioral Science, 48, 283298.
Roth, W.-M. (2014c). Science language Wanted Alive: Through the dialectical/dialogical lens of Vygot-
sky and the Bakhtin circle. Journal of Research in Science Teaching, 51, 10491083.
Roth, W.-M., & Alexander, T. (1997). The interaction of students' scientific and religious discourses:
Two case studies. International Journal of Science Education, 19, 125146.
Roth, W.-M., & Bowen, G. M. (1995). Knowing and interacting: A study of culture, practices, and
resources in a grade 8 open-inquiry science classroom guided by a cognitive apprenticeship meta-
phor. Cognition and Instruction, 13, 73128.
Roth, W.-M., Goulart, M. I. M., & Plakitsi, K. (2013). Science during early childhood: A cultural-
historical perspective. Dordrecht, The Netherlands: Springer.
Roth, W.-M., & Hsu, P-L. (2010). Analyzing communication. Rotterdam, The Netherlands: Sense Pub-
lishers.
Roth, W.-M., Lawless, D., & Masciotra, D. (2001). Spielraum and teaching. Curriculum Inquiry, 31,
183207.
Roth, W.-M., & Mavin, T. J. (2014). Peer assessment of aviation performance: Inconsistent for good
reasons. Cognitive Science. DOI: 10.1111/cogs.12152
Roth, W.-M., & Middleton, D. (2006). The making of asymmetries of knowing, identity, and accounta-
bility in the sequential organization of graph interpretation. Cultural Studies of Science Education,
1, 1181.
Roth, W.-M., & Radford, L. (2010). Re/thinking the zone of proximal development (symmetrically).
Mind, Culture, and Activity, 17, 299307.
Roth, W.-M., Ritchie, S. M., Hudson, P., & Mergard, V. (2011). A study of laughter in science lessons.
Journal of Research in Science Teaching, 48, 437458.
Roth, W.-M., & Roychoudhury, A. (1993). The development of science process skills in authentic
contexts. Journal of Research in Science Teaching, 30, 127152.
Roth, W.-M., & Thom, J. (2009). The emergence of 3d geometry from children's (teacher-guided) clas-
sification tasks. Journal of the Learning Sciences, 18, 4599.
Roth, W.-M., Tobin, K., & Ritchie, S. (2008). Time and temporality as mediators of science learning.
Science Education, 92, 115140.
Roychoudhury, A., & Roth, W.-M. (1996). Interactional processes in a constructivist physics lab. Inter-
national Journal of Science Education, 18, 423445.
Sacks, H. (1995). Lectures on conversation vols. I & II. Malden, MA: Blackwell.
 REFERENCES 237

Selting, M., Auer, P., Barden, B., Bergmann, J., Couper-Kuhlen, E., Gnthner, S., Meier, C., Quasthoff,
U., Schlobinski, P., & Uhmann, S. (1998). Gesprchsanalytisches Transkriptionssystem [Conversa-
tion analytic transcription system]. Linguistische Berichte, 173, 91122.
Smith, D. E. (1981). On sociological description: A method from Marx. Human Studies, 4, 313337.
Smith, D. E. (1983). No one commits suicide: Textual analysis of ideological practices. Human Studies,
6, 309359.
Smith, D. E. (1990a). Conceptual practices of power: A feminist sociology of knowledge. Toronto,
Ontario: University of Toronto Press.
Smith, D. E. (1990b). Texts, facts and femininity: Exploring the relations of ruling. London, UK:
Routledge.
Smith, D. E. (1999). Writing the social: Critique, theory, and investigations. Toronto, Ontario: Univer-
sity of Toronto Press.
Stith, I., & Roth, W.-M. (2008). Students in action: Cogenerative dialogues from secondary to elemen-
tary schools. Rotterdam, The Netherlands: Sense Publishers.
Suchman, L. (2007). Human-machine reconfigurations: Plans and situated actions. Cambridge, UK:
Cambridge University Press.
Tateo, L., & Marsico, G. (2013, May). Navigating the unknown river of development. Paper presented
at the 15th meeting of the International Society for Theoretical Psychology, Santiago, Chile.
