 

 

 

PLACE ll RETURN BOX to remove this chock” from your record.
TO AVOID FINES Mom on or More data duo.

DATE DUE DATE DUE DATE DUE

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

PSEUDO-ARGUMENT
By

Brian Kleiner

A DISSERTATION

Submitted to
Michigan State University
in partial fulﬁllment of the requirements
for the degree of

DOCI‘OR OF PHILOSOPHY

Department of Linguistics and Germanic, Slavic,
Asian and African Languages

1 997

ABSTRACT

PSEUDO—ARGUMENT
By

Brian Kleiner

This work carries out a discourse analysis of naturally occurring conversations
between undergraduate students about issues of race, many of which reveal recurrent
identiﬁable underlying structures, linguistic features and functions which lead to the
conclusion that what is involved is a fairly well-deﬁned type of discourse, or ‘speech
activity.’ This type of discourse shares many features in common with argumentative
discourse in that both types involve sequences of the same underlying ideational units
(positions, supports and disputes) which are conﬁgured within larger units (arguments).
On the other hand, the data analyzed in this work are quite distinct from ordinary
argumentative discourse on structural, linguistic and functional grounds. The kind of
discourse involved in the data is called “pseudo-argument,” because participants borrow
elements of ordinary argument in order to shield themselves against negative attributions or
inferences which might arise among the participants. By making their talk appear
argumentative, participants are able to appear rational, unbiased, and nonracist in an

essentially one-sided discussion in which controversial beliefs and attitudes are

continuously put forward and go unchallenged.

Pseudo-argument is a collaborative activity in which participants jointly construct
arguments (in the sense of to “make” an argument), and also simulate arguments (in the
sense of to “have” an argument) in which the views of an absent antagonist are imported
into the conversation and subsequently disputed. It is this collaboration within the
organizational framework of pseudo-argument which allows participants to diffuse,
reinforce, practice and validate their beliefs and values. It is for this reason that the study
of pseudo-argument may contribute to our understanding of discoursal processes having to

do with the reproduction of “modern racism” in everyday talk.

Copyright by
BRIAN KLEINER
1997

ACKNOWLEDGMENTS

I would like to express my gratitude to Dennis Preston, who has been my
chairperson, mentor and role model. It was his work and inspiration which brought the
particulars of this dissertation into existence, and I owe much to him for this. I also thank
Barbara Abbott, Grover Hudson, David Dwyer and James Stalker for their helpful
comments and support. I am indebted to the College of Arts and Letters, which generously
provided two semester-long fellowships over the last two years. This work would not
have been completed in a timely way were it not for their ﬁnancial support. Thanks,
ﬁnally, to Mahide Demirci, for standing by me and sharing so much with me over these

graduate student years.

TABLE OF CONTENTS

CHAPTERl
INTRODUCTION ................................................................................... 1
1.0. Nature of the work .................................................................... 1
1.1. Aims of this work: theoretical and applied ......................................... 2
1.1.1. Theoretical goals: contribution to discourse analysis .................. 2
1.1.2. Applied goals: contribution to the understanding of discoursal
processes in the reproduction of prejudice in everyday talk .......... 3
1.2. Data collection ......................................................................... 8
1.3. Plan of work ......................................................................... 10
CHAPTERZ
DISCOURSE ANALYSIS ........................................................................ 13
2.0. Introduction — analytic tools for investigation .................................. 13
2.1. Discourse analysis — structure versus function ................................. 13
2.2. Coherence and interaction .......................................................... 17
2.3. Larger units of discourse ........................................................... 18
2.3.1. Narrative structure (Labov, 1972) ..................................... 19
2.3.2. Ethnography of communication — speech events ................... 24
2.3.3. Speech activities ......................................................... 26
2.4. Linguistic analysis .................................................................. 29
2.4.1. Schiffrin (1987) .......................................................... 31
2.4.2. Labov (1972) ............................................................. 37
2.4.3. Gumperz (1982) ......................................................... 40
2.5. Two approaches to discourse ...................................................... 44
2.5.1. Speech act theory ........................................................ 44
2.5.2. Conversation analysis ................................................... 49
2.6. Summary ............................................................................. 54
CHAPTER3
ARGUMENTATION .............................................................................. 57
3.0. Review and background ........................................................... 57
3.1. O’Keefe’s ‘two senses of argument’ ............................................. 61
3.2. Features of argumentation ......................................................... 62
3.2.1. Argumentation occurs in interaction ................................... 62
3.2.2. Argumentation is cooperative .......................................... 63
3.2.3. Argumentation is “public” .............................................. 66
3.3. Definitions of argument ............................................................ 68
3.4. Example of argumentation from the data ......................................... 72
3.5. Argument as speech activity ....................................................... 74
3.5.1. The speech act approach (van Eemeren and Grootendorst, 1984;
van Eemeren et al, 1993) ............................................... 75
3.5.1.1. Argument as illocutionary act ............................... 75

vi

3.5.1.2. The stages of critical discussion (arguments made) ...... 80

3.5.1.3. The reconstruction of argumentative discussions ......... 81
3.5.2. Conversation Analytic approach
(Jackson and Jacobs, 1982, 1986) .................................... 86
CHAPTER4
THE STRUCTURE OF PSEUDO-ARGUMENT ............................................. 90
4.0. A structural approach ............................................................... 90
4.1. Why data do not fit the discourse models ........................................ 91
4.1.1. Van Eemeren and Grootendorst’s speech act approach ............. 91
4.1.2. Jackson and Jacobs’ CA approach .................................... 96
4.2. The structure of pseudo-argument ................................................ 98
4.2.1. Collaborative argumentslz making a case together ................. 104
4.2.2. Collaborative Argumentszz Importing and
disputing the antagonist ............................................... 111
4.3. Pseudo-argument versus ordinary argument ................................... 126
4.4. Summary ........................................................................... 129
CHAPTERS
LINGUISTIC PROPERTIES OF PSEUDO-ARGUMENT ................................ 130
5.0. Introduction ........................................................................ 130
5.1. Linguistic analysis of ordinary argumentation ................................. 132
5.1.1. Schiffrin (1985) ........................................................ 132
5.1.2. Schiffrin (1980) ........................................................ 134
5.1.3. Schiffrin (1987) ........................................................ 138
5.2. Linguistic realization of pseudo-argument ...................................... 141
5.2.1. Some features of collaborative argumentl ........................... 142
5.2.1.1. Endorsement and alliance ................................... 144
5.2.1.1.1. Theuseof “exactly”: bracketing and
endorsing general positions ..................... 144
5.2.1.1.2. The use of repetition — endorsing and
bracketing the supports of others .............. 151
5.2.2. Some features of collaborative argument; ........................... 157
5.2.2.1. Importing the antagonist ................................... 158
5.2.2.1.1. Direct quotation .................................. 158
5.2.2.1.3. The discourse marker “whatever” ............. 162
5.2.2.1.3.1. WHATEVER] — listing and
labeling difficulties ..................... 163
5.2.2.1.3.2. WHATEVER2 — marker of
irresolvabilty ............................ 166
5.2.2.133. WHATEVER3 — importing
the antagonist ............................ 168
5.3. Summary ........................................................................... 175

vii

CHAPTER6

THE FUNCTION OF PSEUDO-ARGUMENT .............................................. 178
6.0. Introduction ........................................................................ 178
6.1. Balancing instrumental and identity goals ...................................... 181
6.1.1. Disclaimers .............................................................. 184
6.1.2. Higher level strategies ................................................. 186
6.2. The function of collaborative argumentl ........................................ 187
6.3. The function of collaborative argumentz ........................................ 191
6.4. Summary -— why call it “pseudo-argument” ? ................................. 197
6.5. Conclusion ......................................................................... 200
6.5.1. Possible objections .................................................... 201
6.5.2. Contribution to discourse studies .................................... 205
6.5.2.1. Pseudo-discourse ........................................... 205
6.5.2.2. Discourse analysis of structure,
function and linguistic features ............................ 209
6.5.3. On the applied side — contribution to our
understanding of race talk ............................................. 211
6.5.3.1. Van Dijk’s work on the reproduction of ethnic
prejudice in discourse ...................................... 212
FOOTNOTES ..................................................................................... 222
APPENDICES .................................................................................... 225
LIST OF REFERENCES ........................................................................ 228

viii

Chapter 1

INTRODUCTION

1.0 Nature of this work.

This work attempts to carry out a discourse analysis of natural language data which
involve conversations between undergraduate students about issues of race. Many of these
conversations about race reveal recurrent identiﬁable underlying structures, linguistic
features and functions which lead to the conclusion that what is involved is a fairly well-
deﬁned type of discourse, or ‘speech activity.’ This type of discourse shares many
features in common with argumentative discourse. In particular, both involve sequences of
the same underlying ideational units (positions, supports and disputes) which are
conﬁgured within larger units (arguments). On the other hand, despite this superﬁcial
resemblance, the data to be analyzed in this work are quite distinct from ordinary
argumentative discourse on structural, linguistic and functional grounds. For reasons
which will become apparent, the kind of discourse involved in the data is called “pseudo-
argument.” This is because the resemblance between the two discourse types is more than
coincidence — it will be argued that participants in pseudo-argument borrow elements of
ordinary argument in order to shield themselves against negative attributions or inferences

which might arise among the participants. To put it another way, by making their talk

2
which might arise among the participants. To put it another way, by making their talk
appear argumentative, participants are able to appear rational, unbiased, and nonracist in an
essentially one-sided discussion in which controversial beliefs and attitudes are
continuously put forward and go unchallenged.

Pseudo-argument arises in talk about delicate issues (such as race) in which all
participants are allied in their nonnorrnative beliefs and at the same time are mindful of
countervailing social norms and therefore work to forestall the negative impressions which
can result from holding views which are not entirely acceptable to the larger society.
Pseudo-argument is a collaborative activity in which participants jointly construct
arguments (in the sense of to “make” an argument), and also simulate arguments (in the
sense of to “have” an argument) in which the views of an absent antagonist are imported
into the conversation and subsequently disputed. It is this collaboration within the
organizational framework of pseudo-argument which allows participants to diffuse,
reinforce, practice and validate their beliefs and values. It is for this reason that it will be
claimed in the ﬁnal chapter that in terms of its application, the study of pseudo-argument
may contribute to our understanding of discoursal processes having to do with the

reproduction of “modern racism” in everyday talk.

1.1. Aims of this work: theoretical and applied
1.1.1. Theoretical goals: contribution to discourse analysis

i. The ﬁrst aim of this work is to make a contribution to the developing area of

3

discourse analysis by illustrating how a principled analysis of a larger unit of talk may be
carried out which examines the interrelated structural, functional and linguistic properties of
the discourse. It will be shown that an analysis which examines these different aspects of a
discourse unit together may shed light not only on the unit itself but also on how that unit is
related to other units. In the case at hand, the combined analysis of the structure, function
and particular surface linguistic features of pseudo-argument sheds considerable light on
this kind of discourse, on how these aspects of the discourse are integrated in practice, and
further shows exactly the ways in which the unit is similar to and distinct from ordinary
argumentative discourse.

ii. The second aim of this work is to show how a unit of discourse may take on the
appearance of another unit in order to proﬁt from stereotypical features associated with it.
In this way, various kinds of pseudo-discourse arise in which one type of talk simulates
another’s structures and certain linguistic features, but remains distinct in function.
Pseudo-argument is an example of this process in that it simulates the structures and certain
linguistic features of ordinary argumentative discourse in order to proﬁt from associations
of objectivity, fairness, and rationality, while at the same time fulﬁlling a function quite

different from ordinary argument.

1.1.2. Applied goals: contribution to the understanding of discoursal processes in the
reproduction of prejudice in everyday talk.
Although the bulk of this work treats the content of the data as secondary and the

form as primary in the analysis of pseudo-argument, it will become apparent (in chap. 6)

4

that the particular content of pseudo-argument plays a signiﬁcant role in the construction of
this kind of discourse. And although it is not claimed that pseudo-argument is limited to
talk about race, it will be claimed that it is limited to talk about “delicate” issues (content) in
which participants jointly take nonnormative stands. It is this necessary concern with
content in the study of pseudo-argument which leads to the applied aspects of this work. I
would like to show how the study of pseudo-argument may contribute to our
understanding of discoursal processes having to do with the reproduction of what social
psychologists have called “modern racism” or “aversive racism.” It appears that pseudo-
argument may be an important player in the diffusion and reinforcement of modern racist
beliefs in everyday face-to—face communication.

The conversations analyzed in this work show characteristics of what has variously
been called “modern racism” (McConahay, 1986), “aversive racism” (Gaertner and
Dovidio, 1986) and “symbolic racism” (Kinder and Sears, 1981) by social psychologists.
Unlike “old-fashioned racism,” which was characterized by the straightforward, open

expression of hostility in both word and deed, “modern racism” involves a mix of

conﬂicting values, beliefs and feelingszl

In our view, aversive racism represents a particular type of ambivalence in
which the conflict is between feelings and beliefs associated with a sincerely
egalitarian value system and unacknowledged negative feelings and beliefs about
blacks (Gaertner and Dovidio, 1986, p.62).

According to Gaertner and Dovidio (1986, p.62), because of their strong egalitarian values,

5
aversive racists support public policies which in principle promote racial equality. They
identify with a liberal political agenda, and they think of themselves as nonprejudiced. On
the other hand, aversive racists possess unacknowledged negative feelings and beliefs
about black people. This negative affect is not hatred or hostility but rather discomfort,
disgust or fear.
According to McConahay (1986, pp.92-93), the ideology of modern racism in the

United States includes the following tenets:

1) Discrimination is a thing of the past because blacks now have the
freedom to compete in the marketplace and to enjoy those things they can
afford.

2) Blacks are pushing too hard, too fast and into places where they are not
wanted.

3) These tactics and demands are unfair.
4) Therefore, recent gains are undeserved and the prestige granting

institutions of society are giving blacks more attention and the concomitant
status than they deserve.

McConahay adds two more tenets adhered to by modern racists:

5) Racism is bad and the other beliefs do not constitute racism because these
beliefs are empirical facts.

6) Racism, as deﬁned by modern racists, is consistent only with the tenets
and practices of old-fashioned racism: beliefs about black intelligence,
ambition, honesty, and other stereotyped characteristics, as well as support
for segregation and support for acts of open discrimination.

In sum, aversive or modern racists in the United States are ambivalent between

6

strong egalitarian values and negative feelings and attitudes towards Blacks and other
minorities. This ambivalence helps determine a set of beliefs which compose what may be
called a modern racist ideology. This ideology is characterized mainly by the conviction
that discrimination and racism no longer exist and therefore that any “advantages” given to
Blacks are unfair.

American culture has historically been a racist one, and the continued presence of
anti-Black sentiment in the United States is well documented (see reviews of Katz et al,
1986; Crosby et al, 1980; Dovidio and Gaertner, 1986). Although racial attitudes and
beliefs in this country have progressed beyond the “old-fashioned” kind, issues of race and
discrimination are highly salient ones in contemporary American culture. This is due in
large part to the perception of many White Americans that their long-standing privileges as
members of the dominant group are being threatened. The institutional and interactional
mechanisms underlying the current forms of racism in the United States are less obvious
and more subtle, often masked by appeal to values defendable on nonracial grounds.
Therefore, the question of how modern racism is reproduced at different levels is one
which is worthy of consideration. One facet of this question is how modern racism is
reproduced at the level of face-to-face verbal interaction, since this activity seems to be a
prominent venue for the diffusion and reinforcement of beliefs and attitudes associated with
the modern racist orientation (van Dijk, 1987).

Little attention has been given so far by scholars to how modern racist beliefs are

diffused and reinforced in day-to—day, face-to-face verbal interaction by members of the

7
dominant White majority. With the exception of a few studies (e. g., Van Dijk, 1987;
Essed, 1991; Potter and Wetherell, 1992; Verkuyten et a1, 1994), there appears to be little
known about how the ambivalence involved in modern racism becomes manifest in
everyday discourse. That is, few scholars have addressed the questions of how the conﬂict
of egalitarian values and anti-Black feeling shapes discourse in talk about race and,
conversely, how that particular discourse serves as a vehicle for the reproduction of
modern racist beliefs and attitudes in face-to-face interaction.

It will become apparent in chapters 4-6 that the data of this study give strong
evidence for the modern racist orientation described by social psychologists. In particular,
the content of many of the conversations repeatedly involves appeals to egalitarian values in
the face of “unfair advantages” given to African-Americans. In addition, the conversations
show subtle evidence of the negative feelings attributed to modern racists by the social
psychologists mentioned earlier. Further, participants in the conversations often explicitly
make statements such as “Racism is in past,” “They get all the advantages,” “It’s so
unfair,” “They’re always whining and complaining...,” etc., all of which serve to conﬁrm
McConahay’s summary of the tenets of modern racism. Finally, ample evidence will be
given to support the claim that participants are extremely concerned about avoiding the
appearance of being “racist,” and that much of the form and content of their talk is devoted
to forestalling this impression.

While this work is intended primarily to contribute to our understanding of

discourse and discourse analytic methods, I also believe that it will (secondarily, but

8

perhaps more importantly) contribute to our understanding of how modern racism is
reproduced in everyday talk. We will return to this issue in the ﬁnal chapter of this work
where it will be shown that the speech activity of pseudo-argument is a suitable medium for

the diffusion and reinforcement of modern racist beliefs.

1.2. Data collection

The data were collected in the fall of 1994 as part of a project designed by Professor
Dennis Preston, assisted by the author of this work, and funded by an AURIG (All
University Research Initiation Grant) from the Ofﬁce of the Vice President for Research
and Graduate Studies, and was approved by UCRIHS (University Committee on Research
Involving Human Subjects), both of Michigan State University. The initial pool of
respondents were drawn from a humanities course taught by Professor Preston entitled
‘The roles of language in society.’ Of the 300 students attending the course, 76
volunteered to participate. The students (data gatherers) were instructed to get together with
2 or 3 friends in a relaxed setting, (e.g., a dorm room), and tape record at least 15 minutes
of conversation. The topics of the conversation were to be proposed at the discretion of the
data gatherers following a prepared script which included questions having to do with
issues of race on the Michigan State University campus. The prepared instructions and
questions given to the data-gatherers are included in appendix A. Data-gatherers were
instructed to initiate topics and then to allow the discussion of each issue take its course. In
addition, the data-gatherers were encouraged to actively participate in the discussions.

The students and their respective friends who participated were undergraduates, most

9

of whom were White, presumably middle class and young. All of the conversations
recorded were between ethnic sarnes. That is, all conversations were either all-White or all-
Black (and in a few cases all-Asian). We requested this because we were after the most
uninhibited, spontaneous speech possible, and mixing races (ethnic groups) would
probably result in a (no less interesting but) less honest, less forthright discussion of the
issues. Our method of having students interview their friends was deliberately chosen for
the same reason. That is, this method allows for a minimizing of observer’s inﬂuence on
the speech behavior of respondents. Without the presence on an authoritative outsider as
interviewer, we hoped that the conversations would approach more closely the natural, less
self-conscious level of ordinary talk among friends (see chap. 6). We hoped to capture
spontaneous behavior, natural discourse and honest beliefs.

These hopes were not disappointed — the conversations were surprisingly open,
spontaneous and sometimes apparently racist. It seems that despite the presence of a tape-
recorder, the groups of friends engaged in the kind of sincere, animated, heartfelt
discussions which only take place behind closed doors. There were very few references in
the transcriptions to the tape-recorder, and while the inﬂuence of recording cannot be
ignored, it seems safe to say that the speech of the respondents reﬂects their ordinary way
of carrying on in similar situations.

The tapes were collected and each was summarized for tapical content. As a result, we
have a comprehensive description of the most commonly discussed topics and respondents’

positions regarding those topics. Although this work is concerned less with the content of

10

these discussions than their organization and structure, this topical analysis is useful in that
it may be used as a gauge of the degree to which the positions taken by participants in the
data are representative of the views of the population of students sampled in this project.
Although the corpus overall reﬂects a broad spectrum of beliefs and attitudes by White
college students about issues of race on the University campus and beyond, it should be
noted that the excerpts to be presented and analyzed in what follows in this paper
(especially in chapters 4, 5, and 6) are representative of many, if not most, of the
conversations in the larger corpus both in terms of their content and form.

After the topical analysis was completed, the tapes with the best quality recording as
well as the richest content were transcribed. Tapes were chosen with the hope of their
being representative of a wide range of views, from the most anti-racist to the most
(modern) racist conversations. Six of the tapes were fully transcribed (about 120 pages)
according to transcription conventions set down by Jefferson 1973, with certain
modiﬁcations. The detail of the transcription was limited to representing interruptions,
pauses, emphasis, clause-ﬁnal intonational contours, and overlaps. All segmental material
from false starts, self repairs, etc., was preserved in orthographic form. The transcription

conventions are included in appendix B.

1.3. Plan of this work
This work will proceed as follows; chapter 2-will present the discourse analytic
tools and assumptions which will guide the analysis of pseudo-argument. Here, the roles

of underlying structure, function and surface linguistic detail in discourse will be examined,

11
as well as how these aspects of discourse have been treated in previous work. In addition,
discussion will touch on the notions of coherence and interaction in discourse, and it will
be shown that these properties of discourse are closely interrelated. Then discussion will
focus on the notion of the discourse unit, that is, a bounded larger unit of talk with distinct
organizational and linguistic properties. Finally, this preliminary chapter will review two
prominent approaches to discourse which will come into play later in the work; speech act
theory and conversational analysis.

Chapter 3 will also be preliminary in that it will introduce the contemporary ﬁeld of
argumentation theory and show that there is some consensus among scholars that
argumentative discourse has certain properties which distinguish it from other types of
discourse. After providing a deﬁnition of argumentation, two discourse analytic models of
argument will be presented; the ﬁrst will be the speech act theory approach of van Eemeren
and Grootendorst (1984) and van Eemeren et a1 (1993), and the second will be the
conversational analytic approach of Jackson and Jacobs, (1982, 1986).

In chapter 4 the analysis proper begins by comparing several episodes from the data
with ordinary argumentation with respect to the speech act theory and conversation analytic
models reviewed in chapter 3. It will be shown that the data do not neatly ﬁt either of these
models. The remainder of chapter 4 will involve an analysis of the underlying structure of
pseudo-argument. The data give evidence for two distinct recurrent structural patterns or
conﬁgurations of underlying units (positions, supports and disputes); these conﬁgurations

are called collaborative argument] (CA1) and collaborative argumentz (CA2).

12

Chapter 5 will involve the analysis of certain syntactic and lexical surface linguistic
features of pseudo-argument. It will be argued that these features serve the purpose of
bracketing underlying structures of the discourse. It is further claimed that these linguistic
phenomena provide evidence for the underlying structures proposed for pseudo-argument
(PA), and that they are important resources for participants in the construction and
understanding of the discourse they are involved in. It is also claimed that these features
serve as further evidence that PA is distinct from ordinary argumentation (OA), because the
features analyzed do not commonly occur in 0A.

The ﬁrst part of chapter 6 will examine the function of pseudo-argument. It will
ultimately be argued that much of the structure of PA (including the structure of CA1 and
CA2) is due to identity work on the part of participants in an effort to forestall negative
inferences by others and to project an image of rationality, objectivity and fairness. That is,
it will be shown that argument structures are employed by participants in PA primarily for
the purpose of face-work and only secondarily for the instrumental purpose of persuasion
(as in ordinary argument). It will then be argued that beyond this facade of fairness and
rationality is an activity whose main purpose is to allow participants to simulate, exchange,
reinforce and practice beliefs and arguments in a nonhostile environment. At the end of
chapter 6, conclusions are drawn regarding ways in which this work contributes to the
developing area of discourse analysis and how the study of pseudo-argument may give
insight into the discoursal processes involved in the reproduction of the modern racist

ideology in everyday face-to-face interaction.

Chapter 2

DISCOURSE ANALYSIS

2.0. Introduction — analytic tools for investigation

The aim of this chapter is to provide the discourse-analytic tools which will be
employed in the investigation to follow in later chapters. This will involve spelling out the
theoretical background(s) and some of the assumptions which underlie this research
project. Following some discussion of what is meant by ‘structure,’ ‘function,’ and ‘larger
units’ in discourse analysis, a case will be made for the contribution of low-level linguistic
analysis to discourse studies. Following that, I will brieﬂy review two prominent
approaches to discourse (speech act theory, and conversation analysis) which will play a
role in the analysis of the data of this study. Finally, I will summarize by presenting a list
of discourse-analytic assumptions which will later serve to guide the analysis of pseudo-

argument.

2.l. Discourse analysis — Structure versus Function
Discourse analysis is a vast and underdeveloped area, and the various approaches to
discourse which have arisen are based on scholarship from many disciplines including

linguistics, philosophy, anthropology, and sociology among others (Schiffrin, 1994, p.5;

l3

14

van Dijk, 1987, p. 13). Although a thorough review of discourse analysis is beyond the
scope of this paper (for reviews, see Schiffrin, 1994; Stubbs, 1983; Potter and Wetherall,
1987), in what follows I would like to touch upon those elements of discourse analysis
which will be relevant to the arguments set forth in later chapters.

The various approaches to discourse are quite different in terms of their origins,
methods, theoretical assumptions, concepts, and goals (Schiffrin, 1994, p.13). Following
Schiffrin (1994), these approaches (including Conversation Analysis, Ethnography of
Communication, Pragmatics, Variation Analysis, Interactional Sociolinguistics, and Speech
Act Theory, among others) take different positions with regard to a variety of issues such
as the relationship between structure and function, text and context, and the role of
intentionality and intersubjectivity in communication. Given this fragmentation, discourse
analysis is far from becoming a uniﬁed ﬁeld. Nonetheless, there are recurrent themes and
assumptions which are shared by the different approaches which give shape to the ﬁeld,
and allow for certain generalizations about the parameters and goals of discourse analysis.

For example, Stubbs (1983, p.1) asserts that:

Roughly speaking, it refers to attempts to study the organization of language above
the sentence or above the clause, and therefore to study larger linguistic units.... It
follows that discourse analysis is also concerned with language in use in social
contexts, and in particular with interaction or dialogue between speakers.

Most of the approaches to discourse would abide by this general two-part deﬁnition,

although probably with emphasis on either the ﬁrst part or the second. The ﬁrst part claims

15

that discourse analysis studies the organization of language above the sentence level.
Those approaches which emphasize this part are more concerned with structure and form,
whereas those approaches which emphasize language in use are more concerned with
function.

Schiffrin (1994) notes that these divergent emphases derive from two ‘paradigms’
in linguistics — the ‘structuralist’ paradigm and the ‘functionalist’ paradigm; “These two
paradigms make different background assumptions about the goals of a linguistic theory,
the methods for studying language, and the nature of data and empirical evidence” (p.20).
Structuralist analyses of discourse seek to discover levels of structure larger than the
sentence by identifying units or constituents with particular relationships with each other.
Structuralist approaches employ traditional methods of linguistic analysis with the
assumption that discourse structures belong to a uniﬁed hierarchy of structures beginning
with sounds and morphemes and moving to clauses, sentences and so on.

On the other hand, those approaches committed to the functionalist paradigm view
discourse as inseparable from the purposes of language in human life. Deﬁning discourse
as language use entails viewing discourse as a “system (a socially and culturally organized
way of speaking) through which particular functions are realized” (Schiffrin, 1994, p. 32).
In this way, the study of discourse requires looking beyond formal relations within the
linguistic system to consider the embeddedness of language in interaction between
individuals with particular aims in particular contexts and cultures.

Schiffrin (1994) summarizes these distinct emphases as follows:

16

The availability of two different perspectives - stemming from two different ways
of deﬁning discourse - is partially responsible for the tremendous scope of
discourse analysis. If we focus on structure, our task is to identify and analyze
constituents, determine procedures for assigning to utterances a constituent status,
discover regularities underlying combinations of constituents (perhaps even
formulating rules for producing those regularities), and make principled decisions
about whether or not particular arrangements are well formed. If we focus on
function, on the other hand, our task is to identify and analyze actions performed by
people for certain purposes, interpret social, cultural, and personal meanings, and
justify our interpretations of those meanings for the participants involved. Dealing
with either structure of function alone is thus a hefty task: but dealing with both can
take us into two different analytical worlds that are often difﬁcult to integrate.

(p.42)

That there may by some bridge between the ‘structuralist’ and ‘functionalist’ approaches to
discourse is suggested by Stubbs’ deﬁnition given above in which he states that the study
of language use follows from the study of larger units of language. Although Stubbs does
not make this connection explicit, I take it to mean that the study of (the structure of) larger
units necessarily leads us to investigate the social and interactive forces which shape and at
the same time make use of those units. Schiffrin explicitly attempts to bridge the gap
between the ‘structuralist’ and ‘functionalist’ approaches by deﬁning discourse as
‘utterances.’ She views utterances as “...units of language production (whether spoken or .
written) that are inherently contextualized...” (p.41). By viewing discourse as comprised
of utterances rather than sentences, “... we can suggest that discourse arises not as a
collection of decontextualized units of language structure, but as a collection of inherently
contextualized units of language use” (p.39).

Schiffrin’s use of ‘utterance’ seems to provide a good meeting point for the two
paradigms, and her insight will be made use of in this paper. In the analysis which will

follow in subsequent chapters, it will be shown that both a Structuralist approach (chap. 4)

17

and a functionalist approach (chap. 6) shed considerable light on the data, and, further, that
the views from both of these perspectives are not only compatible but also necessary for a

full understanding of the discourse.

2.2. Coherence and Interaction

There are several other general aspects of discourse analysis which will be
discussed at this point, given their relevance to the issues to follow in this work. First of
all, discourse analysts are centrally concerned with how coherence is created in stretches of
talk. That is, they try to discover the principles which underlie the production and
recognition of coherence by determining the ways in which discourses differ from random
sequences of utterances (Stubbs, 1983, p. 15). To discover the source(s) of coherence in
discourse, some discourse analysts look to surface forms (e.g., Gumperz, 1982), the
relations of underlying abstract units (e. g. Labov and Fanshel, 1977), pragmatic principles
(e.g. Grice, 1975), the sequential organization of talk (e.g., Schegloff and Sacks, 1973),
and shared cultural knowledge (e.g., Hymes, 1974).

Another central aspect of discourse which analysts from all approaches recognize is
that “the structures, meanings, and actions of everyday spoken discourse are interactively
achieved” (Schiffrin, 1994, p.416). That is, spoken discourse is jointly produced and
sustained in real time by participants in interaction. Levinson (1983, p.44) characterizes
interaction as “the sustained production of chains of mutually—dependent acts, constructed
by two or more agents each monitoring and building on the actions of the other.” Thus,

any approach to discourse must attend to the ways in which ‘structures, meanings and

18

actions’ emerge in the process of social interaction between goal-motivated participants
playing particular roles in particular activities. Further, any approach must consider the
emergent nature of spoken discourse whereby participants respond to and build on the
contributions of each other. The meanings of utterances are discovered in part by (hearers)
referring back to prior utterances, and those interpretations of meanings allow hearers to
project forward to what will follow in the discourse. In this way, the creation of meaning

and coherence in discourse results in large part from its interactive nature.

2.3. Larger units of discourse

Discourse analysts are not only interested in local structures which create coherence
in discourse and guide interpretations (e.g. adjacency pairs (Schegloff and Sacks, 1973),
exchange structures (Stubbs, 1983)). Some approaches to discourse have recognized the
importance of investigating even larger structural units which play a role in creating
coherence and meaning (e.g., lectures, debates, sermons, jokes, chats, news reports,
interviews, narratives, etc.). However these larger units are labeled by analysts (e.g.,
‘speech events’ (Hymes, 1972), ‘activities’ (Levinson, 1992), ‘genres’ (Hanks, 1990),
‘discourse units’ (Linde, 1981, Schiffrin, 1984), ‘global structures’ (van Dijk, 1987)), it is
clear that such higher level units have their own particular organization and internal
structure; “... one distinguishing feature of discourse units is that they have an internal
structure — their constituent parts are related in predictable ways — that differentiates them
from the surrounding discourse” (Schiffrin, 1984, p.211). Furthermore, the particular

organization of larger discourse units serves as a resource for participants in producing and

19
interpreting utterances (Levinson, 1992). In the following sections, three perspectives on
such larger units will be presented. First, Labov’s work on narrative will be presented as a
particular example of a structuralist analysis of larger units. Second, Hymes’ functionalist
approach within the ethnography of communication will be presented with regard to the
notion of ‘speech event.’ Finally, Levinson’s ‘activity types’ will be discussed. It will be
argued that Levinson’s notion of ‘activities’ provides the most useful analytic tool for
distinguishing and classifying larger units of discourse since it incorporates both

structuralist and functionalist points of view.

2.3.1. Narrative structure (Labov, 1972)

Labov’s (1972) inﬂuential work on oral narrative is structuralist in that it attempts to
identify the external boundaries of an autonomous larger unit of talk, as well as the
relationships between smaller embedded units within it. Labov deﬁnes narrative as “...one
method of recapitulating past experience by matching a verbal sequence of clauses to the
sequence of events which (it is inferred) actually occurred” (p.360). According to Labov,
fully-formed narratives are composed of different smaller sections or units which perform
different functions and which are differentiated from other sections in part linguistically
(see section 2.4.2.). Narratives are normally bounded at the beginning by an abstract ,
which is ordinarily a clause that summarizes for a listener what the narrative will be about.
The abstract is usually followed by an orientation section which provides background

information about the story (who was where at what time, etc.). The skeleton of the

2O

narrative is delivered in the complicating action, which describes on a clause by clause
basis the actual sequence of events which occurred. Labov asserts that a minimal narrative
contains a sequence of at least two temporally ordered clauses such that reversing the order
of the clauses would result in changing the order of the events described. The result or
resolution marks the termination of the complicating action. All narratives also contain
evaluation, which comments on the story world from outside and tells the listener of the
signiﬁcance of the story (why is was worth telling). Evaluative clauses may also occur
outside of an evaluation section and be intertwined with other sections of the narrative.
Finally, narrative is sometimes bounded at the end by a coda which brings the speaker and
audience out of the story world back to the present and indicates that any action which may
have followed is not relevant to the story.

To illustrate these smaller subparts of personal experience narratives, consider the

following excerpt from the data of this work in which K. tells about an experience she had

atwork:1

(1)

l K: I think that going back to what L. was saying
2 abou:t like discrimig- discrimination against

3 like Blacks or Whites, I think — that - on this

4 campuzs - Blacks are more racist against Whites.
5 And I have an instance that I can talk about

6 where - where I work we have to check I.D.'s for
7 you to get into the building, and - whether

8 you're White, Black, Mexican, whatEVer, if you

9 don't have an I.D. you can't get in. And there

10 was this one time when I was checking an I.D. ,
11 and it was - a Black guy. And he didn't HAVE it.
12 Well I told him to go home and GET it. And he
13 started YELLing at me, calling me a racist bitch
14 in front of everyone. And - I mean just before I
15 didn't allow HIM in, I didn't allow a WHITE guy

21

16 in for not having his I.D. And I mean I was

17 really embarrassed - that he was like going off
18 in front of like TWEnty people saying "YOU are a
19 bitch. YOU are a racist bitch. I can't believe
20 you.” And he was going on and on about it.

First it should be noted that K. employs her story as support for her position (in
lines 34) that “on this carnpu:s - Blacks are more racist against Whites.” Personal
experience narratives are commonly used as evidence in arguments in prejudiced talk about
minority ethnic groups, especially in high-contact areas (van Dijk, 1987). Van Dijk argues
that such narratives ordinarily portray the storyteller as a powerless victim and the minority

group member as a deviant or problematic villain.2 Indeed, in K.’s narrative, she portrays

herself as an innocent victim and portrays the Black student as a problematic oppressor.

In any event, K.’s personal experience narrative displays many of the subparts
which are identiﬁed by Labov. First, in lines 5-6, K. presents an abstract (“And I have an
instance that I can talk about where -”) which appears to be self-interrupted after “where.”
At this point, K launches into the orientation section of the narrative which continues

through line 9:

(2)

where I work we have to check I.D.’s for you to get into the building, and -
whether you’re White, Black, Mexican, whatEVer, if you don’t have an I.D. you
can’t get in.

22

Here the orientation establishes certain features of the situation which later prove
relevant to K.’s presentation of herself as a victim —- everyone who comes into the
building where she works must present a student I.D. At the end of line 9, K. begins the
complicating action with “and there was this one time...” The complicating action then

lasts until line 14, and the narrative clauses contained within it are listed as follows:

(3)

a. I was checking an I.D., and it was - a Black guy.

b. And he didn’t HAVE it.

c. Well I told him to go home and GET it.

d. And he started YELLing at me, calling me a racist bitch in front of everyone.

Following the complicating action, K. then begins the evaluation section of the narrative in

line 14:

(4)

And - I mean just before I didn’t allow HIM in, I didn’t allow a WHITE guy in for
not having his I.D.

Here K. attempts to prove her own innocence (and as a consequence the Black student’s
guilt) by demonstrating her impartiality. This is buttressed by the job description which
she strategically placed in the orientation section. In lines 16-20, the evaluation section
continues, and here K. emphasizes how she suffered as a result of the verbal assault of the

Black student:

23

(5)
And I mean I was really embarrassed - that he was like going off in front of

like TWEnty people saying “YOU are a bitch. YOU are a racist bitch. I

can’t believe you.” And he was going on and on about it.
K. emphasizes the harm of the verbal assault of the Black student by admitting her
embarrassment at being attacked in front of a large crowd. In her evaluation, she portrays
herself as a powerless victim at the mercy of a problematic villain.

The previous excerpt contained many of the sections which are claimed by Labov to

occur in personal experience narratives. According to Labov, each section of the narrative

performs a different function in the sense that different underlying questions are answered:

a. Abstract: what was this about?

b. Orientation: who, when, what, where?

c. Complicating action: then what happened?
d. Evaluation: so what?

e. Result: what ﬁnally happened?

The coda does not answer any speciﬁc question, but functions to close off the narrative and
put off any further questions (such as ‘is that it?’). Not only do the sections perform
different functions within the narrative, but utterances in part gain their interpretations
depending on the section in which they occur. For example (taken from Schiffrin, 1994,
p.285), if the utterance “she was only eight years old” occurred in the orientation section, it
would be interpreted as background information. On the other hand, if the same utterance

occurred in the evaluation section (often mingled with the complicating action), it would be

24

interpreted as the storyteller’s evaluation.

Labov’s treatment of narrative serves to illustrate that larger units in discourse may
be analyzed as larger bounded structural units containing smaller interrelated structural
parts. Similar work shows that other larger units are amenable to the same kind of
structuralist analysis (see for example, Schiffrin’s work on lists, (1994), and Linde and
Labov’s work on apartment descriptions (1975)). Furthermore, the structure of such
textual units may be studied independently from how they are situated in the larger social,
interactive and discoursal context. One reason which contributes to this is that narratives

are relatively monologic, and therefore mainly the product of one person, unlike dialogic

discourse which is comprised of the interleaved contributions of at least two people.3

However, it is clearly the case that even dialogic discourse may be bounded and have
internal structure. For example, interviews are dialogic and clearly have beginnings,
middles and endings, as well as related internal structural parts (e.g., introductions,

question-answer pairs, narratives, etc.).

2.3.2. Ethnography of communication — speech events

It seems to be the case that the structural analysis of monologic or dialogic
discourse ultimately requires reference to function. Even Labov’s analysis of narrative
deﬁnes the sections (abstract, orientation, etc.) in terms of their function within the larger
unit. A fully functional approach to such larger units is found in the ethnography of

communication, which was developed by Hymes in the 1960s and 1970s, and is based on

25

insights from both anthropology and linguistics. Although a full review of this approach is
beyond the scope of this paper, we can say that one of its central assumptions is that
communication (language use) is part of our cultural repertoire which allows us to make
sense of and interact with the world, and therefore the study of language use is the study of
culture. Cultural norms are created, sustained, and negotiated in concrete kinds of verbal
interaction. In addition, such norms themselves are constrained by the particular patterns
of language use within a culture. Such patterns and their appropriate use in situations
compose what Hymes calls “communicative competence,” which is the knowledge that
governs the use of language. The particular patterns of language use may be discovered by
the analyst by way of “participant observation,” that is, a method of learning what is natural
and expected for members of a speech community by immersing oneself in that culture.

The ethnography of communication is a functional approach because it gives
theoretical priority to language use over structure (Schiffrin, 1994) in that the structures
themselves (i.e., within speech events) are the product of norms of language use. At the
same time these structures constrain language use and so, although speech is said to be
primary to code, there is an integral relation between structure and function.

According to Schiffrin (1994), the ethnography of communication “seeks to
discover and analyze the structures and functions of communicating that organize the use of
language in speech situations, events, and acts.” (p.185) Speech situations, events, and
acts are communicative units of varying size. Speech situations are the largest unit, are

governed by a single set of rules, and provide the setting or scene in which the speech

26

occurs. Speech events are embedded in speech situations and are governed directly by
norms of language use. The smallest unit is the speech act, which is the same as the unit
within speech act theory, and which is embedded within both speech events and speech
situations. Different speech situations, events and acts can be distinguished according to a
classiﬁcation system which Hymes call the SPEAKING grid. The SPEAKING grid is a

list of components which comprise any communicative situation:

S setting physical circumstances
scene subjective deﬁnition of an occasion

P participants speaker/sender/addressor
hearer/receiver/audience/addressee

E ends purposes and goals

A act sequence message form and content

K key tone, manner

I instrumentalities channel (verbal, nonverbal, physical)
forms of speech drawn from community repertoire

N norms of interaction and interpretation
speciﬁc proprieties attached to speaking
interpretation of norms within cultural belief system

G genre textual categories

The SPEAKING grid may be employed to discover, analyze and compare larger units of
discourse (e. g., a speech event). In this way, any larger unit may be dissected into its

component parts (i.e., setting, participants, ends, etc.).

2.3.3. Speech activities.
Levinson (1992) proposes the notion of “activity type,” which he claims is roughly
equivalent to, but preferable to Hymes’ speech event. His notion of “activity” is derived

from Wittgenstein’s latter work (1958) in which meaning is said to be inseparably tied to

27

the “language games” in which people engage. That is, the understanding of an utterance is
heavily dependent on the role it plays within the larger activity in which it is embedded.

Levinson introduces the concept of “activity type” as follows:

...I take the notion of an activity type to refer to a fuzzy category whose focal
members are goal—deﬁned, socially constituted, bounded, events with constraints
on participants, setting, and so on, but above all on the kinds of allowable
contributions. Paradigm examples would be teaching, a job interview, a jural
interrogation, a football game, a task in a workshop, a dinner party, and so on.

(p.69)

According to Levinson, the notion of “activity type” is to be preferred to Hymes’ notion of
speech event in part because it is not limited to an activity coextensive with speech. That is,
there may be nonverbal activities with the same status as verbal ones, and Hyme’s
approach does not capture this idea. In addition, Hymes’ approach claims that each activity
should be described according to constraints of the classiﬁcatory “speaking” grid.
However, Levinson argues that although the speaking grid is a valuable tool for comparing
different kinds of activities within and across cultures, it does not capture the fact that the
various constraints are not of equal importance. Finally, Levinson claims that Hymes’
approach is too atomistic and taxonomic in that the relation between the various constraints
of the speaking grid do not reveal the underlying force which binds them together — the
goal of the activity. This weakness in Hymes’ approach comes to light when considering
applications to problems in second language learning where pupils must be taught each and

every structural detail of an activity even though such details would naturally fall out of a

28

simple understanding of the purpose of the activity.