Tobin, K. (1987). The role of wait time in higher cognitive level learning. Review of Educational Re-
search, 57, 6995.
van Eijck, M., & Roth, W.-M. (2011). Cultural diversity in science education through novelization:
Against the epicization of science and cultural centralization. Journal of Research in Science Teach-
ing, 48, 824847.
Voloinov, V. N. (1930). Marksizm i filisofija jazyka: sonovye problemy sociologieskogo metoda b
nauke o jazyke [Marxism and the philosophy of language: Main problems of the sociological meth-
od in linguistics]. Leningrad, USSR: Priboj.
Vygotskij, L. S. (1934). Mylenie i re: Psixologieskie isslodovanija [Thinking and speech: Psycho-
logical investigations]. Moscow, USSR: Gosudarstvennoe Socialno-Ekonomieskoe Isdatelstvo.
Vygotskij, L. S. (2005). Psyxologija razvitija eloveka [Pyschology of human development]. Moscow:
Eksmo.
Walkerdine, V. (1998). Counting girls out: Girls and mathematics. London, UK: Falmer Press.
Wittgenstein, L. (1978). Tractatus logico-philosophicusLogisch-philosophische Abhandlung 13. Aufl.
Frankfurt/M, Germany: Suhrkamp.
Wittgenstein, L. (1997). Philosophische Untersuchungen / Philosophical investigations (2nd ed.). Ox-
ford, UK: Blackwell.
Woititz, J. G. (1976). A study of self-esteem in children of alcoholics. Unpublished doctoral dissertation,
Rutgers University.
Index
A
15, 23, 24, 42, 48, 114, 123, 127, 141,
Abduction, 97, 99, 166, 167, 168, 170,
179
Account, xi, xiii, 14, 16, 17, 18, 21, 24,
66, 67, 69, 71, 80, 90, 91, 96, 113,
126, 143, 147, 149, 150, 160, 163,
173, 183, 193, 194, 196, 198, 199,
200, 214
Accountability, xi, xii, xiii, xiv, xv, 7, 9,
10, 14, 16, 17, 18, 19, 21, 24, 30, 31,
37, 41, 45, 48, 49, 54, 66, 67, 69, 71,
74, 80, 90, 91, 96, 113, 126, 143, 147,
149, 150, 160, 162, 163, 173, 183,
193, 194, 196, 198, 199, 200, 208,
214
Activity theory, 80, 111
Assertion, 130, 174, 175, 176, 177, 178,
179, 180, 182
Attribution, 22, 52, 84, 90, 110, 114,
115, 119, 154
Audience, viii, 29, 41, 43, 47, 52, 53, 54,
56, 57, 60, 63, 71, 73, 90, 109, 110,
111, 113, 119, 120, 121, 153, 154,
159
B
Bakhtin, M. M., 111, 146, 147, 153, 154,
161
Bayesian, 41, 46
Bias, 46, 62, 123, 124, 131
Bourdieu, P., xv, 1, 5, 7, 8, 9, 10, 11, 12,
142, 143, 169, 194, 198, 208, 209,
211, 214
Bracketing, xi, 24, 118, 127, 131, 137,
138, 140, 145, 158, 159, 160, 219,
223
C
Circularity, 143, 144, 174, 209, 218
Common sense, 7, 30, 48, 157, 158, 168,
181
Constructivism, 7, 9, 14, 16, 71, 155,
207, 220, 222
Conversation analysis, xi, 8, 31, 65, 117,
118, 150, 159, 160, 161, 217, 231
D
Data session, vii, 33, 73, 93, 109, 225
Deduction, 166, 167
Dialectics, 22, 113, 131, 135, 151, 182,
220
Dope, 70, 74, 109
Drama, 12, 126, 131
Durkheim, ., xii, 79, 81, 183