Levinson (1992) therefore prefers a different approach in which activities may be
identiﬁed and distinguished mainly in terms of their “structure” and “style,” although he
does not discuss the latter term. The structure of an activity includes its subdivision into
subparts, where those subparts may involve conventionalized, prestructured sequences.
Further, there may be constraints on who may participate, the roles required of such
activities, as well as on when and where the activity may take place. Most signiﬁcantly,
Levinson makes the claim that such structures are motivated by the goal of the activity in
question; “...I would like to view these structural elements as rationally and functionally
adapted to the point or goal of the activity in question.” (p.71). Thus Levinson’s claim
seems to be that the mutually recognized goal of an activity is the glue which binds the
structural subparts.

According to Levinson, one key question with regard to the structure of an activity
is; “in what ways do the structural properties of an activity constrain (especially the
functions of) the verbal contributions that can be made towards it?” (p.71). Such
constraints on the functions of contributions dictated by the structural properties of the
activity are the key to understanding the activity itself and distinguishing that activity from
others. The other side of this claim involves the ways in which the knowledge of such
constraints provides for the correct interpretation of the contributions of others by hearers;
“Because there are strict constraints on contributions to any particular activity, there are

corresponding strong expectations about the functions that any utterances at a certain point

29

can be fulﬁlling” (p.79). In other words, understanding the point of an utterance requires
an understanding of the nature of the activity one is involved in — the structure of an
activity informs interpretation of the utterances within it.

Levinson’s notion of activity types will henceforth be used in this paper when
referring to larger units of discourse. One reason for this is that this notion incorporates
both structuralist and functionalist points of view, both of which are invaluable in the
analysis of larger units. In addition, Levinson’s emphasis on activities as goal-deﬁned
events with constraints on participants and allowable contributions makes this approach an
attractive one in terms of the analysis of pseudo-argument in chapters 4-6 where it will be
shown that pseudo-argument may be established as a speech activity distinct from ordinary

argument on just these grounds.

2.4. Linguistic analysis
In this work I follow Stubbs (1983) in assuming that the analysis of surface
linguistic features (syntactic, prosodic, discourse markers, etc.) is a valuable and necessary

endeavor for discourse analysis:

...at least in some discourse types... a large proportion of the talk comprises
precisely such superﬁcial indices of underlying organization... it is precisely such
items which are of interest to the discourse analyst, since they are the items which
indicate the underlying structure of the discourse or the underlying functions of
individual utterances. (178)

 

Although underlying structures and functions are the main elements which provide

30

coherence to discourse, surface linguistic features are utilized by interactants to point to
such structures and functions, and therefore such features also contribute to coherence. In
addition to their use in helping analysts to identify underlying structures and functions,
linguistic features are (more importantly) used and oriented to by participants themselves in
the emergent, interactive construction of discourse. Therefore, linguistic features are clues
given by speakers and interpreted by hearers which make plain underlying discourse
structures and functions, and which are available for analysis by discourse analysts.

With these aims and assumptions in mind, analysts have studied various linguistic
features with regard to their role in signaling discourse structures and functions. For
example, Labov and Fanshel (1977) studied mitigation, Tannen (1989) analyzed the roles
of repetition in discourse, and Pomerantz (1984) studied the shape of dispreferred turns in
conversation. In what follows, I will discuss several other works which also are
concerned with the relationship between surface linguistic features and underlying
structures and functions. These are Schiffrin’s (1987) work on discourse markers,
Labov’s (1972) work on narrative, and Gumperz’s (1982) work on contextualization cues.
It should be noted ﬁrst that all of these studies represent different approaches to discourse
(i.e., conversation analysis, variation analysis, interactional sociolinguistics) which rely on
naturally occurring data. It is not surprising then that those approaches to discourse which
rely instead on fabricated data (i.e. speech act theory, Gricean pragmatics) or none at all
(i.e., ethnography of communication) are not very concerned with surface linguistic

phenomena. This may be because fabricated data are normally stripped of the messy (but

31

most interesting) features of spoken interaction (e. g., delays, repetitions, back-channel

cues, repairs, interruptions, rrritigations, etc.)

2.4.1. Schiffrin (1987)

Schiffrin’s (1987) work illustrates the utility of investigating certain surface
linguistic forms called discourse markers with regard to their role in bracketing units of
discourse, and thereby guiding interpretations of utterances. The English markers which
Schiffrin discusses include “well,” “now,” “so,” “but,” “oh,” “because,” “or,” “I mean,”
“and,” “y’know,” and “then.” According to Schiffrin (1987, p.326), discourse markers
are “contextual coordinates for utterances: they index an utterance to the local contexts in
which utterances are produced and in which they are to be interpre .” “Local contexts”
here refers to an utterance’s place within components or ‘planes’ of discourse which
cooccur and emerge in spoken interaction. That is, utterances occur simultaneously on
different discourse planes (according to Schiffrin, these are: exchange structure, action
structure, ideational structure, participation framework, and information state (see pp.24-29
for discussion», and discourse markers serve to display the role of utterances within those
discourse planes.

For example, the marker “oh” primarily marks information state transitions where a
bearer displays recognition of familiar information, or else the receipt of new information.
In addition, “oh” also plays a role in participation frameworks since it displays its user as a
bearer who is “...an active recipient of information who acknowledges and integrates

information as it is provided” (p.99). Finally, “oh” may play a role on the ‘action

32

structure’ plane of discourse by marking an utterance as an action, e.g., a clariﬁcation,
which helps manage information state transitions. Consider the following excerpt from the

data which illustrates a typical use of the discourse marker “oh”:

(6)

1 B: Sounds like your family is kind of against Black
2 people.

3 C: No. no not at all.

4 B: Not agAINST, but have had some bad experience

5 with- or whatever.

6 C: N— no, my m- see my mom’s like the kind of

7 person th- you know she went to like Catholic

8 highschool (and that) and she was just brought up

9 to believe that everybody's equal and that sort of
10 thing. And my dad pretty much feels the same way.
11 he went to Catholic private highschool and all this
12 and that - So my parents aren't like that at all.
13 But it's just like MY experience - it's like - =

I
14 —>B: oh that’s good

In this excerpt, C. denies B.’s assertion (in lines 1-2) that his parents are prejudiced
against Black people. After B.’s attempt to reforrnulate her assertion in a more acceptable
way, C. attempts in lines 6-12 to persuade B. that she is wrong — since his parents were
raised to believe that “everyone is equal,” and since they both went to a Catholic
highschool, they are not prejudiced people. In line 14, B. orients to this new information
about C.’s parents, and marks her new understanding with the discourse marker “oh,”
which is followed by an evaluative tag. Thus B. uses “oh” to mark what Schiffrin calls an

“information state transition” and displays the receipt of new information which has

33

presumably altered her view of C.’s parents.

As a second example of how discourse markers index utterances to local contexts,
consider the marker “but.” This marker’s primary function is to locate an utterance on the
ideational plane of discourse (for Schiffrin, this means the semantic level of propositions or
‘ideas’) by marking an upcoming unit as a contrasting idea or action (p.152). In such cases
the utterance prefaced by “but” may contrast with either the actual semantic content of a
prior utterance, or else with implicit expectations or assumptions by the speaker or hearer.

The following excerpt from the data illustrates the use of “but” as a contrast marker:

(7)
1 G: Um, I actually don’t know the aesthetics of
2 it. I don’t know how much - how many: minority
3 scholarships there are out there:, - an- and what
4 have you:, - I DON'T know quite how much - leeway
5 they’re given as to getting INto the university, -
6 um but I do think they need to have some kind of an
7 advantage coming from the inner-city.= It's — =

l
8 —>A: But =
9 G: = got to be tough.

10 A: = why are you assuming that all Black people are
11 coming from the inner-CIty: .

In this excerpt, A. uses “but” in line 8 to mark the utterance it precedes as a
contrasting unit. In this case, “but” signals contrast with G’s assertion in lines 6-7,
speciﬁcally with its presumed implicit assumption that all Black students are from

the inner city.

34

“But” may also function at the exchange structure plane in that it may be used to
mark speaker-return to a point which has been interrupted or postponed due to competing
conversational demands. Consider the following example from the data which involves the

use of “but” in its resumptive capacity:

(8)
1 G: [[Yeah. The minority aids?= They have them right
2 next to- right- They have the residential advisor,
3 AND they have the minority ad- aids and advisors or
4 whatever, - - I can understand, It's GOT to be =

[
5 A: Uhhuh,
6 G: = harder to adjust for them, in some ways. SOME
7 of them. I'm not saying ALL Black- I am
8 generalizing right now, and it probably comes off
9 as pretty - rude and obnoxious.

10 T: It's because you're a racist pig!
11 ((T and A laugh loudly))

12 —>G: But I do think that - chances are - they— =

[
((telephone rings))

13 G: = they- they run into more problems, and need
14 someone to TALK to a lot of the times I think.

In this excerpt, G. is interrupted by T.’s tease in line 10, followed by T. and A.’s laughter.
It seems plausible here to suppose that G.’s use of the marker “but” in line 12 signals a
return to his point which has been interrupted by the other participants. In this case, “but”
does not appear to be used to signal contrast.

Schiffrin asserts that some markers have semantic content whereas other do not.

35

She argues that the meanings of those markers with some semantic content (e. g., “but,”

99 u

“because,” “so, now”) contribute to their discourse function, but:

...markers themselves do not convey social and/or expressive meanings. Rather,
markers are situated in very different discourse slots, and it is the utterance within
that discourse slot which is interpreted for social and/or expressive meaning: but ,
for example, does not itself mean ‘challenge’ — although the utterance which it
precedes may certainly be interpreted as a challenge. (p. 318)

In this way, a marker itself does not convey meaning, but rather contributes to the
interpretation of utterance relations. In Schiffrin’s terms, markers ‘select’ and ‘display’
meaning relations between sequences of utterances. To demonstrate what these terms

mean, consider the following simple example provided by Schiffrin, (1987, p.318):

a. Sue dislikes all linguists
b. I like her

The meaning relation between these two utterances is on the surface indeterminate. That is,
one might select a contrastive relation between (b) and (a) (which could be marked by “but”
before (b)), or alternatively a resultative relationship (which could be marked by “so”
before (b)). Schiffrin argues that the interpretation of the meaning relation between these
utterances in context would already be constrained by the background conditions. That is,
the correct interpretation would be available based on the hearer’s knowledge of the
speaker’s beliefs (e. g., that the speaker has a high opinion of linguists, or alternatively a

very low one). Therefore, the addition of markers like “but” or “so” before (b) in this case

36

would not determine the meaning relation between (a) and (b), but rather it would display a
relation which was already inferable from the context; “This means that although a marker
may be able theoretically to select any number of implicit and potential relationships, in
actuality, that relationship is already fairly constrained, such that the marker acts more to
display the relationship” (Schiffrin, 1987, p.319).

In addition to displaying meaning relations, Schiffrin suggests that markers may
also serve to display structural relations between utterances. That is, markers may display
the identity of structural units in talk (which may be the size of single utterances, or may be

the size of larger units). Consider the following example given by Schiffrin:

a. I believe in fate.
b. I won the grand prize in a sweepstakes.

There are two possible interpretations of the meaning relations between (a) and (b) — (a)
may be the cause of (b) or else (b) may be support for (a). There are also accompanying
structural relations which are involved in that either utterance (a) could be the structural unit
of a ‘position’ with (b) as a unit of ‘support’, or alternatively (a) could be a unit of ‘cause’
with (b) as a unit of ‘result.’ Either of these interpretations are possible, but when markers
like “because” or “so” are added, one interpretation of the meaning and structural relations
between the utterances is selected and displayed to the exclusion of the other. Either of
these interpretations was available without markers, and although one interpretation would

be more likely in actual context, the addition of markers makes clearer the relations

37

between utterances.

In sum, discourse markers bracket units in discourse by selecting and displaying
the meaning and structural relations between utterances. For Schiffrin, markers locate
utterances on different discourse planes and serve as useful linguistic devices in the display
and interpretation of discoursal organization and structure. Schiffrin also notes that
discourse markers promote coherence in that they integrate different discourse planes by
locating utterances at different planes simultaneously; “... a discourse marker — with one
indexical function, anchors an utterance into more than one discourse component at once.
By so doing, it provides a path towards the integration of those different components into

one coherent discourse” (p.330).

2.4.2. Labov (1972).

Next consider again the work by Labov on narrative. In section 2.3.1. it was
shown that Labov’s work treated personal experience narrative as a bounded and structured
unit, with interrelated internal units (orientation, complicating action, etc.). In addition to
this structural emphasis, Labov was also interested in the acquisition of a ‘syntax’ of
narrative. That is, speakers as storytellers acquire the skills which allow them to modify
the syntax of constructions in ways which differentiate the various underlying smaller
structural units. Individuals respond to such correlations between surface form and
underlying discourse structure, and therefore surface syntax serves as a guide to hearers

about the meaning and function of utterances within a narrative.

38

For example, Labov argues that clauses which contain the complicating action
(narrative clauses) are simple and regular. He describes the basic narrative clause as a

series of eight possible elements. These include (in order) conjunctions (e. g., “so ” “and,”

“but,” “then”), simple subjects, simple past tense, no auxiliary at surface structure, preterit
verbs, complements (direct and indirect objects), manner or instrumental adverbials,
locative adverbials, and temporal adverbials and comitative clauses.

For instance, in the following excerpt from the data, the narrative clauses in lines 6-

7 and 9 contain many of these elements:

(9)

A: Okay then, tell me if THIS is reverse
discrimination. My freshman year. , my brother floor
was - very - African-American, there were only a
couple White guys on the floor, and I became good
friends with my R.A., and some of his friends, and.
- one day I went to sit with them at the TABle, -
the Black table in Wilson. He was Black, and I =

[
T:Was your R.A. Black or White.

A: = went to SIT there, and people got up and left
because they didn't want a White girl sitting
there. - What's that.

I—‘I—‘KO co \IOSUIDWNH

I-‘O

In this excerpt, A. provides a personal experience narrative which is meant to support her
contention that reverse discrimination exists on the campus. After some orientation in lines
2-5, A. begins the complicating action in line 6. The narrative clauses of the complicating
action (which are interrupted by more orientation in line 7 in response to T.’s question in

line 8) are as follows:

39

(10)

a. one day I went to sit with them at the TABLE, - the Black table in Wilson.

b. and people got up

c. and left
These clauses contain many of the 8 possible elements Labov attributes to basic narrative
clauses, including the conjunction “and,” simple subjects (“1” and “people”), simple past
tense, no auxiliaries at surface structure, the comitative “with them” and the locative
adverbial “at the table.”

Compared to the syntactic complexity of ordinary conversation, Labov argues,
narrative clauses are extremely simple in that they usually lack such elements as modals,
negatives, and embedding. Furthermore, Labov argues that given the lack of complexity of
narrative clauses, deviation from that basic syntax has a marked effect, and serves as a
linguistic means to signal evaluation. Recall that although there may be an independent
evaluation section, it is often the case that evaluation will pervade other sections of a
narrative. By departing from the basic syntax of the narrative clause, a speaker can embed
evaluative force within narrative clauses.

To illustrate that there is an association between syntactic complexity and
evaluation, Labov (p.392) discusses various syntactic constructions (of varying
complexity) which are not part of the basic narrative syntax and which carry evaluative
force. These include intensiﬁers (quantiﬁers, repetition, and ritual utterances, etc.),
comparators (constructions containing negatives, modals, questions, imperatives, etc.,

which compare the actual events with other possible ones), and correlatives (such as

40
progressives in be...ing , or appended participles which conjoin within an independent
clause two events).

One reason for analyzing the syntax of clauses within narratives is to show that
different utterances have speciﬁc syntactic and semantic properties which contribute to their
function and their identity as units within the narrative (Schiffrin, 1994, p.284). In other
words, the surface linguistic constructions are suited to different structural parts and
functions. The narrative clauses of the complicating action section, for example, are
identiﬁed by their very basic syntactic features. Evaluation (within the complicating action)
is identiﬁed by the marked departure from basic narrative syntax, and orientation clauses
are often identiﬁed in part by the progressive tense. In sum, this kind of structural
linguistic analysis at the discourse level serves to illustrate that both underlying structure
and surface linguistic features contribute to the organization and coherence of larger units in
talk. Furthermore, Labov’s work shows that linguistic features are important clues for the

analyst in discovering underlying structures and functions in a text.

2.4.3. Gumperz (1982).

Whereas Labov’s study of narrative is structuralist in that it seeks to discover
underlying units of organization and their interrelations, the work of Gumperz falls
squarely into the functionalist camp in that it is concerned, in Schiffrin’s (1994) terms, with
“...the social, cultural, and expressive meanings stemming from how utterances are
situated in contexts” (p.350). Where Labov’s work was concerned with the structure

evidenced within a text, Gumperz’ (1982) work is concerned with the situated interpretive

41

procedures employed by interactants:

This book attempts to deal with such issues by concentrating on the participants’
ongoing process of interpretation in conversation and on what it is that enables them
to perceive and interpret particular constellations of cues in reacting to others and
pursuing their communicative ends. (p.5)

The ‘what it is’ in this case refers not just to grammatical competence, but also to such
extralinguistic factors as background knowledge of context and culture-speciﬁc norms,
interactive goals, and interpersonal relations which contribute to creating ‘frarnes’ for
interpretation:
in determining what is meant at any one point in a conversation, we rely on
schemata or interpretive frames based on our experience with similar situations as

well as on grammatical and lexical knowledge... such frames enable us to
distinguish among permissible interpretive options. (p.21)

Interpretative frames allow hearers to select among many possible interpretations by
locating an utterance within a particular activity type, which according to Gumperz is a
“basic socially signiﬁcant unit of interaction” (p.130). That is, frames (activity types) tell a
bearer what the talk is about and “... constrain interpretations by channeling inferences so
as to foreground or make relevant certain aspects of background knowledge and to
underplay others” (p.131).

A central feature of Gumperz’ work is the ‘contextualization cue,’ which is a

(linguistic) signaling mechanism for identifying interpretive frames:

42

...constellations of surface features of message form are the means by which
speakers signal and listeners interpret what the activity is, how semantic content is
to be understood and how each sentence relates to what precedes or follows.
These features are referred to as contextualization cues. (p.131)

Contextualization cues are implicit and do not carry meaning out of context, but rather serve
to signal ‘contextual presuppositions.’ The linguistic realizations of contextualization cues
are diverse; code switching, prosodic phenomena, choice of lexical or syntactic options,
conversational openings and closings, etc. Such cues are implicit because they do not have
conventional meaning and interactants are rarely aware of them. Rather, in Schiffrin’s
terms (1994, p. 100), “...they signal the speaker’s implicit deﬁnition of the situation, and
more importantly, how the propositional content of talk is to be understood.”

Since contextualization cues are learned in face-to face interaction, they‘may be
group or culture-speciﬁc, and therefore interaction between individuals who do not share
the same contextualization cues may lead to misunderstanding. One method therefore of
discovering contextualization cues is to study interactions across groups or cultures where
communication fails. Gumperz points out that such rrriscommunication in interaction can
have disastrous consequences for members of disenfranchised groups who do not share the
same contextualization cues as dominant groups (e.g., in gatekeeping situations).

The following example taken from Gumperz (1982, p. 147) illustrates how the lack

of shared (prosodic) contextualization cues may lead to miscommunication:

43

(1 1) [don’t wanna read
In a taped elementary school classroom session, the teacher told a student to read.
The student responded, “I don’t wanna read.” The teacher got annoyed and said,
“All right, then, sit down.”

Gumperz played the taped exchange for both Black and White respondents. Most of the
White respondents interpreted the student’s utterance “I don’t wanna read” as being
uncooperative. Most of the Black respondents on the other hand interpreted the utterance as
meaning “I can do it, but I need you to encourage me a little ﬁrst.” The teacher obviously
chose the former ‘uncooperative’ interpretation and responded accordingly. The Black
respondents agreed that the rising intonation at the end of the student’s utterance was what
led them to their ‘encourage me’ interpretation, and that if the student had intended to be
uncooperative he would have stressed ‘wanna’ instead. In addition, Gumperz notes
(p.149) that ethnographic evidence from urban schools indicates that formulaic expressions
like “I don’t wanna” and “I don’t know” in addition to the speciﬁc prosodic contour
mentioned above have formulaic interpretations as requests for cooperation or support
which may be speciﬁc to the Black speech community. Further, those not familiar with
such cues use their own system of cues and interpret such utterances as expressions of
inability or uncooperativeness. Obviously, such cases of miscommunication may lead to
unwarranted negative evaluations by teachers.

The attention to surface linguistic forms (as contextualizations cues) is an important

aspect of Gumperz’ approach because such forms are considered an essential part of the

44

interpretive process in discourse. As Gumperz puts it, “The issue as we have deﬁned it,
concerns the discovery of hitherto unstudied connections between perception of surface
linguistic signs and interpretation” (p.32). Discourse is coherent and meaningful for
interactants in part because they share contextualization cues which allow them to share
interpretations and agree on the kinds of talk that they are engaged in.

In sum, it is assumed in this paper that the analysis of surface linguistic features is a
valuable and necessary pursuit for discourse analysis. Such features (syntactic, prosodic,
lexical, etc.) are resources for both participants and analysts in the creation and
interpretation of meaningful, coherent discourse. Chapter 5 will be devoted to the analysis

of some of the speciﬁc linguistic properties of pseudo-argument.

2.5. Two approaches to discourse.

The following two sections will present two prominent approaches to discourse
which will be come into play in later chapters.
2.5.1. Speech act theory.

Speech act theory may aptly be called a general theory of language usage
(Levinson, 1983), and although the philosophers who originated and advanced the theory
were not directly concerned with discoursal phenomena, it is fair to say that speech act
theory is one approach which provides the theoretical units which may be viewed as the
building blocks of discourse. These units are social actions like commanding, requesting,

promising, and questioning which may be identiﬁed and labeled, and since they each

 

 

45

respond to and initiate other social actions, they provide “...a local, sequentially emergent

basis for discourse” (Schiffrin, 1994, p.91).4

Speech act theory was conceived by the ordinary language philosopher John Austin
in the early 1960’s as part of a larger reaction against the logical positivist view in
philosophy that sentences which cannot be given a truth-value are meaningless. Austin’s
work aimed to undermine the prevalent view that the central function of language is to
describe states of affairs (Potter and Wetherell, 1987, p.15). His main insight was that
language is used to perform communicative actions and not simply to describe things in the
world. Therefore speech act theory is concerned with the sorts of things that people do
with language. Austin’s work was signiﬁcant and inﬂuential in part because it directed
attention to a view of language as a human practice, dependent on social convention and
context.

Austin ﬁrst noted that certain kinds of sentences (‘performatives’) are mainly
actions which cannot be said to be true or false but rather may be defective in one way or

another. Consider the following perforrnatives:

(12)
a. I sentence you to 10 years

b. I hereby grant you immunity

In these cases, the purpose of the sentences is to get something done, not to describe some

state of affairs. Instead of making the odd claim that these sentences may be true or false, it

46

should be said that they are successful or unsuccessful in performing the actions they
propose. If for example, you haven’t committed any crime, or the judge is joking, then the
acts will be ‘defective’ (but not false). The conditions which must be met in order for an
act to be successful are called ‘felicity conditions.’

Austin called those utterances which primarily describe states of affairs
‘constatives’, but he denied that perforrnatives and constatives may be easily distinguished.
That is, all utterances both describe things and do things. Speech act theory asserts that all
utterances have both a meaning (a sense and reference) and a force, which Austin called
respectively a ‘locutionary act’ and an ‘illocutionary act.’ The illocutionary act is “...what
is directly achieved by the conventional force associated with the issuance of a certain kind
of utterance in accord with a conventional procedure, and is consequently determinate...”
(Levinson, 1983, p.237). Another type of act which an utterance performs is the
‘perlocutionary act’, which has to do with the indeterminate effects (intended or
unintended) caused by utterance. For example, the utterance ‘Jump!’ may have the
illocutionary force of an order or suggestion, but may have the perlocutionary effect of
shocking, humoring or forcing the bearer. (See van Eemeren and Grootendorst (1984) for
detailed discussion of the status of perlocutionary acts in speech act theory.)

Austin’s main ideas were later systematized by Searle (1970, 1976), and it is
Searle’s work which has had the most inﬂuence in linguistics (Levinson, 1983, p.238).
One of Searle’s main contributions is the idea that felicity conditions are constitutive of

illocutionary force, so that meeting the felicity conditions for a particular act effectively

47

counts as doing that act. If felicity conditions are constitutive of different acts, then it
becomes possible to compare different speech acts in terms of their felicity conditions.
Searle divides the felicity conditions for any speech act into four kinds: The ‘propositional
content condition’ concerns the reference and predication of the utterance. The ‘preparatory
conditions’ are the features of context which must be met for an act to be successful — for
example, knowledge of and about participants as well as of the situation. The ‘sincerity
conditions’ have to do the speaker’s psychological state at the time of utterance, and the
‘essential condition’ is what the utterance ‘counts as’, that is, the action that the utterance
performs. Levinson (1983, p. 240) provides the following example to illustrate how

Searle’s constitutive rules may be used as a kind of grid to compare different speech acts:

A comparison of felicity conditions on requests and warnings

Conditions REQUESTS WARNINGS
propositional content Future act A of H Future event E
preparatory 1. S believes H can do A 1. S thinks B will occur and is

not in H’s interest
2. It is not obvious that

H would do A without 2. S thinks it is not obvious to H
being asked that B will occur
sincerity S wants H to do A S believes E is not in H’s best
interest
essential Counts as an attempt to Counts as an undertaking that E
get H to do A is not in H’s best interest

For Searle, part of linguistic competence is having knowledge of the constitutive rules for

speech acts. Schiffrin (1994, p.59) summarizes the import of constitutive rules as follows:

48

Communication relies upon shared knowledge of the name and type of a speech act:
speaker and hearer share knowledge of how to identify and classify an utterance as
a particular “type” of act, as a unit of language that is produced and interpreted
according to constitutive rules.

Searle was also concerned with providing a principled taxonomy of speech act
types, and because certain speech acts are related in that they share particular constitutive
rules, it was thought that related speech acts could be classiﬁed under several larger
categories. Searle (1976) distinguishes ﬁve major categories (cited in Levinson, 1983,

p.240):

(i) representatives, which commit the speaker to the truth of the expressed
proposition (paradigm cases: asserting, concluding)

(ii) directives, which are attempts by the speaker to get the addressee to do
something (paradigm cases: requesting, questioning)

(iii) commissives, which commit the speaker to some future course of action
(paradigm cases: promising, threatening, offering)

(iv) expressives, which express a psychological state (paradigm cases: thanking,
apologizing, welcoming, congratulating)

(v) declarations, which effect immediate changes in the institutional state of
affairs and which tend to rely on elaborate extra-linguistic institutions
(paradigm cases: excommunicating, declaring war, christening, ﬁring from

employment)

To summarize this brief, simpliﬁed sketch, speech act theory as originated by Austin and

schematized by Searle takes the illocutionary act to be the basic unit of communication.

49

Language is above all a tool for social action, and language users share the constitutive
rules which make the production and interpretation of speech acts possible. Although
speech acts are deﬁned according to psychological and social functions independent of
discourse (Stubbs, 1983, p. 149), speech act theory may serve as an approach to discourse
because it supplies the illocutionary act, and it is proposed that discourse may be segmented
entirely into such communicative units. Furthermore, discourse is coherent because speech
acts initiate and are responsive to other speech acts, and therefore form the basis for the

sequential organization of talk.

2.5.2. Conversation analysis

Conversation analysis is an approach to discourse which has its roots in sociology
and the framework of ethnomethodology (Garﬁnkel, 1967). Garﬁnkel proposed that
“...shared methods of reasoning generate continuously updated implicit understandings of
what is happening in social contexts...” (Heritage, 1988, p. 128). These methods of
reasoning are tacit, normative, and allow people to create a common, mutually intelligible
social world. As a subdiscipline of ethnomethodology, conversation analysis (CA) uses
conversational interaction as a resource in order to investigate such normative methods of
reasoning, because in conversation people display their analysis and understanding of what
is happening in the interaction as it happens, on a turn by turn basis. Thus, according to
Heritage (1988, p. 129), “...CA is centrally concerned with the study of the sequential

organization of interaction and of the reasoning that is inherently displayed in it.”

 

50

Sequential organization means that each turn in talk occurs in response to an
immediately prior turn and further provides a new context for a next turn. In Schiffrin’s
terms; “This notion of context as being both retrospective and prospective can be seen as
yet another way that meanings (and knowledge) are continually adjusted and sequentially
emergent” (1994, p.235). Conversation analysis is a structural approach which seeks to
discover the patterns and kinds of organization that underlie and result from this
sequentially emergent nature of talk. The basis for discovering such organization is the

conduct of participants themselves:

it is because actors succeed in using the sequential progression of interaction to
display their understandings of its events and rules that the shared world that has

been jointly achieved is publicly available for analysis. (Taylor and Cameron,
1987 :104, cited in Schiffrin, 1994, p.234.)

Therefore, conversation analysts (unlike speech act theorists who normally focus on
decontextualized single utterances) study empirical, naturally occurring conversations.
Conversation analysts avoid premature generalization by proposing underlying structures
for which there is evidence within the talk itself that participants are methodically
producing and orienting to those structures “as normative organizations of action”
(Heritage, 1988, p.131). The notion of ‘normative,’ as used by conversation analysts,
means that certain actions are expected and that their absence may lead to sanction. The
methods of CA therefore attempt to discover the underlying structures in a text by

examining evidence for the ways in which participants produce and orient to those

51

structures as well as how they orient to departures from them.

One fundamental structure which has been proposed to underlie conversation is the

‘adjacency pair’ :5

adjacency pairs are sequences of two utterances that are:

(i) adjacent

(ii) produced by different speakers

(iii) ordered as a first part and a second part

(iv) typed, so that a particular ﬁrst part requires a particular second (or range of
second parts) — e. g. offers require acceptances or rejections, greetings

require greetings, and so on (Schegloff and Sacks, 1973, p.295)

Such paired utterances are part of the local, normative organization of action in that the
absence of a second pair part will usually incur either a repetition of the ﬁrst part, or some
sanction by ﬁrst speaker. In addition, the expectation of a second part is so strong that its
absence will often by read as having a particular meaning. For example, if speaker A asks
speaker B “Do you want to go visit your mother next week?”, silence on the part of speaker
B could well be interpreted as a negative response. That is, participants in conversation
expect answers to follow questions. The ﬁrst part of an adjacency pair sets up a next slot
which must be ﬁlled by a second speaker, and creates a ‘conditional relevance’ such that
whatever follows that ﬁrst part will be interpreted as the particular expected second (e.g. an

answer to a question):

52

...the fact that answers to questions normally and properly follow the question to
which they respond is a facet of the social organization of action, but it is also a
resource in the social organization of interpretation — a resource which supplies the
presumptive basis on which to interpret utterances which follow questions as
‘answers’ (Heritage, 1988, p.139).

To say that particular ﬁrst parts require a particular second is somewhat of an
oversimpliﬁcation in that there may be a range of possible second parts to some ﬁrst part.
However, it is not the case that all possible second parts are equal. That is, different
possible second parts to some ﬁrst part are ranked in relation to one another — there is
ordinarily at least one ‘preferred’ second and at least one ‘dispreferred’ second to any given
first pair part. For example, in response to an offer the preferred second is an acceptance
whereas the dispreferred second might be a rejection. The notion of preference is a
structural one and not a psychological one (Levinson, 1983, p.307) in that preferred
seconds are structurally simple (unmarked), whereas dispreferred seconds are structurally
more complex (marked). For instance, dispreferred seconds (e.g. refusals to requests,
disagreements to assessments) are normally delayed, prefaced by some marker (such as
‘well’), and accompanied by some account (e.g. “I have plans already for the weekend”).
Participants orient to such structural marking and are responsive to it. For example, if a
speaker makes an invitation which is followed simply by a slight delay or ‘well,’ that
speaker might quickly withdraw the invitation in anticipation of a dispreferred response.

In this way, preferred and dispreferred second parts are distinguished primarily on
structural grounds. Levinson (1983, p.333) suggests that in addition to this structural

criterion, we also need a rule of speech production which says essentially “Avoid the

53

dispreferred action.” Pomerantz (1984) seems to distinguish preferred from dispreferred
actions in psychological terms: “A next action that is oriented to as invited will be called

‘preferred next action’; its alternative, 3 ‘dispreferred next action’” (p.63). For Pomerantz,

preferred actions are agreeable whereas dispreferred actions are disagreeable:6

...across different situations, conversants orient to agreeing with one another as
comfortable, supportive, reinforcing, perhaps as being sociable and as showing that
they are like-minded... Likewise, across a variety of situations conversants orient
to their disagreeing with one another as uncomfortable, unpleasant, difﬁcult,
risking threat, insult or offense. (p.77)

To return now to the notion of the adjacency pair, it is not always the case that

second pair parts immediately follow ﬁrst pair parts. Consider the following invented

example:
(13)
A: Do you want to take a walk?
B: Where to?
A: The park.
B: Okay

In this case, A’s ﬁrst pair part (question) is followed not by an answer, but rather by
another question which is followed by an answer. B returns to the second pair part after
the inserted question-answer pair. Such embedded pairs which separate another adjacency
pair are called ‘insertion sequences’ (Schegloff, 1972). Insertion sequences allow the

postponing of a second pair part while preliminaries are sorted out. However, given the

54

conditional relevance of the original ﬁrst pair part, a second pair part is still expected even

though it does not follow immediately.

2.6. Summary.

This chapter has attempted to provide only the discourse analytic tools which will
be employed in later chapters of this paper. Chapters 4 and 6, for example, will require
some familiarity with the basic discourse analytic assumptions discussed earlier (i.e., about
interaction, structure and function, the organization of larger units) as well as with two
approaches to discourse — speech act theory and conversation analysis. Chapter 5 will
require some understanding of the motivations for and the methods of linguistic analysis of
discourse, as discussed here with regard to the works of Schiffrin (1987), Labov (1972)
and Gumperz (1982). In addition, there are other important aspects of discourse and
discourse analysis which were not discussed above, but which will be brought into play in
later chapters as needed (e.g., “participant frameworks” in chap. 5, and “identity wor ” in
chap. 6)

This chapter has not presumed to be a comprehensive review of all of the
approaches to discourse, nor has it argued for one approach over others. Moreover,
discussion of even a portion of the array of methods, assumptions and concepts in an area
as vast as discourse analysis is a difﬁcult task, and selecting tools from such an array is a
particularly signiﬁcant and delicate ﬁrst step in the discourse analysis of a text. Therefore,
in what remains in this chapter I will summarize some of the assumptions which will serve

as guidelines for the analysis to follow in later chapters. These assumptions are culled

55

from the various approaches and are shared by many, if not all of them (although there may
be differences in emphasis).

First, I follow Schiffrin (1994, p.415) in deﬁning discourse analysis as the study
of “...the language of utterances in relation to its function in social interaction.” This
deﬁnition ﬁnds the middle ground between structuralist and functionalist approaches to
discourse. Second, it is assumed that coherence (in spoken interaction) results in part from
the sequential properties of interaction and in part from surface linguistic forms in concert
with the speech activities in which they are embedded. Third, it is assumed that structure
and function are mutually constitutive in that function constrains structure and vice versa.
Forth, I assume that structure emerges in terms of the local, sequential relations of
utterances; “...each utterance in an interaction receives part of its meaning from what
another person oﬂ’ered before, and gives part of its meaning back to that other person to
use in what comes next : it is from this chain of self/other reciprocity that the structure of
interaction emerges.” (Schiffrin, 1994, p.352). In addition, structure emerges at a global
level in terms of larger units — speech activities are structured, goal-deﬁned events with
constraints on participants and allowable contributions.

Fifth, speech activities may be compared to and distinguished from other activities
in terms of their structure, function, and linguistic properties. Sixth, the organization of a
speech activity is an interpretative resource for hearers. Finally, linguistic features
(syntactic, prosodic, lexical, etc.) are also resources for participants and analysts in the

creation and interpretation of meaningful, coherent discourse.

56

The analysis of the data in the following chapters will be carried out with respect to
these assumptions and guidelines. Chapter four will involve the analysis of the data from a
structuralist perspective, seeking to discover underlying units of organization and their
interrelations. It will be argued that pseudo-argument is a speech activity distinct from
ordinary argument in terms of the constraints on the sorts of allowable contributions.
Chapter ﬁve will analyze the data with regard to recurrent linguistic features which serve to
signal underlying structures and guide interpretations. Chapter six will analyze the data
from a more functionalist perspective with regard to how the goals and strategies of the
participants bring about the particular structures of pseudo-argument. Before beginning the
analysis however, some more preliminaries are in order. Since a goal of this work is to
establish a speech activity which is distinct from ordinary argumentation, the nature of
argumentation must ﬁrst be made clear. The following chapter will review early
scholarship as well as current deﬁnitions and assumptions about the nature of
argumentation, and ﬁnally will provide two contemporary discourse-analytic models of

argumentation against which the data will be compared in chapter 4.

Chapter 3

ARGUMENTATION

3.0. Review and Background

The aim of this chapter is ﬁrst to brieﬂy review the ﬁeld of argumentation in order
to provide a background for the chapters which follow, and second to provide a
characterization of argumentation as a speech activity against which ‘pseudo-argument’
may later be compared. This latter goal will involve the search for an appropriate deﬁnition
of argument, followed by the provision of two inﬂuential discourse-analytic models of
argument (the speech act approach of van Eemeren and Grootendorst (1984), and the
conversational analytic approach of Jackson and Jacobs (1982, 1986)).

Although argumentation has been studied since antiquity (Schriffrin, 1985; Willard,
1989), it was not until the end of the 19th century that the study of argumentation became a
‘ﬁeld’ in itself (Cox and Willard, 1982). Early works on argumentation (e.g., George
Pierce Baker’s Principles of Argumentation , 1895) were textbooks which prescribed the
means for research, case-building and refutation in public speaking, modeled by famous
interscholastic debates. Underlying such works was the implicit philosophical assumption

that debate is a truth-seeking method of inquiry which has its analog in practical decision-

57

58

making processes.

Drawing on British rhetorical theory, these early writers and writers well into the
20th century also assumed a sharp distinction between ‘conviction’ and ‘persuasion,’
where conviction issued from appeals to logic and intellect, and persuasion appealed to
emotion. Successful argumentation was thought to involve both conviction and
persuasion, with persuasion being employed only secondarily in the service of logical

positions:

Normatively, argumentation was the study of the proper or “best grounds ” for
conviction — the truth conditions of propositions, the requirements for valid
reasoning, and the rational selection of belief or action from among the expressed
alternatives” (Cox and Willard, 1982, p. xvii.).

Early work considered argumentation (in debate) to be a procedure for the rational
discovery of truth, and therefore the unearthing of the logical form of arguments was
thought to give guidance in correct thought and decisions. Although argumentation was
recognizably different from formal logic, it was assumed that the process of debate, though
dealing with probabilistic facts and not analytic certainties, could result in something akin to
truth. Because it was assumed that in debate truth would naturally win out over falsehood,
early textbooks optimistically asserted that argumentation as applied logic would always
disclose fallacious reasoning and lead to truth (Cox and Willard, 1982, p. xx.).

In sum, according to Cox and Willard (1982, p.xxi), from the late 19th century

until at least the mid-20th century, argumentation “had become the study of the logical

59

forms underlying ordinary language claims, disputes, and debate.” Therefore the emphasis
of this tradition was on logical structure abstracted from the interactional setting and
purposes of actual argumentation. The aim of argument analysis was seen to be the
reduction of actual debate to its underlying logical form, with the subsequent prescription
of rules of ‘rational inference.’ However, by the mid-1960’s, the underpinnings of early
argumentation research began to be questioned. The works of Toulmin (1958), and
Perelman and Olbrechts-Tyteca (1958,1969) had a considerable inﬂuence on scholars of
argumentation, and led to a rethinking of the aims of the ﬁeld as well as a reconceptualizing
of argumentation itself. First, there was a turn in this period away from prescription
towards describing argument in use. Second, with the reassessment of the assumption of
the ‘a priori’ status of logic, argumentation scholars assailed the philosophical notion that
beliefs may be arrived at along the same lines as proofs in mathematics. Rather, one must
begin an investigation of argument empirically to examine the “contextual and/or audience-
dependent sources of a claim’s correctness, reasonableness, or acceptance.” Perelman and
Olbrechts-T‘yteca (1971, p.4) argued that since an argument develops in terms of an
audience, argumentation theory should involve “the study of the discursive techniques
allowing us to induce or to increase the mind ’s adherence to the theses presented for its
assent ” (Cited in Cox and Willard, 1982, p. xxiv).

This turn towards the social and contextual embeddedness of argument and the
dependence of argument on audience brought the study of argument back closer to

traditional rhetorical theory, since rhetoric, as the science of persuasive communication,

60

also focuses on context and audience. According to Alexandrova (1986, p.267), the
purposes of rhetoric and argumentation are identical in that both involve the same system of
subpurposes; the main purpose for both is to persuade the audience/recipient of the validity
or acceptability of an assertion, to confute counterassertions, and to prognosticate the future
behavior of the audience. The secondary purposes of both rhetoric and argumentation are
to air one’s own knowledge, values and attitudes as a member of some social group in
order to excite trust and conﬁdence in the assertions put forward. In addition, in both
rhetoric and argumentation, the speaker attempts to appear unbiased, honest and objective

in order to provoke sympathy with his/her own assertions as well as an emotional reaction

against the opponent and the opponent’s assertions.l

Scholarship on argumentation in the last few decades has attempted to conduct
programmatic research which relies on more rigorous attention to theory-building,
conceptual clariﬁcation, and methodology. While the various issues which have been
raised and the differences among argument theorists are many and far beyond the scope of
this paper (see Cox and Willard, 1982, for review), in what follows I will touch upon only
those common assumptions and aims of contemporary argument theory which are germane
to the (discourse-analytic) purposes of this paper. Speciﬁcally, I will discuss O’Keefe’s
distinction between the two senses of argument, the communicative/interactive nature of

argumentation, and the focus of contemporary argument theorists on ordinary discourse.

61

3.1. O’Keefe’s ‘two senses of argument’

In a self-conscious attempt to clarify concepts within the ﬁeld, argument theorists
have debated the meaning of the concept ‘argument.’ Following the work of O’Keefe
(1977, 1982), there is now a general consensus among argument theorists that care must

be taken to distinguish two distinct senses of argument;

In everyday talk the word “argument” is systematically used to refer to two different
phenomena. On the one hand it refers to a kind of utterance or a sort of
communicative act. This sense of the term I will call “argumentl.” It is the sense

contained in sentences such as “he made an argument.” On the other hand,
“argument” sometimes refers to a particular kind of interaction. This sense,
“argumentz,” appears in sentences such as “they had an argument.” (1977, p. 121)

While an ‘argumentl’ may be said to be refuted, valid, or fallacious, it would not
make sense to characterize an ‘argumentz’ in this way. Similarly, while an ‘argumentz’

might be said to be pointless, unproductive or to come to blows, it would be odd to

characterize an ‘argumentl’ in this way (O’Keefe, 1977, p.121). Further, the distinction is

evidenced in everyday talk in the difference between ‘arguinglthat’ and ‘arguingz about.’
According to O’Keefe, much of the earlier discussion and characterization of

argumentation had been muddled because of the failure to keep these two senses separate.