E
Endogenous, vii, xiv, xv, 82
Ethnographical adequacy, 16, 17, 18, 43,
143, 146, 159, 161
240
Ethnomethodology, xii, xiii, xiv, xv, 7,
10, 31, 71, 74, 81, 142, 150, 160, 209,
215
Explanandum, 62, 140
Explanans, 62, 140
F
Interaction ritual, 119, 152
Fact: social, xi, xii, xiii, 11, 24, 79, 81,
86, 171, 183
Familiarity, vii, viii, xiv, 8, 10, 15, 18,
32, 38, 39, 40, 46, 47, 48, 55, 58, 67,
73, 91, 95, 97, 101, 121, 124, 125,
129, 134, 142, 147, 149, 160, 164,
167, 171, 174, 175, 177, 178, 184,
194, 225
Feyerabend, P., 7, 9, 29, 30
First-time-through, xiii, 58, 59, 83, 140
Formal Analysis, xii, xiii, 10, 141, 150
Formulating, xi, xiii, 16, 18, 39, 48, 54,
82, 102, 115, 125, 135, 137, 147, 149,
155, 158, 159, 177, 180, 181, 196,
198, 211
Foucault, M., viii, xv, 24, 35, 61, 123,
124, 126, 131
G
184, 186, 189, 208, 218, 224
Garfinkel, H., xi, xii, xiii, xiv, 7, 30, 37,
70, 79, 96, 139, 141, 142, 143, 145,
150, 158, 161, 165, 172, 173, 174,
182, 186, 208
Gesticulation, 82
Glossing, xiii, xiv, 129, 131, 137, 145,
147, 150, 151, 152, 160, 196, 197,
231
Grammar, viii, 31, 44, 52, 66, 77, 141,
161, 172, 184, 187, 196, 201
Grounded theory, 167
H
Lebenswelt (lifeworld), xi, xiv, 31, 68,
Heidegger, M., 44, 84, 90, 115, 155, 189
Higher psychological functions, 22, 61
Husserl, E., xiv, 69, 113, 127, 158, 160
Hypothesis, viii, 33, 37, 39, 40, 41, 42,
45, 46, 47, 51, 52, 53, 54, 55, 56, 57,
59, 60, 61, 62, 65, 66, 67, 68, 69, 72,
74, 78, 80, 81, 86, 88, 89, 92, 93, 97,
99, 100, 101, 102, 103, 120, 127, 129,
135, 138, 160, 165, 167, 180, 181,
186
INDEX
I
Ideology, 24, 174, 200, 215, 218, 219,
223
Incommensurability, 14
Induction, 166, 167
Intelligibility, 47, 115
Interjection, 130, 133, 134, 136, 193,
223
Interrater reliability, 142
Intersubjectivity, 7, 71, 107, 115, 116,
207, 214
Intonation, 232
Invariant, 11, 13, 15, 37, 39, 60, 98, 100
IRE, xiii, 41, 44, 129, 140, 146, 147,
148, 149, 150, 151, 152, 154, 155,
156, 165, 171
Irreducibility, xv, 66, 79, 116, 120, 139,
146, 148, 149, 162
K
Knowledge, vii, 13, 14, 20, 23, 24, 30,
35, 45, 46, 52, 59, 86, 87, 123, 125,
126, 128, 129, 130, 131, 138, 139,
140, 142, 146, 165, 172, 173, 182,
Knowledgeability, 67, 85, 134, 135, 137,
139, 142, 144, 172, 184, 186, 209
Knowledge-power, viii, xv, 107, 124,
126, 134, 135, 136, 137, 138, 139,
140, 146, 158
L
Laissez faire, 5, 9, 10
Language game, 20, 21, 22, 31, 87, 89,
91, 120, 137, 164
Language-in-use, 66, 78, 87, 111, 113,
114, 116, 117, 120
70, 71, 163
M
Manifestation, xii, xv, 10, 22, 37, 65, 86,
111, 113, 139, 144, 165, 168, 169,
170, 171, 173, 174, 177, 182, 183,
184, 185, 186
Mannheim, K., xi, 37, 165, 168, 170,
171, 174
 INDEX
Marx, K., 11, 22, 71, 74, 80, 111, 113,
205, 207, 214, 224