While earlier argument research was mostly interested in ‘argumentl,’ (and prescribing the
ways to become a good ‘arguer1’), O’Keefe warns that the expanded (descriptive) interest
of contemporary argument scholars in ‘argumentz’ requires that the two senses not be

elided:

62

Yet I think it is obvious that a coherent description of everyday “argument” will turn
on recognizing that distinction: it is one thing to describe or explain an argument]

that someone makes, and something quite different to describe or explain an
argumentz that two persons are having. (p. 127)

The distinction between the two senses of argument is useful in that it serves to clarify the

issues to be addressed by argument theorists, such as “How are argumentsz conducted in
everyday life? What strategies are employed in making argumentsl? Along what
dimensions do argumentsz differ?” (p.128). We will see later that the distinction between

the two senses of argument will help to clarify several differences between ordinary

argument (the activity which encompasses both argumentl and argumentz) and pseudo-

argument.2

3.2. Features of argumentation
3.2. 1. Argumentation occurs in interaction

Contrary to the prescriptive and logical aims of earlier argument theorists,
contemporary treatments of argument are mostly descriptive and view argument as a
communicative activity embedded in human interaction. Van Eemeren and Grootendorst

(1984) locate argumentation squarely within the domain of interaction:

Unlike logicians, argumentation theorists concern themselves primarily with
argumentation in colloquial speech This means that the object of their research is
not constituted from combinations of premises and conclusions formulated with the
help of formal symbols whose meaning is established unambiguously beforehand,
but from constellations of statements by language users... (p.4)

63

Instead of treating argument as a static, abstract entity, it is viewed by argument theorists as
a structured activity which emerges in interaction, between actors with particular goals,
strategies, and relationships in a variety of settings.

Viewing argument as an activity in interaction means viewing it as something which
unfolds in time, where arguers are seen as “creative participants acting in concert,
collaboratively creating, shaping, and changing events by interpreting their options and
strategically adapting to the expectations and actions of others” (Willard, 1989, p.67).
Willard warns that the analyst’s fascination with reiﬁed, abstract structure may easily
obscure the emergent nature of real arguments in which real people strategically adjust their
goals, modify their stances, and respond to and anticipate the reactions of others in real
contexts. Arguments may end abruptly or may end in resolution, and the same argument

may be sustained over several encounters.

3.2.2. Argumentation is cooperative.

Another key assumption of many contemporary argument theorists is that argument
is a cooperative enterprise in which participants attempt to arrive at an agreement. Theorists
have challenged the notion that the verbal conﬂict involved in argumentation is disorderly,
unruly and inherently face-threatening. For instance, Willard (1989) states that
“Arguments cannot succeed without strong underlying cooperative bonds that keep the
dissensus psychologically and sociologically manageable” (p.61). He argues that while

arguments do allow kinds of acts which are not entirely acceptable in ordinary conversation

64

(e. g., challenges, demands and attacks), these usually occur within conventional bounds of
civility deﬁned by such things as one’s ethics, standards of professionalism, etc. In
addition, Willard states that arguments are intrinsically cooperative achievements because
the dissensus upon which argumentation is based minimally requires some common
ground and shared assumptions such as agreement that the parties disagree in the ﬁrst
place.

Van Eemeren et al (1993) argue that since argumentation is resolution-oriented, it
must be sufﬁciently regimented in order for resolution to be attainable. This regimentation
involves rules of conduct and certain rational procedures and assumptions (e.g., standards
of reasonableness) which must be adhered to for the argumentation to make any progress
towards resolution. That is, resolution is unlikely unless participants in argument
cooperate to a certain extent.

Jackson and Jacobs’ work on conversational argument (1982, 1986) places
cooperation at the heart of this activity in that argumentation is seen as a regulatory
mechanism primarily employed by participants to restore alignment between them;
“...argument is not so much a way of disrupting the working consensus of
conversationalists, as much as it is an effort on the part of conversationalists to enforce it”
(Jacobs, 1986, p.234).

To illustrate how participants in argument attempt to restore alignment between

themselves, consider the following excerpt which was presented in chapter two:

65

(1)

1 B: Sounds like your family is kind of against Black
2 people.

3 C: No. no not at all.

4 B: Not agAINST, but have had some bad experience

5 with- or whatever.

6 C: N— no, my m- see my mom’s like the kind of
7 person th— you know she went to like Catholic
8 highschool (and that) and she was just brought up
9 to believe that everybody’s equal and that sort of
10 thing. And my dad pretty much feels the same way.
11 He went to Catholic private highschool and all this
12 and that - So my parents aren't like that at all.
13 But it’s just like MN experience — it's like - =

[
14 B: oh that’s good

In this excerpt, B. makes an assertion (in lines 1-2) which is directly disputed by C. (in line
3). In lines 4-5, B. attempts to restore alignment between herself and C. by rephrasing her
assertion in a manner which she presumably expects C. to ﬁnd more acceptable. In the ﬁrst
part of this turn, B. self-repairs (“Not agAINST”), thus withdrawing her initial assertion in
response to C.’s straightforward dispute. In the second part of her turn, B. modiﬁes her
assertion and in this way makes a second bid for C.’s agreement.

This example shows how participants in argumentation actively attempt to achieve
realignment by withdrawing and modifying assertions. In this case, in line 6, C. disputes
B.’s modiﬁed assertion, and then goes on (in lines 6-12) to explain why B. is wrong to
assert that his parents are prejudiced against Black people. In line 14, B. displays her
acceptance of C. explanation (and her approval), and the alignment between participants is

restored. Thus, this excerpt illustrates how when opposition arises in conversation,

66

participants cooperatively work towards reestablishing alignment. Furthermore, the
excerpt shows that argumentation is not meant to disrupt “a working consensus,” but rather

is employed as a method to enforce it.

3.2.3. Argumentation is “public.”

Contemporary argument theorists also seem to agree that argumentation is a
“public” activity. That is, to argue, one must “go public” by overtly committing oneself to
the meanings of one’s words, being accountable for what one says. In this way, speakers
may be forced to become committed to the outcomes of the positions that they take, even
when this is not intended or desired. As soon as one stakes out a position in argument, he
or she must live with the results (or face the repercussions of withdrawing that position,
i.e., embarrassment, humiliation). Thus, positions put forward in argument are public in
that they take on a life of their own, and to adopt a position is to be responsible for its
outcomes (Willard, 1989, p. 258).

To illustrate, consider the following example from the data in which T. holds A.

accountable for a prior position:

(2)

T: A----, you just stated that though that you
wanted it to be a more integrated greek system.
Reverse discrimination is MEANT is help integrate
the society. Not segregate the society. Um -
Reverse discrimination - is Black - is Black =

Lﬂrbtallwl"|

67

(5 lines deleted)

11 T: So if you're against reverse discrimination, you
12 want them to TAKE White people. So what you’re

13 more interested in is the Black sororities and

14 fraternities i— inviting the White people, or

15 asking the White people - to join their =

This excerpt is part of a larger argument between T., G., and A. about the nature of reverse
discrimination. The relevant point concerning this example is that T. (in lines 1-2) invokes
a position taken earlier by A. in order to catch her in what he takes to be a contradiction. In
this case, T. orients to the public, accountable nature of positions put forward in argument,
and attempts to use A’s own words against her.

Argumentation is public in another way — it involves positions and supports
which are overtly expressed and not the underlying thoughts or beliefs of a speaker; “What
people argue over is not so much the actual positions of the parties, but the ones that they
can be held to have expressed” (Jacobs, 1986, p.237). It is for this reason that
indirectness, vagueness or ambiguity may be strategically employed in order to avoid being
held publicly accountable for some implied argument.

Van Eemeren et a1 (1993) argue that while attitudes, beliefs, and processes of
reasoning certainly underlie argumentation, their inﬂuence on argument must be seen as
“channelized by a system of public commitment and public accountability” (p. 12).
Similarly, van Eemeren and Grootendorst (1984) claim that this emphasis on the public,
“external” nature of argument is essential to guard against the danger of “internalizing” or

psychologizing the subject of investigation:

68

We believe that argumentation theory must concern itself with differences of
opinion and with efforts to resolve disputes about expressed opinions by verbal
means. This means that the argumentation theorist must concern himself with
expressed opinions and argumentative statements and not primarily with the
thoughts, ideas and motives which may underlie them. (p.6)

According to van Eemeren and Grootendorst, this stress on what they call
“externalization,” coupled with their theoretical apparatus (see section 3.4.1.), allows for a
scientiﬁcally interesting way to “reconstruct” (a procedure to be discussed in section
3.4.1.3.) underlying intentions, thoughts, etc. from what people actually say in

argumentation.

3.3. Definitions of argument

Before going on to present two discourse analytic models of argumentation, in what
immediately follows the various contemporary characterizations of argumentation will be
summarized by discussion of several prominent deﬁnitions of argument which have
surfaced in the literature. It will become apparent that all of these deﬁnitions share certain
features, most prominent of which is the condition that at least two people are in
disagreement.

Willard (1989) deﬁnes argument as follows; “argument is a kind of interaction in
which two or more people maintain what they construe to be incompatible positions.”
Similarly, Blair (1987) claims that argumentation begins with a disagreement between two
or more people about a point of view, where “the point of the argumentation is to convince

the other side to change its mind. The argumentation proceeds by each party criticizing the

69
other’s point(s) of view, and defending its own” (p.191).

Schiffrin (1985) deﬁnes argument as “...a discourse genre through which
individuals support disputable and disputed positions” (p.45). Given Schiffrin’s
discourse-analytic background, it is not surprising that she characterizes argument ﬁrst as a
discourse genre, unlike Willard who sees argument as primarily and more generally as a
kind of interaction. However, it is clear that Schiffrin also recognizes that argument (as a
kind of discourse) occurs through interaction. She distinguishes between ‘rhetorical’
argument and ‘oppositional’ argument (a distinction very similar to O’Keefe’s). She
deﬁnes ‘rhetorical’ argument (O’Keefe’s argumentl) as “discourse through which a
speaker presents an intact monologue supporting a disputable position” (p.37), and she
deﬁnes ‘oppositional’ argument (O’Keefe’s argumentz) as “an interaction in which an
opposition between speakers creates an extended polarization that is negotiated through a
conversation” (p.41). Clearly Schiffrin’s deﬁnition of oppositional argument closely

resembles Willard’s deﬁnition of argument (which seems to pertain only to argumentsz) in

that in both there is reference to interaction between parties in opposition. In both cases,
that opposition is said to occur where speakers ‘maintain incompatible positions,’ or in
Schiffrin’s terms, where “speakers support openly disputed positions” (p.37).

Although Schiffrin treats rhetorical arguments as monologues, implicit in her
discussion is the assumption that rhetorical arguments also occur in interaction, and are
triggered by a speaker’s expectation of dispute by others. Schiffrin claims that rhetorical

argument is a discourse genre in which speakers attempt to establish disputable positions

70

such that “...speakers orient their presentations to a potential opposition” (p.40). In this
way, rhetorical arguments are dependent on features of the interactions in which they take
place.

Van Eemeren et a1 (1993) make the same point by stating that their “socialized”
perspective of argument leads to the conclusion that both arguments made (rhetorical
arguments) and arguments had (oppositional arguments) should be seen as social and
collaborative activities; “Argumentation is not so much a process whereby a single
individual privately draws a conclusion as it is a procedure whereby two or more
individuals try to arrive at an agreement” (p.12). Thus even arguments made by a single
individual emerge from real or projected dialogic interaction in which positions taken
respond to actual or anticipated doubts, objections, and questions of an antagonist.

Van Eemeren and Grootendorst (1984) discuss “argumentative discussions” (see below for
explanation of terms) which centrally involve disputes over an “expressed opinion” (or

position) where:

at least one language user has committed himself to that expressed opinion either
positively or negatively by advancing a positive or negative point of view in respect
of it and if at least one language user has explicitly made it known that he does not
subscribe to that point of view or at least has expressed doubt as to its
acceptability. (p.79)

Preston (1993) also acknowledges two kinds of argument, one which involves the

9 (6

support of positions (similar to O’Keefe s argumentl”) and another which involves overt

disputes of positions and supports (O’Keefe’s “argument;”). It is the latter kind which

71

Preston regards as closer to the folk sense of argument, and therefore his focus is on
oppositional argument. Preston goes on to identify arguments as; “...obligatorily having at
their core the denial of an assertion or of a belief behind an assertive or even non-assertive
act (loosely a presupposition), including belief about the connectedness or relevance of a
support.” (p.206). Thus, Preston’s deﬁnition emphasizes that opposition may arise in a
variety of ways and is not limited to the denial of the propositional content of an assertion.
We will see many examples in later chapters which illustrate that opposition in conversation
may be due to a variety of factors, including the denial of presuppositions and the relevance
of supports. It should be noted that while Preston’s characterization of (oppositional)
argument does not explicitly refer to participants, his discussion and examples assume that
argumentation requires at least two present parties in opposition.

Despite the differences in orientation and emphasis of the various deﬁnitions
presented above, they all share a small set of characteristics. First, most of the deﬁnitions
make reference to two or more participants who must be present for an argument to occur.
For those deﬁnitions which explicitly distinguish the two senses of argument (for example,

Schiffrin’s), this condition applies to argumentsl as well as to argumentsz. Each of the

deﬁnitions also implicitly assume or explicitly state that at least one participant will play the
role of protagonist with respect to some position and at least one will play the role of
antagonist. Second, all of the deﬁnitions refer in one way or another to the condition that
there be some opposition/disagreement between the participants. Willard (1989) states that

“It may seems trivial to say that arguments require dissensus, but this claim is one of the

72

few points of agreement among argumentation theorists” (p.53).

These two conditions, that there be two or more participants who are in opposition
or disagreement are shared by all of the deﬁnitions of argumentation discussed above, and
these conditions for argument will be assumed in this paper. In the following chapters it
will be shown how these conditions are not satisﬁed when it comes to the data analyzed in

this study.

3.4. Example of argumentation from the data.
Consider the following example from the data, which illustrates many of the

properties of argumentation discussed so far in this chapter:

(3)

1 A lot of Black people have to go out on their OWN

2 and get their own fraternities and sororities. Why

3 can't they just join - THE greek system. But they =

[

4 G: Cause they

5 feel COMfortable.

6 A: = to have their own BLACK greek system.

7 G: Cause they feel COMfortable. Cause - cause

8 they're a TWELVE percent minority, so why are they

9 going to want to be: - among twelve percent - of

10 people of their own KIND, Whereas ah - they could =
[

11 A: Because =

12 G: = (ever-)

13 A: = that' 3 what college is about. Meeting new

14 people. I' m JEWish. It' s not like the whole -
15 fricken world is JEWish,

16 G: Yeah! But the sorority you chose happens to be
17 mostly JEWish! And I think that that- - I think
18 that MOST people choose - a house based on people
19 that are similar to oneSELF!

73

This excerpt exempliﬁes many of the features of ordinary argumentation discussed
throughout this chapter. Most importantly, it involves positions, disputes of positions, and
supports of those disputes. These occur between two present participants who assume the
roles of protagonist and antagonist with regard to certain standpoints. Here, in lines 1-3,
6, A. takes the position (partly in the form of a rhetorical question) that Black students
should not establish a separate greek system, but rather should integrate into “THE greek
system.” Playing the role of antagonist, G. indirectly disputes this position (in lines 4-5,
7-10) by arguing that Black people are justiﬁed in joining Black fraternities and sororities,

because as a minority they may feel more comfortable with others like themselves.3 A.

then disputes the position implicit in G.’s rhetorical question (in lines 8-10) by arguing (in
lines 11, 13-15) that Black students should want to integrate because “that’s what college is
about.” In other words, meeting people who are different from yourself is an important
aspect of college life. She provides a second support (in lines 14-15) by arguing that just
because she is Jewish (another minority), she doesn’t think that everyone is Jewish, which
seems to imply that Black people unreasonably narrow the range of their social relations.
At this point G. then disputes both of A.’s supports by pointing out that A. is guilty of
exactly the behavior she criticizes (lines 16-19); she belongs to a Jewish sorority and for
precisely the same reason that Black students belong to Black fraternities and sororities —
because people choose to be with others who are like themselves.

Second, it should be noted that the excerpt shows participants as cooperatively

seeking to restore an alignment among themselves by attempting to persuade each other by

74

means of arguments. Although resolution (realignment) is not reached in this case, the
argumentation itself should be viewed as an effort on the part of participants to come to

terms with their disagreement, and to establish a “working consensus.”

3.5. Argument as speech activity

It should be clear by now that once it is viewed within its interactive context, the
analysis of argumentation must contend with it as a linguistic/discoursal activity. That is,
argument must occur between language users. It is for this reason that argument theorists
have turned to examine the discoursal context in which argument takes place. Van Eemeren
and Grootendorst (1984) claim that argumentation should be viewed not just as product,
but also as process; “...we think that theoretical thinking about argumentation should be
concerned not only with the structure of completed argumentative constellations but also
with argumentation as a verbal activity” (p.7)

Willard (1989) states that two necessary conditions for “having an argument” are
that the arguers have adequate communicative competence to employ comprehensible
utterances, and know the rules of the conversational system; “Seeing argument as
interaction thus requires that we see it as a matter of conversational rules and structure...”
(p.45). Similarly, Schriffrin (1985, p.37) claims that since arguments are a discursive
activity, they depend on some of the same underlying assumptions and principles of
discourse organization as other kinds of discourse.

It is claimed here that argumentation is a well-deﬁned speech activity (in Levinson’s

75

sense), in part because it is a goal-deﬁned, socially constituted event. In addition, it will
become apparent in what follows that argumentation is a speech activity because it involves
constraints on the kinds of allowable contributions by participants, and further because
these constraints provide grounds for the correct interpretation of the contributions of
others by hearers. In the rest of this chapter, two discourse-analytic models of

argumentation which incorporate these insights will be presented. It will be shown in the

following chapter that the data of this work do not neatly ﬁt either of these models.4

3.5. 1. A speech act approach (van Eemeren and Grootendorst, 1984; van Eemeren et al,
1993).

Current research within a modiﬁed speech act theory approach has tried to show
that at the structural level of action sequences, having an argument is a speech activity
composed of sequences of speech acts, while making an argument is a complex speech act
itself (van Eemeren et a1, 1993, p.4). The following sections will attempt to ﬂesh out the
properties of this speech act approach to argumentation. Finally, the method of
“reconstruction” advocated by van Eemeren et a1 will be presented, since it will serve as a

tool in the analysis of the data in later chapters.

3.5.1.1. Argument as illocutionary act . Van Eemeren and Grootendorst (1984) deﬁne

argument (in the sense of “making an argument”) as:

76

...a speech act consisting of a constellation of statements designed to justify or
refute an expressed opinion and calculated in a regimented discussion to convince a
rational judge of a particular standpoint in respect of the acceptability or
unacceptability of that expressed opinion. (p.18)

Here an “expressed opinion” is taken to consist of a proposition (positive or negative)
which may refer to facts, ideas, actions, and attitudes. A “standpoint” means the (positive
or negative) stance or attitude taken up with respect to the expressed opinion. Thus a
language user may take a positive or negative standpoint toward a positive or negative
expressed opinion. For example, I might take a negative standpoint with respect to the
negative expressed opinion that there should not be an increase in taxes by saying “I do not
believe that there should not be an increase in taxes.” My argumentation then, according to
van Eemeren and Grootendorst, would consist of a “constellation of statements” designed
to convince a rational judge (the antagonist) of the unacceptability of the expressed opinion.

Note however, that van Eemeren and Grootendorst’s treatment of argument as an

illocutionary act requires several important modiﬁcations of Searle’s approach.5 First,

Searle’s theory regards the sentence as the characteristic grammatical form of the
illocutionary act such that there is a one-to-one relation between the utterance of a sentence
and the performance of an illocutionary act. Obviously, van Eemeren and Grootendorst’s
treatment of argument requires that the single illocutionary act consist of more than one
sentence. In addition, the illocutionary act of argument may consist of a constellation of
statements which are themselves illocutionary acts (e. g., assertions). Thus van Eemeren

and Grootendorst argue that the assumption of a one-to-one relation between sentence and

77

illocutionary act should be abandoned in favor of an approach which distinguishes between
sentence level and higher textual level illocutionary forces.

According to this view, “compound illocutions” (or “illocutionary act
complexes”), which consist of ordered sequences of sentences, act at a higher level
in the same way as lower level “elementary” (sentence level) illocutions. That is,
such higher level illocutions serve single functions (have a single force) in the same
way as lower level illocutions such as promises, questions or requests. Van
Eemeren and Grootendorst argue that the notion of compound illocutions preserves
the essential insight of Searle’s work that there is a one-to—one correspondence
between functional units and linguistic units.

Van Eemeren and Grootendorst (as well as van Eemeren et a1, 1993) treat
argument as a compound illocution composed of a constellation of statements which
are assertives. Furthermore, this constellation of statements would constitute the
illocutionary act of argumentation if certain conditions are fulﬁlled. The complex

act of making an argument, according to van Eemeren et a1, is structured by the

following felicity conditions:6

1. Pr0positional content conditions:

a. The speaker, S, has put forward a standpoint (or opinion) 0.
b. S has advanced a constellation of assertives, A1, A2,..., An, in which

propositions are expressed.

2. Essential condition: Advancing this constellation of assertives counts as an
attempt to justify O to the satisfaction of the hearer, H, i.e., to convince H
of the acceptability of O by means of A1, A2,..., An.

78

3. Sincerity conditions:
a. S believes that O is acceptable.
b. S believes that the propositions expressed in A1, A2,..., An

are acceptable.
c. S believes that A1, A2,..., An constitute an acceptable justiﬁcation of O.

4. Preparatory conditions: ‘
a. S believes that H will not or does not accept O at face value.
b. S believes that H will accept the propositions expressed in A1, A2,...,An

c. S believes that H will accept A1, A2,..., An as a justiﬁcation of 0.

d. S believes that either the propositions expressed in A1, A2,..., An are
not already obvious to H, or A1, A2,..., An constitute a
justiﬁcation of O that is not already obvious to H, or both.

These conditions (propositional content, essential, sincerity and preparatory) are based on
Searle’s (1970) model of the “constitutive” rules which must be carried out in order for
some illocutionary act to be performed successfully. These rules or conditions are
constitutive in the sense that if they are all fulﬁlled, then the act is carried out successfully.
According to van Eemeren et a1 (1993), the proposition content condition for
argumentation is that a speaker has put forward a standpoint and a number of assertives.
The essential condition states that the constellation of assertives counts as an attempt by the
speaker to justify the standpoint to the satisfaction of the bearer. From the hearer’s point of
view, if the utterances are recognizable as argumentation, then the propositional content and

essential conditions have been fulﬁlled. In addition to these, for the argument to be

correctly carried out, certain preparatory and sincerity conditions must be satisﬁed.7

The sincerity conditions state that the speaker must believe that the standpoint

(opinion) is acceptable, that the propositions underlying the assertions used to support the

79

standpoint are in themselves acceptable, and ﬁnally that these assertions are acceptable as
justiﬁcation for the standpoint. If any of these sincerity conditions is not satisﬁed, then the
argumentation will be defective. For example, if the speaker does not really ﬁnd the
expressed standpoint acceptable, then he or she misleads the hearer into believing that there
is a dispute whereas there is actually no dispute.

The preparatory condition (4a) requires that the speaker believe that the hearer does
not accept the standpoint at face value, so that in order for the argumentation to arise, the
speaker must anticipate opposition or doubt from the bearer with regard to the standpoint.
Preparatory condition (4b) states that the speaker must believe that the hearer will accept the
propositions of the assertions put forward, and condition (4c) states that the speaker must
believe that the bearer will accept the assertions as justiﬁcation for the standpoint. Finally,
condition (4d) requires that the speaker believe that the assertions and their use as
justiﬁcation for the standpoint are not already obvious to the bearer.

According to van Eemeren and Grootendorst (and van Eemeren et a1), if any of
these preparatory conditions is not satisﬁed, then the argument will be defective. For
example, if condition (4c) is not met, this means that the speaker does not believe that the
bearer will accept the assertions as justiﬁcation for the standpoint, and therefore the
argumentation will be pointless since it implies that the speaker believes in advance that the

argumentation lacks cogency (van Eemeren and Grootendorst, 1984, p.46).

80

3.5.1.2. The stages of critical discussion (arguments made). So far, the presentation of
van Eemeren and Grootendorst’s (1984) and van Eemeren et al’s (1993) speech act model

of argument has touched upon only argument] as a compound illocution, and has ignored
the larger discoursal context of argumentsz. Van Eemeren and Grootendorst do address the
issue of how the compound speech act of argument is employed in argumentsz. Since they

reserve the term “argument” for the compound illocutionary act itself, they prefer to refer to
the activity in which an attempt is made to resolve a dispute about an expressed opinion as
an “argumentative discussion.” An argumentative discussion will involve not only the
argument speech act, but also other speech acts as well, and will vary in terms of its
complexity. The simplest argumentative discussion will involve a single dispute regarding
a single expressed opinion. More complex types will involve multiple disputes having to do
with multiple expressed opinions; “One argumentative act complex may support or
complement another, and it may be only together that they constitute a complete attempt at
justiﬁcation or refutation” (van Eemeren and Grootendorst, 1984, p.76).

Van Eemeren and Grootendorst argue that any argumentative discussion, no matter
how complex, will be composed of one or more single disputes, and any single dispute
within an argumentative discussion will itself be composed of certain stages, including the
confrontation, the opening, the argumentation, and the concluding stage. The
confrontation stage involves the recognition by the parties that there is some dispute —— that
is, one participant must cast doubt on some standpoint. If the parties wish to attempt to

resolve the dispute, they may embark upon the opening stage of an argumentative

81

discussion where a participant challenges the other to support his or her standpoint. Thus,
in stage two (the opening stage) the participants assume the roles of protagonist and
antagonist with respect to the standpoint in question.

The ‘discussion proper,’ which involves the actual move towards resolution of the
dispute begins in the argumentation stage, whereby the protagonist attempts to justify the
standpoint by putting forward argumentation. The antagonist is then entitled but not
obligated to cast doubt on aspects of the protagonist’s argumentation (on the assertions, or
on their potential as justiﬁcation for the standpoint). This process of calling into question
aspects of the protagonist’s argumentation may in principle continue indeﬁnitely, although
it is more common for one of the parties to quickly abandon his or her initial position. The
ﬁnal stage of a simple argumentative discussion is the concluding stage in which it must be
decided by the participants whether or not the dispute was resolved. Although many
simple disputes are never actually resolved (in that neither participant has a change of
mind), the participants must still establish its outcome, and the mutually recognized lack of

resolution is an outcome.

3.5.1.3. The reconstruction of argumentative discussions. The approaches of van Eemeren
and Grootendorst (1984) and van Eemeren et a1 (1993) are “normative” or “critical” in that
they attempt to propose an ideal (speech act) model of argumentation against which to
describe, analyze and evaluate actual cases of argumentation in naturally occurring

discourse. One of the main concerns of van Eemeren et a1 (1993) is to reconcile normative

82

and descriptive approaches to argumentation by relying on the assumption that:

...speech act performance is a real-world phenomenon whose natural organization
is normative. In analyzing discourse as speech act performance, we describe actual
practices which are themselves embodiments of underlying standards of
reasonableness and rationality. (p.2)

According to this assumption, actual argumentation has an underlying normative
organization, and therefore it is possible to analyze an actual instance of argumentation in
light of how it lives up to such standards as reasonableness and rationality (upon which
speech acts are based). It is clear that not all instances of actual argumentation measure up
to the ideal model proposed within this speech act approach. For example, many real
arguments do not follow all of the stages of argumentative discussions proposed by van
Eemeren and Grootendorst (1984), and many do not involve objective, disinterested,
cooperative, resolution seeking participants (i.e., some arguments end in bloodshed).
Further, real arguments do not always involve participants of equal status or power. Thus,
since a normative model of argument is based on “rational” principles which allow
participants to maximize efﬁcient resolution (and which presumably are oriented to by
participants themselves), it also allows the analyst to determine the ways in which ordinary
(non-ideal) argument deviates from the ideal.

Further, according to van Eemeren et a1 (1993, p.34), the normative model offers a
framework for interpreting and “reconstructing” argumentation in actual discourse. In this

approach, reconstruction is taken to mean any representation of discourse fashioned to

83

some theoretical perspective so that it involves a coding of discourse which highlights
certain features and ignores others. The reconstruction of argumentative discourse then,
involves a coding of an instance of argumentation which highlights those features that are

relevant to the resolution of a dispute:

...the primary interests of a normative reconstruction of argumentative discourse are
to recover the propositions that make up the substance of the arguments, to
determine how the arguments are used to justify or refute the standpoints at issue,
and to examine how the performance of speech acts can function at a particular
stage of discussion to help resolve a dispute. (p.38)

Moreover, when we direct these analytic aims towards an instance of actual argumentation,
it becomes apparent that actual argumentative discussions do not completely correspond to
the ideal model. For example, the purposes of contributions are rarely explicitly stated, and
stages are rarely openly announced. In addition, there is much that occurs in actual
argument which is not considered in the ideal model, such as repetitions of arguments, and
speech acts other than assertives employed to present standpoints (e. g., offers or
expressives), or arguments (e. g., questions, complaints). Nonetheless, van Eemeren et al
argue that in most cases it is intuitively apparent that an argument is taking place, even
though what occurs does not always explicitly and completely correspond to the ideal
model. Thus the problem for reconstruction of an actual instance of argumentation is to
recover our intuitions about how the discourse can be seen to‘ have an argumentative
function while maintaining a sensitivity to the ways in which the argumentation is

communicated (p.39).

84

According to van Eemeren et a1 (1993), any reconstruction of actual argumentation
must be empirically justiﬁed in that it “imputes structures, functions, and content to the
discourse that the participants themselves and natural-language users in general intuitively
recognize” (p.40). Grounding for claims in reconstruction may come from ethnographic
evidence as well as from implied propositions within the text which may be inferred from
our knowledge of conventional structures and discourse strategies. Another source of
empirical grounding for claims in reconstruction comes from cues from the participants
themselves which tell us how they interpret the functions of utterances within the
argumentative discussion (e. g., pauses, discourse markers, etc.). According to van
Eemeren et a1, these different sources for empirical grounding can, taken together, help to

justify the analyst’s claims about the speciﬁc features of an instance of argumentation:

None of these sources works alone, and all work against the background of the
analyst’s and reader’s own cultural knowledge and intuitive competence as a native
speaker. Ultimately, the acceptability of any particular reconstruction of a
discourse will depend on its overall coherence, its accountability to the details of the
text, and its consistency with other information about how this case works in
particular, how related cases of this type work in general, and how discourse in
general is known to work. (p.44)

 

Reconstruction of argument must not only be empirically grounded, but also
normatively grounded, since it must concentrate on those features of the discourse which
relate to the resolution of disagreement and ignore other (equally empirically grounded)
discourse functions and meanings which do not bear on the considerations of the ideal

model. In addition, it may be the case in reconstruction that ambiguity arises, where there

85

is no way to empirically determine with any conﬁdence some meaning in the discourse. In
such cases, van Eemeren et al suggest that the reconstruction lean towards the strongest
interpretation which provides the best ﬁt with the normative model. That is, in ambiguous
cases the reconstruction should select the interpretation best in keeping with the assumption
that the discourse is designed to resolve disagreement in a rational manner.

According to van Eemeren et a1 (1993, p.60), the reconstruction of argumentative

discourse should be guided by a search for:

1. the points at issue,

2. the different positions that the parties concerned adopt with respect to these
points,

3. the explicit and implicit arguments that the parties adduce for their
standpoints,

4. the structure of the argumentation of each of the parties.

Van Eemeren et a1 (1993) identify four “transformations” which are applied in the
reconstruction of an argumentative discussion — deletion, addition, permutation, and
substitution. “Deletion” involves the selection of those elements from the original
discourse which are relevant to the resolving the dispute at hand. Other elements which are
not directly relevant to this purpose (i.e., elaborations, clariﬁcations, repetitions, etc.) are
omitted. “Addition” involves supplementing the actual discourse by ﬁlling in those
elements which are relevant to resolution of the dispute but which are implicit or
unexpressed. These implicit elements in the actual discourse might be unexpressed

premises or conclusions. “Permutation” involves rearranging elements in the discourse in a

86

way which makes the process of dispute resolution as clear as possible. This could
involve, for example, moving a late occurring confrontation out of the argumentation stage
or concluding stage back to its “ideal” place at the beginning of the argumentative
discussion. Finally, the transformation of “substitution” involves the attempt by the analyst
to produce a clear and explicit presentation of the argument. This could involve, for
example, creating uniformity in the reconstruction by replacing rhetorical questions which
function as arguments with assertions, and reducing different formulations of a standpoint
with a single formulation.

According to van Eemeren et al, the application of these transformations in a
reconstruction of argumentative discourse ought to be motivated by empirically sound
observations concerning the discourse. The reconstruction should not render a discourse
incoherent, should deal with only those aspects of the discourse related to the resolution of
dispute, and should involve as little change to the original discourse as necessary to
faithfully represent the argumentative discussion. Van Eemeren et al’s method of
reconstruction will be employed as a tool in the analysis of the data of this study in later

chapters.

3.5.2. Conversation Analytic approach (Jackson and Jacobs, 1982, 1986)

Another inﬂuential approach to argumentation in discourse is the work of Jackson
and Jacobs (1982, 1986, henceforth J&J) within the framework of conversational analysis.
The essential insight of this approach is that conversational argument is a kind of repair

mechanism where participants attempt to achieve a realignment with each other; “Argument

87

is a way of managing the practical problems presented by the actual or potential
withholding of an agreement response and by the failure to withdraw or suppress the kinds
of acts that elicit disagreeable responses” (Jacobs, 1986). Jackson and Jacobs propose that
argument in conversation is the sequential expansion of an adjacency pair which results
from the actual or anticipated inability to provide a preferred second-pair part. According to
J&J (1982), conversational arguments are ‘disagreement—relevant’ speech events, which
means that; “They are characterized by the projection, production, suppression, or
resolution of disagreement, so that they function not only to manage cases of expressed
disagreement, but also to regulate the occurrence of disagreeable speech acts” (p.224).
Argument is seen as a regulatory device or procedure which may be employed to avoid
overt disagreement in the service of the preference for agreement (Pomerantz, 1984) and
cooperation between interlocutors.

J&J incorporate speech acts into their conversational analytic approach by showing
that arguments often result from the necessity of giving dispreferred responses to certain
speech acts (like requests, offers, commands, etc..) because of the perceived failure to meet
certain felicity conditions associated with those acts. For example, an argument might arise

in response to an infelicitous command issued by an employee to a boss:

(4)
employee: “Go to lunch now”

boss: “Who are you to tell me to go to lunch!”

88

In this way, arguments are seen to result not only from propositions, but also from
infelicitous speech acts (but see Preston (1993) for criticism of this approach).

Although there are a variety of types of expansion discussed by J &J which are
available for avoiding overt disagreement and achieving realignment between interlocutors
(presequences, embedded sequences, postsequences, etc.), most relevant to the purposes
of this paper are ‘within-turn expansions.’ ‘Within-tum expansions’ are supports which
are built into f'n'st-pair parts in which disagreement or doubt from another party may be
anticipated. That is, if a speaker anticipates disagreement, he or she can provide supporting
arguments which could avert that disagreement. One important outcome of the
presumption of the preference for agreement is that supporting arguments will not be
proffered unless there is some real expectation of disagreement or doubt, hence the
enthymematic nature of conversational argument; “Speakers do not offer arguments for
their acts unless they have reason to anticipate some particular objection or unless they are
challenged” (Jacobs, 1986, p.234).

Further, the same conversational principles which regulate the production of
arguments also serve as interpretive resources for hearers — supporting arguments
(sequential expansions) will likely be interpreted by hearers as attempts to avert some
potential disagreement. That is, hearers are likely to attribute sequential expansion to the
anticipation of disagreement or doubt on the part of the speaker. Consider a hypothetical

example to illustrate this point:

89
(5)
A: Where were you this afternoon?

B: At the library...
You can ask Jim... he saw me there.

In this example, B.’s sequential expansion (supporting argument) would likely be
interpreted by A. as B.’s anticipation of doubt (in this case, if there were actually no doubt
on A’s part to begin with, it would certainly arise after B.’s expansion!).

In sum, according to the conversation analytic approach of Jackson and Jacobs,
argument in conversation is the sequential expansion of an adjacency pair which results
from the actual or anticipated inability to provide a preferred second-pair part. Furthermore,
argumentation in conversation arises either when one party objects to what another has
said, or else when one party anticipates disagreement and seeks to forestall it by providing
unsolicited arguments. Argumentation is thus viewed as a way of managing the practical
problems which arise in conversation where an agreeable response is actually or potentially
withheld. Most importantly for the purposes of this paper; “...conversationalists offer
arguments only if they encounter disagreement or if they have some reason to expect
disagreement; when arguments are offered, they are typically limited to just what is needed
to satisfy expressed or expected disagreements” (Jackson, 1986, p.218). We will see in the
following chapter that in some kinds of conversation, this is not always the case — there
are many instances in the data of this study where unsolicited arguments are provided

without any expectation of disagreement at all.

Chapter 4

THE STRUCTURE OF PSEUDO-ARGUMENT

4.0. A structural approach

The previous chapter was devoted to reviewing scholarship on argumentation with
the goal of showing that despite differences in orientation, most contemporary argument
theorists are in consensus that argumentation is an interactive, discursive activity between at
least two people who are in disagreement. The ﬁnal part of chapter 3 presented two
discourse-analytic models of argument which incorporate these essential features of
argumentation. Both van Eemeren and Grootendorst’s (1984) and Jackson and Jacob’s
(1982, 1986) work treat argument as a type of discourse which is uniquely structured, and
governed by pragmatic and conversational principles. With these discourse-analytic
models of argument in mind, the stage is set for an analysis of the data of this study. It will
be shown that although the conversations which comprise the data resemble arguments in
many respects, they clearly do not conform to either model in a neat way. While the
conversations give evidence for the continuous presence of positions, supports and
disputes, there is no opposition or disagreement between participants — the disputes are

aimed not at other (allied) participants, but rather at a nonpresent antagonist whose

90

91

positions and supports are continuously imported into the conversation. In addition, unlike
the argumentation depicted by the two discourse-analytic models in which one individual
presents a position followed by support, in the data of this study positions may be
collaboratively supported by two or more participants.

In sum, it will become clear that the conversations being analyzed resist
characterization in terms of either discourse model of argument. Further, a case will be
made that these conversations involve a speech activity which I will call “pseudo-
argument,” for reasons that will become apparent in this chapter and the ones to follow.
This speech activity has features of prototypical argumentation but is distinct in terms of its
structure and function. The remainder of this chapter will then aim to characterize and
formalize the structure and interactional framework of pseudo-argument. It will be shown
that the patterns evidenced by the data reveal a kind of collaborative “us versus them”
argument which takes place between two or more allied protagonists and some conjured

absent antagonist.

4.1. Why data do not fit the discourse models.
4.1.1. Van Eemeren and Grootendorst’s speech act approach.

The data are problematic in terms of van Eemeren and Grootendorst’s speech act
approach in that there are apparent violations of preparatory conditions (a) and (d), repeated

here for the reader’s convenience:

92

a. S believes that H will not or does not accept O at face value.

d. S believes that either the propositions expressed in A1, A2,... An are
not already obvious to H, or A1, A2,..., An constitute a
justiﬁcation of O that is not already obvious to H, or both.

It appears to be the case that speakers and hearers in many of the conversations are
in accord from the beginning with regard to positions put forward (violation of condition
(a)), and further are already familiar with many of the positions and their respective
supports provided by other participants (violation of (d)). To illustrate this, consider the

following excerpt from the data in which the topic of discussion is “unfair advantages”

given to Black students:

(1)

1 K: Why should they be coming to — a top ten uni- -

2 a big ten university, getting like ALL the money,

3 you know and I know people that get these checks, =
I

4 ?: ( )

5 K: = and don' t even need everything they need. And

6 the thing that I really don't like is how - I mean

7 this is something minor, but like - they get this

8 government money it’s like (to pay for) books, but yet

9 they return the books at the end of the year and

10 they keep that money. They don’t have to return
11 it. You know what I mean? They're getting like =
I
12 ?: Yeah
13 K: = all this free money and all this stuff when
14 like - here I am taking out loans that - the
15 interest is just adding up, and I'm going to
16 eventually have to pay it all BACK. You know, not
17 saying that I'm.WORried about it, cause I know —
18 that I'll be successful with MY job, but it’s just
19 not fair that
[

20 B: But they’re getting a free ride AND =

[
21 K: exACle

93

22 B: = they could be successful in THEIR job TOO. AND
23 plus they get this stupid point five increase in

24 their GRADES, so now they're going to be graduating
25 with- I

26 C: PLUS they have a better chance
27 of getting a job, so they're - going to be - more
28 successful ( )
I
29 B: So it' 3 just- i- their advantages just keep

30 adding UP! Their - their advantages totally keep
31 adding UP. Their grade advantages,
I

32 C: Yeah that’ 5 true. They do have - more
33 time to relax because - they don' t have to WORK.
34 You know, like everybody else does to

35 pay off their loans.

I
36 B: They'll be under less STRESS,

37 C: They- cause they’re just getting money.

This ﬁrst excerpt involved three White female participants. It reﬂects the high degree to
which participants in pseudo-arguments are in accord from the outset. Although positions
(“Opinions” in van Eemeren and Grootendorst’s terms) are staked out and supported, there
is no indication that the participants expect their positions to not be accepted at face value.
In fact, the opposite seems to hold true — participants expect and encourage each other to
collaboratively provide supports for stated positions. In addition, there is no indication that
participants are hearing these positions and their respective supports for the ﬁrst time.
Although this is difﬁcult to prove based on the transcripts alone, it appears that these
participants are familiar with the arguments that they are practicing.

Following the method advanced by van Eemeren et al (1993), a reconstruction of

the argumentation contained in the excerpt helps to shed some light on these points (here

94

instances of repetition of the main position and supports have been included, where ‘r’

represents repetition):

(K) POS: Black students receive unfair advantages (“all the money” and stuff) (1,2)
(K) SUP 1: They get checks that they don’t need (3,5)

(K) SUP 2: Keep the money from the books bought by the government (6-11)
(K) SUP 3: I have to take loans with interest, and pay it all back (13-17)

(B) POS r: They’re getting a ﬁee ride (not fair) (20)

(B) SUP 4: they could be successful too (22)

(B) SUP 5: they get .5 increase in their grades (23-24)

(C) SUP 6: better chance of getting a job (26-28)

(B) POS r: advantages keep adding up! (29-31)

(B) SUP 5 r: grade advantages (31)

(C) SUP 7: more time to relax (32-35)

(B) SUP 7 r: under less stress (36)

This reconstruction of positions and supports illustrates that this excerpt contains 7 distinct
supports for the one position that African Americans receive ‘unfair advantages.’ All three
participants provide at least one support, the main position is repeated twice (in line 20 and
lines 29-31), and supports are repeated (in line 30 and line 36). (For more detailed
discussion of this reconstruction, see section 4.2.1.)

Clearly there is no dispute which occurs between the participants in this excerpt,
and there is every indication that the participants are allied in their beliefs. Therefore,

although these exchanges have the appearance of argumentative speech acts, they are

95

defective (or ‘infelicitous’) in that preparatory conditions (a) and (d) are not respected. Van
Eemeren and Grootendorst state that if condition (a) is not met, then the speech act is
‘superﬂuous,’ and further; “...in that case, S’s performance of that illocutionary act
complex is in fact a waste of time and effort and both S and L know beforehand that it is”
(1984, p.45). It will become apparent later (in chapter 6) that this conclusion is too strict
— other secondary functions may be served by argumentative structures (e.g., identity
work), and so the violation of certain preparatory conditions does not necessarily constitute
a waste of time and effort.