Merleau-Ponty, M., xii, 182
Metaphysics, 18, 20, 84, 155, 224
Mind, 7, 21, 32, 36, 38, 40, 42, 52, 73,
90, 110, 115, 117, 173, 174, 207, 208,
211, 212
Monday morning quarterbacking, vii,
139, 140, 194
Mundaneity, xii, xiii, xiv, xv, 9, 15, 19,
20, 31, 32, 62, 71, 79, 81, 88, 97, 100,
109, 116, 117, 127, 134, 141, 142,
157, 158, 159, 163, 164, 171, 172,
173, 194, 207, 209, 215, 219, 221,
223
N
174, 186, 201, 209, 214, 215, 216,
Neural network, 184
Nietzsche, F., 84, 118, 123, 155
P
144, 145, 146, 149, 150, 153
Passing theory, 52, 186, 208
Personality, 22, 61, 207
Phenomenography, xv, 160
Phenomenology, xiv, 127
Pitch, 13, 77, 149, 162, 169, 170, 196,
232
Power, vii, xi, 24, 35, 59, 60, 81, 123,
124, 125, 131, 137, 139, 146, 168,
176, 191, 194, 224
Pragmatism, 20, 22, 145, 175, 186, 198,
207, 208, 214
Praxis, xiii, 11, 39, 96, 141, 214, 217,
224
Preconception, xi, 41, 48, 217
Probability, 41, 46, 54, 55, 58, 81, 86, 97
Prosody, 43, 149, 162, 231
Protocol, 29, 37, 44, 45, 66, 68, 71, 96,
120, 131, 171, 173, 198
Psychology, xi, xii, xiv, 10, 12, 20, 22,
23, 32, 61, 65, 66, 70, 74, 109, 113,
150, 154, 157, 200
R
Utterance, 232
Rationality, 44, 45, 48, 85
Recipient, 21, 53, 58, 73, 90, 113, 115,
118, 153, 161, 169, 171, 200
Reflexivity, 45, 48, 68, 73, 79, 80, 96,
214, 217
241
Relation: institutional, 35, 70, 79, 128,
131, 138, 139, 201; ruling, xv, 82,
209, 215, 221; social, viii, xii, 24, 32,
107, 148, 164, 196, 197, 214, 215,
217, 220, 222, 224; societal, vii, xii,
xiii, xv, 22, 24, 30, 61, 70, 80, 207,
220, 223
Relational: analysis, 214, 221, 223;
thinking, xv, 220
S
Shop floor, 143, 144, 145
Signification, 112, 113
Sociology, xii, xiv, 7, 11, 30, 32, 70, 74,
81, 109, 113, 119, 142, 152, 157, 172,
224
Sound-word, 78, 95, 112, 113, 115, 116,
154
Staff, xiii, xiv, xv, 33, 120, 139, 140,
Struggle, 59, 123, 124, 126, 131, 134,
135, 136, 137, 221
Subjectivity, 14, 24, 47, 52, 170
T
Think-aloud protocol, 29, 31, 45, 138
Transactional, 47, 154, 219, 220
Transcription, vii, viii, ix, 17, 18, 19, 29,
31, 32, 33, 35, 36, 37, 38, 39, 40, 43,
45, 48, 49, 51, 53, 62, 63, 65, 66, 67,
68, 69, 73, 74, 77, 79, 80, 87, 90, 91,
92, 96, 97, 98, 100, 102, 103, 110,
112, 114, 115, 116, 117, 118, 119,
120, 125, 128, 129, 131, 132, 140,
144, 145, 159, 161, 165, 169, 171,
174, 177, 178, 180, 185, 189, 191,
193, 194, 196, 209, 210, 211, 212,
213, 231
Turn-taking, xiv, 41, 80, 120, 146, 162
U
V
Variation, 40, 41, 54, 57, 222
Voloinov, V. N., 78, 111, 112, 149, 161,
201, 214
242 INDEX

Vygotsky, L. S., 12, 14, 21, 22, 61, 62, Wittgenstein, L., 20, 21, 23, 66, 89, 120,
66, 69, 78, 111, 112, 126, 149, 152, 149
199
Vyzkazyvanie (statement), 154 Z

W Zone of proximal development, 62, 124,

126
Witness, 7, 30, 67, 70, 79