Next consider this excerpt in terms of van Eemeren and Grootendorst’s notion of
argumentative discussion. Recall that argumentative discussions are said to involve at
minimum a single dispute where at least one participant assumes the role of protagonist and
one assumes the role of antagonist with respect to some standpoint. In the previous
excerpt, there is no indication that any participant has assumed the role of antagonist, nor is
there any sign of a dispute. In addition, simple argumentative discussions were said to
involve four stages (confrontation, opening, argumentation, and concluding). There is no
evidence of the ﬁrst two stages in the previous excerpt. That is, no participant casts doubt
on the standpoint of another participant (confrontation), and no participant challenges
another participant to justify some standpoint in dispute.

In sum, if van Eemeren and Grootendorst’s model is accepted, it should be
concluded that something other than ordinary argumentation is occurring in the data.

Participants in the data of this study appear to perform “defective” argumentative speech

96

acts, and do not appear to engage in the stages of argumentative discussion. Further, van
Eemeren and Grootendorst’s model of the complex argumentative speech act is unable to
account for these data in another way — in their model the speech act of argumentation is
the achievement of a single speaker who presents both the standpoint (position) and the
assertions (supports) in order to gain acceptance from an audience. In the data, this is
confounded since the single speech act may be composed of contributions from two or
more participants. For instance, in the previous excerpt, participants K., B., and C. all
contribute supports to the same position that Black students receive ‘unfair’ advantages.
How this is problematic for van Eemeren and Grootendorst’ s notion of a ‘compound’

speech act is an issue worth taking up, but beyond the scope of this paper.

4.1 .2. Jackson and Jacobs’ CA approach.

In section 3.3.2., it was noted that according to Jackson and Jacob’s CA approach, one
important outcome of the presumption of the preference for agreement is that supporting
arguments will not be provided unless there is some real expectation of disagreement or
doubt on the part of the speaker. The data of this work present a problem for the presumed
enthymematic character of conversational argument in that participants provide supporting
arguments for positions (ﬁrst-pair part assertions) where there is no apparent expectation of
disagreement or doubt. In many of the conversations that were analyzed, participants were
in complete accord throughout, and further collaborate in providing supports for the
positions of others. There is no indication that participants are defensive in the stances that

they take — on the contrary, they express enthusiastic support for each other at every

97

opportunity. Indeed, in the conversations which were analyzed (in particular, those which
will later be characterized as pseudo-argument), there was not a single instance of
disagreement between participants, and yet participants liberally supply supports for
positions.

One might object that in discussions which involve the presentation of beliefs, one is
dealing not with ‘conversational argument’ (and the repair of defective felicity conditions)
but rather ‘substantive argument’ in which interlocutors are pressed by the nature of the talk
they are engaged in to present supports, even for positions for which disagreement is not
necessarily expected (Willard, 1989, p.43). In substantive argument, the point is not to
repress or avoid looming disagreement, but instead to let disagreement ﬂourish within an
atmosphere of general skepticism and confrontation (Willard, 1989, p.44). In such cases,
“the inquisitorial nature of arguments forces interlocutors to be more explicit and thus less
enthymematic than other conversationalists” (p.104). However, even if we accept the
distinction between conversational and substantive argument, it is unclear how such a
distinction can help to account for the data, since the data involve nonconfrontational
discussions where participants are anything but skeptical (of each other’s ideas at least).

Thus, the data show ‘within-tum expansions’ in the form of supporting arguments in
which there is no expectation of doubt or disagreement. Supports are supplied freely,
apparently as a matter of course, without respect for the presumed enthymematic nature of
conversational argument. Further, the data do not involve substantive arguments in that

although beliefs are at issue, participants are not pushed to present unsolicited supports for

98

their beliefs due to any general skepticism of their interlocutors. However, the
intransigence of the data does not prove that 1&1 ’s approach is defective so much as
indicate that in the data something other than prototypical argument is occurring. We turn

to this point in the following section.

4.2. The structure of pseudo-argument

It will be argued here that PA deserves to be treated separately from CA in part at
least because it has different structural properties. Chapter 5 will buttress this argument
with linguistic evidence, and chapter 6 will complete this argument by showing that PA is
distinct from CA on functional grounds as well.

As was discussed in chapter 2, a structuralist approach to larger units of discourse
seeks to discover the ways in which they are bounded and composed of interrelated smaller
units which underlie their organization. For example, Labov’s treatment of personal
experience narrative showed this type of discourse to be bounded by an abstract at the
beginning and a coda at the end. Also, narratives are composed of interrelated smaller units
such as the complicating action, orientation and evaluation, all of which accomplish
different functions in the telling of a story. In the analysis that follows, similar kinds of
underlying structure in pseudo-argument will be argued for. With the aim of distinguishing
PA from CA, it will be shown that although these types of discourse share the same
underlying units (positions, supports, and disputes), they differ in how these units are
conﬁgured and distributed between participants.

Following Schiffrin (1987) and Preston (1993), it will be assumed here that

99

positions, disputes and supports are the underlying (ideational) units which serve as the
building blocks of argumentation in discourse. Further, prototypical argumentation must
minimally contain at least one of each of these units. That is, talk which is missing one of
these units does not constitute prototypical argumentation (that which combines both

argument] and argumentz). (See Preston (1993) for discussion of this point). For

example, talk in which there is a position and a dispute with no support by any participant
is more like childish bickering (“I’m right”, “No you’re not”, “yes I am”, “no you’re not”,
etc.). Talk in which there are positions and supports but no disputes does not constitute the

larger unit (speech activity) of argumentation, but rather involves instances of argumentsl.
Following Jackson and Jacobs, it is assumed here that argumentsl do not ordinarily occur

unless there is at least some anticipation of doubt by the speaker, and therefore the potential

for argumentz. Nonetheless, it is assumed, following Preston (1993), that the speech

activity of argumentation requires at least one explicit dispute. Talk in which positions are
put forward which are neither supported nor disputed does not constitute argumentation,
since in this case we would be forced to accept that any time a position was asserted in talk
argument was occurring — a counterintuitive claim since positions are endlessly asserted in
everyday talk and are seldom supported or disputed (e.g, “it’s going to rain later,”
“Grandparents spoil their grandchildren,” etc.).

Moreover, there is something unusual about speaking of a ‘position’ which is
neither supported nor disputed — it seems preferable to call this kind of assertion simply a

‘belief.’ This is because the constructs of position, support and dispute are interdependent

100

in that utterances become positions only when they supported or disputed (Preston, 1993).
Furthermore, an utterance may be a support only when there is some utterance to provide
support for, and an utterance may be classiﬁed/interpreted as a dispute only when there is
some utterance to be disputed. In other words, utterances acquire the status of position,
support or dispute only within the context of argumentation, and in terms of their semantic
relations with other utterances.

It might be claimed that there is ordinarily a one-to-one correspondence between an
utterance and the ideational unit it represents. Although this generally seems to be the case
for positions and disputes, there are many instances in which several or many utterances
may serve as a single support. For instance, it is often the case that participants in
argument tell narratives in order to support some position (van Dijk, 1987; Schiffrin,
1985). These narratives are composed of utterances none of which would independently
be called supports. Rather, it is their combined force which serves as the support of a
position. Similarly, participants often draw from personal experience in order to support
some position. Consider the following instance from the data in which K. appeals to

personal experience in order to support the position that White students are not as racist as

Black students:

(2)

1 K: = esPECially on this campus, I really think that
2 - a lot of people: - like White students - ARE: -

3 not AS racist as the Black students. Because - I'VE
4 had classes where Black students are just li:ke -

5 oh: - I can't believe- or even like where I work.

6 Where I work a lot of people come in, and White

7 people Black people whatever i- a lot of Asian

8 people. And a lot of Black people are just li:ke —

101

9 if you don't HAVE something, for them, like - if

10 people need to check out basketballs if you don't

11 have something for them they think that it's because
12 "Oh. I'm Black. that - I can't have it." You know?

K. begins her support by telling about some experience from her classes (in lines 3-5), but
aborts this and then provides a support which is derived from her experience at work.
Clearly, the entire stretch of talk from “Where I work...” (line 6) until the end of the
passage constitutes a single unit of support for the position that White students are not as
racist as Black students. That this support may be weak from a logical or rhetorical point of
view is beside the point — most signiﬁcantly, K. herself appears to intend for her personal
experience to serve as support for her position (note the use of ‘because’ before the ﬁrst
aborted support). While none of the utterances independently are supports (e. g, “Where I
work a lot of people come in,”), together they serve as a single unit of support. Thus, it is
incorrect to claim that there is necessarily a one-to-one correspondence between individual
utterances and ideational units (supports especially).

Just as there is no one-to-one correspondence between ideational units and
utterances, there is also no one-to-one mapping relation between particular ideational units
and particular linguistic forms. Although it seems to be the case that positions, supports
and disputes ordinarily occur in the form of declarative constructions with the illocutionary
force of assertives (van Eemeren and Grootendorst, 1984), other sentence types may serve
as positions, supports and disputes (see chap. 5 for discussion of this issue). For

example, sometimes rhetorical questions may serve as positions (from the data -— “Why

102

can’t Black people take out loans”), and imperatives may serve as disputes (e.g., “Dream
on 1”)

Accepting positions, supports and disputes as the underlying ideational units of
argumentation does not entail that other kinds of units do not also play a role in
argumentative discourse. Indeed, in van Eemeren and Grootendorst’s (1984) speech act
terms, ordinary argumentation involves different stages which are units in critical
discussion and are host to a variety of speech acts (smaller units) besides assertives (e. g.,
comrrrissives such as acceptances, and directives such as requests, questions and
challenges). In terms of conversational analysis, besides expanded sequences (of
adjacency pairs), there are other kinds of units of organization (structures) which operate
simultaneously in argumentation (such as the turn-taking and repair systems). In this way,
ordinary argumentative discourse is like narrative in that it minimally contains certain units

and is normally, but not necessarily, host to others:

Argumentative discourse Narrative

Minimally necessary positions, supports narrative clauses

units and disputes (complicating action)
(argumentative stage)

Other units larger: other stages larger: orientation, abstract, etc.
(e.g., confrontation stage)
smaller: other speech acts smaller: evaluative clauses

(e.g., questions, requests)

Furthermore, identifying positions, supports and disputes as the underlying units of

argumentation should not be considered the unwarranted imposition of constructs by

103

argument analysts. There is evidence within the discourse itself that participants in
arguments orient to their activity as argument and furthermore orient to utterances as
positions, supports, or disputes (see chap. 5 for detailed discussion). Arguments are
replete with metacommunicative messages which suggest this (e.g., “So you’re taking the
position that...,” “You can’t deny that...” “I believe that...,” “I can’t go for your
argument”, “There’s no way you can back that up...,” etc.).

However, treating positions, supports and disputes as the minimally necessary
underlying units of argument structure does not help in distinguishing ordinary argument
from pseudo-argument since both activities contain these units (and it is on this point that
they overlap). As suggested above, it will be shown that distinguishing these speech
activities on structural grounds requires reference not only to the ideational units, but also
to how these units are conﬁgured in the sequential organization of the discourse and how
they are distributed among the participants. In what follows, it will be demonstrated that
the data give evidence for two distinct conﬁgurations or patterns of ideational units, neither

of which is found in ordinary argument. The ﬁrst conﬁguration shows features of

argumentsl, whereas the second resembles but is clearly different from argumentsz. More

speciﬁcally, the ﬁrst pattern of ideational units involves the collaborative support of a single
position, whereas the second involves the collaborative dispute of an absent antagonist’s
imported position and support. It will be shown that these two conﬁgurations are recurrent
in and indicative of pseudo-argument, and allow the analyst as well as the participants to

identify the talk as such.

104

The aim of this section then is to present the most general characterization of the
possible conﬁgurations of ideational moves which is consistent with the kinds of data
which I am calling pseudo-argument. I follow van Eemeren and Grootendorst (1984) in
assuming that argumentative discourse can be reconstructed by an analyst in a way which
extracts positions, supports and disputes and ignores all other extraneous material in the
text. While such a reconstruction is subject to the interpretive skills of the analyst, one
important source of evidence available to the analyst in support of these interpretations is
the discourse of the participants themselves as they negotiate their points and react to each

other.

4.2.1. Collaborative argumentsl: making a case together
I call the ﬁrst conﬁguration “collaborative argumentl” (henceforth CA1) since
participants ally themselves in making an argument by collaboratively supporting some

mutually accepted position. Consider the following simple example:

(3)

A: OK. How bout, - how bout the uh - how bout
minority only scholarships. What do you - what
do you think about that.

C: Uh I think they're a great idea, in that-
that all of like - you know, if they get like a
two- suppose like - a Black individual, - uhm,
suppose they get like a - two point two or two
point five, they get a full scholarship,
whereas, you need like a - four point oh, if

0 you’re like a White dude to get a scholarship.

r-uoooqmurrb LUMP

11 B: Unmumn:

I
12 C: I think it's ridiculous.

105

13 B: Yes I think it’s ridiculous too. When I, -
14 was a freshman, I lived in McCollum ( )
15 McCollum. ALL the scholarships were for -—
16 African—American:s, minoritiezs, whatever: -
17 (Now), Asian-Pacific, - =

I

18 A: Asian—Pacific islanders .

In this excerpt, the position taken is that there should not be minority scholarships. C.’s
answer (in line 4) to A’s question is clearly sarcastic, and this interpretation is supported by
C.’s negative evaluation in line 12. What follows in his response (lines 4-10) is a support
for the position which implicitly appeals to the unfairness involved in the presumed
procedure involved in determining who receives scholarships — C. claims that Black
students need a far lower grade point average than do White students in order to qualify for
a scholarship. In line 13, B. shows agreement by seconding C’s evaluation (“Yes I think
it’s ridiculous too”), and then provides another support (in lines 13—16) for the main
position by arguing that “ALL” of the scholarships were awarded to minority students in
her dormitory, invoking the presumed unfairness involved in granting all scholarships to
certain individuals based solely on their minority status.

There are several features of this example worth noting. First, participants are in
agreement and appear to collaborate in supporting a single position. Second, the example
gives evidence of the modem racist ideology, since participants decry the granting of
privileges to Black students and other minorities by appealing to the unfairness of this
practice. Further, implicit in their argumentation is the assumption that the granting of such

privileges to minorities is no longer justiﬁed, presumably because minority groups now

106

stand on equal footing alongside Whites.

(4)

LUMP

Locoqcxur lb

Now consider once again the more complex example given earlier:

K: Why should they be coming to - a top ten uni- -

a big ten university, getting like ALL the money,

you know and I know people that get these checks, =
I

?: ( )

K: = and don't even need everything they need. And

the thing that I really don't like is how - I mean
this is something minor, but like - they get this
government money it's like (to pay for) books, but yet
they return the books at the end of the year and

they keep that money. They don’t have to return

it. You know what I mean? They're getting like =

I
?: Yeah

K: = all this free money and all this stuff when
like - here I am taking out loans that - the
interest is just adding up, and I'm going to
eventually have to pay it all BACK. You know, not
saying that I’m.WORried about it, cause I know -
that I'll be successful with MY job, but it’s just
not fair that

I
B: But they’re getting a free ride AND =

I
K: exACle

B: = they could be successful in THEIR job TOO. AND
plus they get this stupid point five increase in
their GRADES, so now they're going to be graduating
with- I

C: PLUS they have a better chance
of getting a job, so they're - going to be - more
successful ( )

I
B: So it' 3 just- 1- their advantages just keep

adding UP! Their - their advantages totally keep
adding UP. Their grade advantages,

I
C: Yeah that’ 5 true. They do have - more
time to relax because - they don' t have to WORK.
You know, like everybody else does to
pay off their loans.

107

I
36 B: They’ll be under less STRESS,

37 C: 'I‘hey- cause they're just getting money.

One ﬁrst obvious observation regarding this excerpt is that it is replete with
positions, and corresponding supports. Since there appear to be no disputes present in this
stretch of talk, it would be incorrect to claim that an argumentz (in the sense of ‘having an
argument’) was occurring. Clearly participants are actively involved in taking various
positions and justifying them with supports, and in this respect it would be fair to say that

arguments were being made (argumentsl). However, closer inspection reveals that unlike
ordinary argumentsl which are accomplished by only one speaker who provides both the
position and its support, the argument] put forward in this excerpt is the product of a

collaborative effort on the part of three participants who jointly provide supports for a

single position. In the reconstruction of the argument] set forth in this excerpt (repeated

from section 4.1.1.), it is apparent that more than one participant commonly contributes

support for the main position (repetition of the main position and supports is omitted):

(K) POS : Black students receive unfair advantages (“all the money” and stuff) (1,2)
(K) SUP 1 : They get checks that they don’t need (3,5)

(K) SUP 2 : Keep the money from the books bought by the government (6-11)

(K) SUP 3 : I have to take loans with interest, and pay it all back (13-17)

(B) SUP 4 : they could be successful too (22)

(B) SUP 5 : they get .5 increase in their grades (23-24)

(C) SUP 6 : better chance of getting a job (26-28)

(C) SUP 7 : more time to relax (32-35)

108

The reconstruction of positions and supports illustrates that this excerpt contains 7
distinct supports for the one position that African American students receive ‘unfair
advantages.’ All three participants provide at least one support. In lines 1-2, K. uses a
rhetorical question to indirectly take the position that Black students receive unfair
advantages in the form of “ALL the money.” She supports this position in lines 3 and 5 by
claiming that people (presumably Black students) receive checks that they do not need. This
is followed by another support in lines 6-11 which claims that Black students receive
money from the government for books but then exchange the books for cash when they are
done with them. In lines 13-17, K. provides a third support for her position by stating that
unlike Black students, she has to take out loans which she has to eventually pay back.

In line 22, B. joins in by providing another support for the position that Black
students receive unfair advantages by stating that Black students could also be successful in
their jobs, implying that they could afford to pay back loans as easily as anyone else. B.
then adds another support (in line 23-24) for the main position by claiming that Black
students automatically receive a small increase in their grades. In lines 26-28, C. joins the
argument with her own support — she claims that Black students have a better chance of
getting employment, presumably because of afﬁrmative action, a topic discussed earlier in
their conversation. Finally, C. provides another support (in lines 32-35) for the main
position that Black students receive unfair advantages by asserting that they have more time
to relax since they do not have to work to repay loans.

In sum, this excerpt contains seven supports for the single position that Black

109

students receive unfair advantages, and each of the three participants provides at least one
support. Although the reconstruction of positions and supports is subject to the
interpretation of the analyst (given the available linguistic evidence, knowledge of
conventional structures and discourse strategies, ethnographic evidence, etc.), it is argued
here that even a rough reconstruction of the excerpt is sufﬁcient to make clear certain
points. First, participants collaborate in supplying supports for the same position, and
therefore they are jointly building an argument. Second, there seem to be no constraints on
which participants may contribute supports for a position at any particular moment (other
than perhaps ordinary turn-taking rules). That is, any participant may chime in with
relevant support. We may incorporate these points into a structural schema for CA1 as

follows:

A: Position
X: Support 1, 2, 3..., n

This schematization attempts to represent the conﬁguration of ideational units in CA1 in
relation to the various interactional possibilities (who may contribute what and when). A
position in a CA1 is provided by a speciﬁc speaker (represented by A), and must be
followed by one or more supports, any of which may be provided by any participant
(represented by X), including the original speaker. In the data this pattern is found again
and again, and on this basis it seems justiﬁed to claim that CAls play an important role in

pseudo-argument.

110

Finally, note that the content of the previous excerpt reveals several characteristics
of the “modern racist ideology.” Recall that modern racism in the United States is
characterized by an ambivalence between strong egalitarian values and negative feelings and
attitudes towards Blacks and other minorities. This ambivalence helps determine an
ideology which essentially contends that discrimination and racism no longer exist and
therefore that any “advantages” given to African-Americans are unfair. In the previous
excerpt, participants disclaim the “advantages” granted to African-American students by
prestige granting institutions (in this case, both the government and the university).
Implicit throughout their discussion is the premise that such advantages are unfair and
undeserved.

It should be apparent that CAls are not found in ordinary argumentation, and are

distinct from ordinary argumentsl not in terms of the conﬁguration of ideational units, but
rather in terms of the interactional possibilities. Ordinary argumentsl may be schematized

as follows:

A: Position
A: Support 1, 2, 3..., n

Note that the only difference between this schema and the one for CA1 is the constraint in

ordinary argumentl that the position and the supports be provided by the same speaker (A).

It will be shown later that this difference has signiﬁcant linguistic implications (chap. 5)

and functional motivations (chap. 6).

111

4.2.2. Collaborative Argumentsz: Importing and disputing the antagonist

The second conﬁguration of ideational moves which recurs in the data is called

“collaborative argumentz” (henceforth CA2) because it resembles ordinary argumentz in

several respects. That is, in both types of argument, there is open opposition in that there
are disputes of opposing positions and opposing supports, and these disputes are

themselves supported. The main difference is that in ordinary argumentsz, arguments are

held between present participants who assume the roles of protagonist and antagonist with
regard to some position, whereas CA2$ are held not between present participants but rather
between the present participants (protagonists) on the one hand, and some absent
antagonist(s) on the other, whose positions and supports are imported into the talk and then
disputed. It will become clear after an example that it is more accurate to say that CA2$ are
simulated rather than actually held, since the participants must do the work of both the
protagonist and the antagonist. Before considering a rather complex example of CA2, let’s

examine the following simpler example from the data:

(5)
1 B: They totally have this- y- Exactly. And- You
2 know With Asians, you don't see Asians like =

I
3 K: I l
4 B: = pulling an attitude, and I don't think I- I
5 have like Asian friends and — I- I don't think I've
6 Ever seen an Asian - pull- you know and like- Black
7 people bring up like "well - you owe it to us
8 because - you know - you had us as slaves - =

I

9 K: Yeah but
10 how long:

11 B: = thousands of years ago, and blah blah blah:,

112

and you OWE it to us. Blah Blah" Well what about -
what about when we took like during what was it
WOrld war Two, when we took all the Japanese, and
stuck them in prison camps in California. I mean
what about THAT? You don't see Asian people still
all fired up about THAT, — I mean it is totally =

I

K: Yeah and if that IS the
case,
B: = WRONG but— Just- I mean it was a mistake in

the PAST, and it shouldn't - like — I don't know,
it- it- YEAH it does effect us, and YEAH it was
WRONG, but - it wasn't ME, it wasn't even my DAD.
It wasn't even my GRANDpa. It was like-

I
C: And they're-
now they're STILL - you know basically treated =

I
B: ( ) ago.

C: = not just equal, but they get a- the advantage

of - you know in a lot of situations so they can't
complain that they're not treated equal now.

This stretch of talk occurs prior to the talk of the previous excerpt (p.93-94) and

involves the same participants K., B., and C. At ﬁrst sight it is clear that the interactive

quality and prosody of the talk is the same in that participants are each actively and

enthusiastically involved (note for example, the overlaps and emphasis), and in addition,

the participants appear to be allied in their beliefs. Furthermore, the participants do not just

present their own positions and supports, but they also import the presumed argumentsl

(positions and supports) of an absent antagonist (in this case, “Black people”). For

example, using direct quotation (in lines 7-12), B. imports the presumed position of Black

people that White people owe them some compensation, accompanied by the presumed

support that this compensation is due because of the injustice of slavery.

113

Further, participants do not dispute each other, but rather they collaboratively
dispute the imported positions and the supports of the absent antagonist. For example, the
participants in this conversation collaboratively dispute the imported position and support
of the absent antagonist by arguing that although slavery was wrong, it occurred long ago,
and therefore White people now are not responsible. It is interesting to note that the
disputes of imported positions in CA2$ appear not to cooccur with the normal preference
markers characteristic of conversational disagreement (Pomerantz, 1984; Heritage, 1984;
Levinson, 1983), such as delays, mitigators, requests for clariﬁcation, etc. In addition, the
disputes do not subsequently involve deescalation or qualiﬁcation of positions in order to
achieve a working consensus, as is common in conversational disagreement. These facts
are predictable given the absence of the disputed party and the expectation of agreement on

the part of the issuer of the dispute. That is, with no expectation of resistance to the
dispute, preference markers and subsequent qualiﬁcations are unnecessary.1

Consideration of the conﬁguration of ideational units in relation to the interactional

possibilities suggests a schematization of CA2 as follows:

A: POSo / (SUPo l, 2, 3..., n)
X: DIS of POSo/(SUPo)
X: SUP l, 2, 3..., n of DIS

This schematization shows that an “other-authored” position (P080) and any number of

optional corresponding supports (SUPo 1 2,3...,n) may be presented by one participant

114

(A), and that subsequently any participant (X) may issue a dispute (DIS) of the “other-
author ” argument, followed by any number of supports (SUP 1,2,3...,n). In other
words, in CA2, a participant ﬁrst imports the position and optional supports of an absent
antagonist, which is then disputed by any participant where that dispute is followed by any
number of supports for it by any participant.

To illustrate this schematization of CA2, the last sample examined above may be
reconstructed as follows (where ‘r’ indicates repetition, and other-authored positions and

supports are underlined):

(B) E232 : “you owe it to us” (7)

(B) $1130.13 “because you had us as slaves...” (8)

(K) DIS/SUP 1 : It occurred long ago (9-10)

(B) SUP 2 : Japanese-Americans don’t ask for compensation (12-17)
(B) 51111111: Slavery is wrong (17, 20)

(B) SUP 1 r : Slavery was a mistake in the past (20-21)

(B) $111222 : It does affect us (22)

(B) $112111: Slavery was wrong (22-23)
(B) SUP 3 : White people living now are not responsible (23-24)
(C) SUP 4 : Black people have been compensated already (25-30)

In this example of CA2, the antagonist’s argument (position and support) is imported (in
lines 7-8), and then collaboratively disputed by the three participants. While the presumed
antagonist’s support is repeated twice by B. (this time in the explicit form of “Slavery is

wrong”), and an additional support is invoked (“it does affect us”) and agreed to, the value

115

of these supports as justiﬁcation for the position of the antagonist is disputed. Here,
although the content of the supports might in themselves be acceptable, they are not deemed
acceptable or relevant within the argument. Indeed, Preston (1993) notes that it is the case
that “...disputes of supports often deny their relevance rather than their content.” (p.205).

A few signiﬁcant points emerge from the reconstruction. First, participants do not
just present their own positions and supports, but they also present the presumed
arguments (positions and supports) of an absent antagonist (in this case, Black people).
Second, participants do not dispute each other, but rather they dispute the imported
positions and the supports of the absent antagonist.

The notion that the present participants in pseudo-argument may enter into
opposition with an absent antagonist may appear dubious until it is made clear that
opposition here is taken to involve not the disputation of interlocutors, but more accurately
the disputation of the positions or supports attributed to others. Dispute then is deﬁned
here as a relation of between utterances and only secondarily as a relation between the

authors of those utterances. This leaves open the possibility that the positions and supports

presented on behalf of absent parties may subsequently be disputed.2

It should be noted that ‘dispute’ is a crude label since it may involve disputed
propositional content, irnplicatures, or presuppositions of utterances (see Preston (1993)
for discussion). Also, supports may be accepted as true in themselves, but their use as
support for a position may be disputed. For example, while B. in the previous excerpt

admits that slavery is wrong, she disputes the use of the injustice of slavery as support for

116

the position that Black people should be compensated for the injustice.

Here again, in sample 5, there is ample evidence of the modern (American) racist
ideology. In particular, there is the belief among participants that racism and discrimination
are “in the past,” and thus it is claimed that African-Americans are unjustiﬁed in their
demands for compensation for past injustices. Furthermore, implicit in their discussion are
the beliefs that racism is consistent only with the tenets and practices of old-fashioned
racism (i.e., various stereotyped characteristics, support for segregation, support for acts
of open discrimination, etc.), and that racism so deﬁned is “wrong.” Finally, there is the
belief that “advantages” granted to African-Americans are unfair and undeserved.

Now consider a more complex example of CA2:

(6)
1 C: Do you remember - like
2 a few years back, there was like this big thing
3 where the Blacks were getting upset at- at Nike
4 corporation? because they didn't have enough =
I

5 B: No.
6 C: = vice-presidents that were Black? whereas the
7 three highest paid um people in the company, were
8 like Michael Jordan, Charles Barkley and some =

I
9 A: Michael Jordan

10 C: = other - some other basketball player dude. -
11 And Michael Jordan was making over ten million
12 dollars a year, for Nike?

I

13 B: And the vice-presidents are
14 probably making like a hundred thousand a year.
I
15 A: Well what =
I
16 C: ( ) =

17 A: = about-

48

49
50

117

C: = and the Black community was complaining and
whining- whining - oh ‘We d- we don't have enough
ah Black vice—presidents.’ - whereas the three- the
three- the three guys probably made- made more =

I

A: guys

B: [guys made the biggest-

C: = than anybody in the whole corporation

B: Right

C: It's got to-

B: Right

C: You know what I mean? It's riDICulous.

A: How bout the AT&T thing last year.

B: OKAY ( )

C: [How bout the O.J. SIMPson thing. I mean -

he's - he’s Black yet he made ALL this money off
the press. You know what I mean? He was like - you
know: famous reporter now: da da da - and now the
NAACP is getting upset that he's getting too much -
press, now that he murdered somebody?

B: It IS interesting, I mean - say who's another- =

I
A: Especially s-

B: = like Joe Montana. If Joe Montana w- supposedly
killed his wife, or ex-wife - it would be the same
big story - but aren’t we supposed to be talking =

C: I l

= about education,

I
Okay I'm sorry.( )

: Yes,
I

You probably need the conversation =

in that way.

: = that's why I brought up the: thing about -

B
C
A
B
I
A: Well that's- =
B
A
AT&T last year is- i- all th- the Black students =

51
52
53
54
55

56
57

58
59
60
61
62

63

65
66
67
68
69
7O
71
72
73
74

75
76

118

I
B: Oh the monkey thing?

A: = wanted - wanted to drop AT&T because of the -
because they had a monkey on th- on their
advertisement? You didn't know about that? - It's
ridiculous. It's so stupid.

B: But you know= if— if they had a kangaroo for
Austrazlia, nobody w-
I

A: Nobody would ( ) anything

I
C: See th- tha— tha- that’ s the
funny thing though. I mean - it' s as if the Blacks
think that everything about- everything is against
them. It’s like— you know what I mean?

I
A: y- well they’ re- it' 3 like
they' re constantly battling. If they wouldn' t =

I
C: exACle

bring issues up - then - nobody would think =

It’s like they WANT =

about it.

racism to exist. You know what I mean?

: So that they can get ahead.

((snicker))
[[ I don’t know.
[I It's so stu:pid.

[I It' s such a joke. I don’ t - I think it' s a

A

C

A

C

A

C: ExACle.
A

B

A

C:

joke, I think th- it' s a big joke.

((pause))

This excerpt is similar to the previous one in that participants are actively and

enthusiastically involved, and they appear to be allied with respect to the positions put

119

forward. In addition, the participants present positions and supports which they attribute to
some absent party and then dispute. This episode involves the presentation of three distinct
instances in which African-Americans are said to have complained about some apparent
injustice. In each instance, the imported complaint (in the form of an argument that some
injustice occurs) is collaboratively attacked by the participants. In the ﬁnal part of this
episode, conclusions are drawn by the participants to the effect that African-Americans
complain unnecessarily, and further these complaints are attributed to a larger aim (“to get
ahead”).

In the following reconstruction, there are four distinct parts, the ﬁrst three of which
are distinct CA25. The ﬁnal part seems to serve as a kind of resolution in that it
summarizes the points which fall out of the ﬁrst three parts (repetition has been omitted,

and other-authored positions and supports are underlined):

l) CA2 (“Nike”):

(C) 1123.01 : The Blacks were getting upset at Nike corporation (34)
(C) SLIEQI : Because they didn’t have enough vice-presidents that (4,6)
were Black.

(C) DIS / SUPa : The three highest paid people in the company were
Michael Jordan, Charles Barkley and some other
basketball player dude. And Michael Jordan was
making over ten million dollars a year. (7-8, 10—12)

(B) SUPb : And the vice-presidents are probably making like a
hundred thousand a year. (13-14)

120
2) CA2 (“NAACP”)

(C) EQSQZ: Now the NAACP is getting upset (35)

(C) $11202! because he’s getting too much press now that he (36)
murdered somebody. (he’s getting singled
out because he’s Black)

(B) DIS / SUP: If Joe Montana supposedly killed his ex-wife, it (39-40)

would be the same big story, (he would be singled
out too, so it’s not because O.J. is Black, it’s because
he is a famous football player).

3) CA2 (“AT&T”)

(A) P_QS_Q3: All the Black students wanted to drop AT&T (52)

(A) 51.11323: because they had a monkey on their advertisement? (53-54)

(B and A) DIS I SUP: But if they had a kangaroo for Austrazlia, nobody (56-58)
would say anything.

4) Conclusion/Resolution

(C) POS4: The Blacks think that everything is against them. (60-62)

(A) POSS: They’re constantly battling. (63-64)

(A) POS6: If they wouldn’t bring issues up then nobody would (64, 66,68)
think about it. ((There wouldn 't be a problem then.))

(C) POS7: They WANT racism to exist. (67, 69)

(A) SUP of POS7: So that they can get ahead. (70)

First, let’s consider the ﬁrst three CA2s. It is apparent from the reconstruction that
three cases are presented in which African-Americans have supposedly complained about
something. We might gloss the three parts as “Nike,” “NAACP,” and “AT&T.” In each

case, the supposed position and support (argumentl) of the absent antagonist are

presented, and then disputed. In the ﬁrst case, C. imports the antagonist’s position that

“the Blacks were getting upset at Nike corporation,” which is followed by the antagonist’s

121

purported support; “Because they didn’t have enough vice-presidents that were Black.”
After presenting the absent antagonist’s argument, C. attempts to dispute it by pointing out
that there are very high-paid employees (professional basketball players) working for Nike
(the obvious weakness of C.’s support for his dispute is not at issue here). After this, B.
contributes a second support for the dispute by saying that the vice-presidents at Nike do
not make nearly as much as the professional basketball players.

In the second CA2, C. imports the antagonist’s position that “the NAACP is getting
upset,” followed by a support; “that he’s (O.J. Simpson) getting too much - press, now
that he murdered somebody?” In other words, C. presumes that the NAACP (National
Association for the Advancement of Colored People) became upset because O.J. Simpson
was being singled out mainly because he is Black. This interpretation is strengthened by
B’s support of the dispute of the imported argument; “If Joe Montana supposedly killed his
ex-wife, it would be the same big story.” In other words, Joe Montana (a White
professional football player) would be singled out too, and so it’s not because O.J.
Simpson is Black, but rather it is because he is a famous football player.

In the third case, it is A. who initiates the CA2 by presenting the position of the
antagonist that; “All the Black students wanted to drop AT&T,” which is followed by the
presumed support; “because they had a monkey on their advertisement?.” Here again, B.
enters to dispute the absent antagonist’s argument; “But if they had a kangaroo for
Austra:1ia, nobody would say anything.” In other words, B.’s claim is that there was

nothing offensive intended by associating Africa with a monkey in the same way that there

122

is nothing offensive in associating Australia with kangaroos, and therefore the complaints
by Black students were unfounded. (Again, the weakness and offensiveness of B’s
support/dispute is irrelevant to the point at issue here.)

Therefore, in the excerpt above there are three distinct CA2$ which have the same

structure:

A: m and sun: (argumentlo)
X: DIS/SUP of (argumentlo)

In each case, the position and support (argumentj) of the absent antagonist is presented and
is followed by a dispute of the imported argument] with some support for the dispute. In

addition, the three CA2$ have similar content which relates them -— all three involve
instances drawn from media stories in which African-Americans complained or were upset

about mistreatment (not enough executives, too much press for O.J. Simpson, vicious
stereotypes in the media).3 By disputing the imported arguments, the participants in each

case attempt to remove the justiﬁcation for complaint.

Given this connection between the three CA28, it may be argued further that each of
the three CA2$ themselves serve as supports for a higher level position which is implicit
from the start — that Black people complain unnecessarily. This interpretation is
conﬁrmed in the fourth part of the excerpt in which a string of conclusions appear to be

drawn from the three CA25:

123

(7)
(C) POS4: The Blacks think that everything is against them. (60-62)
(A) POSS: They’re constantly battling. (63-64)
(A) POS6: If they wouldn’t bring issues up then nobody would (64, 66,68)

think about it. ((There wouldn ’t be a problem then.))

With the justiﬁcation for complaint cleared away in each case, an underlying motive is

consequently attributed to the antagonist — Black people complain because they want to get

ahead:

(8)

(C) POS7: They WANT racism to exist. (67, 69)
(A) SUP of P087 : So that they can get ahead. (70)

In this way, the three CA2s are connected in a chain and serve as supports for a
higher level position. Van Eemeren and Grootendorst (1984) acknowledge that within their
speech act model, arguments may be combined to form larger structures; “One
argumentative act complex may support or complement another, and it may be only
together that they constitute a complete attempt at justiﬁcation or refutation” (p.76).
Similarly, it is argued here that CA2s may be embedded within even larger argument
structures and are not necessarily ends in themselves. That is, CA28 may serve as
intermediate level argument structures (i.e., supports) which are combined within even
larger argument units (such as larger CAls). This may be illustrated with reference to the

previous excerpt as follows:

124

CA1 = A: POS
X: SUP 1 (CA2),
X: SUP 2 (CA2),
X: SUP 3 (CA2).

In this case, X may represent more than one participant in that a single support (a CA2)
may be the product of more than one person.

Finally, it should be noted that the content of the previous excerpt serves as further
evidence that pseudo-argumentative discourse carries beliefs and attitudes associated with
the modern racist ideology. First, by showing that the various complaints by African-
Americans about discrimination are unfounded, it is concluded by the participants that these
complaints are actually tactics employed to gain unfair advantages (“so they can get
ahead”). Second, implicit in the discussion is the assumption that racism no longer exists
in American society.

In each of the three CA25 of the previous excerpt, more than one participant makes
a contribution, which serves as further evidence for the “collaborative” nature of CA2.
However it does not seem to be necessary that CA2$ be made up of the contributions of
more than one individual. That is, one participant may carry out the work of an entire

CA2, as in the following example:

125

(8)

A: I don’t think it’s being racist, I think they’re
getting a lot more scholarships, and - if they need
them, fine. But maybe have some more- I don't even
know if they have work study? Or if they all — if— a
requirement had been a scholarship? Or financial aid
( ) work study ( )? Maybe they should.be working,
and not just have it all handed to them. - And they
kind of think like they - like they - desERVE it. -
Because - THEY grew up in the inner-city, and WE grew
up in West Bloomfield and Farmington Hills or
whatever, So they- they deserve it. Cause our mommy
and daddies could afFORD it. And they don't know. And
13 I think that THEY generalize a lot too.

HKOmQO‘UlIwaH
O

Hra
aha

Here, beginning with line 7 (“And they kind of think...”) A. seems to be
representing both the position and support of an absent antagonist. She introduces
the ‘other-authored position’ (in lines 7-8) with “And they kind of think... ”, an
initial indication that what follows is a representation of someone else’s belief. This
is followed by the corresponding ‘other-authored support’ (in lines 9-11);
“Because THEY grew up in the inner-city....” Evidence that A. does not subscribe
to the support is that she indirectly disputes it afterwards (in lines 12-13) with the
criticism that such claims overgeneralize and therefore are not convincing.

Further evidence that A. is representing the position and support of an absent
antagonist comes in lines 11-12 where she paraphrases and makes more explicit the
other-authored support of lines 9-11 by saying “cause our mommy and daddies
could afFORD it.” It is unlikely that A. would refer to the White parents of West
Bloomﬁeld and Farmington Hills as “mommy and daddies” if she were the author

of the support. It is more plausible that A. is imitating what is in her conception the

126

mockery of the imagined antagonist who presumably views White college students
from West Bloomﬁeld and Farmington Hills as spoiled children who are pampered
by their wealthy parents. Once again, in this example, a speaker imports the
position and support of an absent antagonist, and then argues against it. In

addition, this example illustrates that a CA2 may be the product of a single speaker.

4.3. Pseudo-argument versus ordinary argument.

Although positions, supports and disputes occur in ordinary arguments, neither
CA1 nor CA2 occurs in ordinary arguments. In pseudo-argument, participants are busy
collaborating either in building cases for their own positions (CA1), or else in identifying

and destroying the antagonists’ argumentsl (CA2). These collaborative activities are

analogous to what happens in ordinary argument where an individual presents his or her

own argumentsl (position and supports) and attempts to destroy his or her opponent’s
arguments (argumentz); “... when two individuals dispute each other’s positions in

oppositional argument, their talk is directed not only to making their own points but also to
challenging their interlocutor’s points” (Schiffrin, 1985, p.45). In addition, pseudo-
argument is superﬁcially similar to ordinary argument in that it contains positions, supports
and disputes in particular sequences.

An enlightening exercise is to take an excerpt from pseudo-argument and

transform it into an ordinary argument by imagining an actual present antagonist

127

who presents his/her own positions and supports and by conﬂating the allied

participants into a single individual. This sort of modiﬁcation requires little more

than changing pronouns — the structure(s) remains intact. Consider a modiﬁed

version of the earlier reconstruction:

1) Ordinary argumentz (“Nike”)

E251 (A: antagonist):
$1.121:

DIS / SUPa (P: protagonist):

SUPb (P):

2) Ordinary argumentz (“NAACP”)

11252 (A):
51122:

DIS/SUP (P):

3) Ordinary argumentz (“AT&T”)

E253 (A):
$11123:

DIS/SUP (P):

We (African-Americans) are upset at Nike
corporation

Because they didn’t have enough vice-presidents that
were Black.

The three highest paid people in the company were
Michael Jordan, Charles Barkley and some other
basketball player dude. And Michael Jordan was
making over ten million dollars a year.

And the vice-presidents are probably making like a
hundred thousand a year.

Now we (NAACP) are getting upset

because he’s (O.J. Simpson is) getting too much
press now that be murdered somebody. (he’s getting
singled out because he’s Black)

If Joe Montana supposedly killed his ex-wife, it
would be the same big story, (he would be singled
out too, so it’s not because O.J. is Black, it’s
because he is a famous football player).

We wanted to drop AT&T
because they had a monkey on their advertisement

But if you had a kangaroo for Austra:1ia, nobody
would say anything.

4) Conclusion/Resolution
Position4 (P):
Position5 (P):
Position6 (P):
Position7 (P):

Support of Position7 (P):

128

You think that everything is against you.

You are constantly battling.

If you wouldn’t bring issues up then nobody would
think about it. ((There wouldn ’t be a problem then.))
You WANT racism to exist.

So that you can get ahead.

This modiﬁed version of the earlier reconstruction illustrates the structural closeness

between pseudo-argument and ordinary argument in that it is possible to change only the

identity of the participants in order to perform the transformation from one type of talk to

the other. This points out again that PA and 0A differ more in terms of interactional

framework (who says what) than in terms of the structure of the text.

There are however, several sharp differences between pseudo-argument and

ordinary argument which have been revealed in the analysis of the data so far;

i) In PA, all present participants are allied throughout. In OA, at least two

participants are in open opposition.

ii) In PA, participants all assume the role of protagonist. In OA, at least two

participants assume the roles of protagonist and antagonist with regard to

some position.

iii) In PA, participants collaborate in supporting certain positions. In OA,

participants do not support each other’s positions.

129

iv) In PA, participants must import the arguments of the absent antagonist. In OA,
the antagonist is present to speak for himself or herself.
v) In PA, antagonists do not have opportunities for rebuttal after the protagonists’

dispute. In OA, participants have ample opportunity for rebuttal.

4.4. Summary

This chapter has shown that the data analyzed do not conform to two prominent
discourse-analytic models of argument, van Eemeren and Grootendorst’s speech act
approach (1984) and Jackson and Jacob’s conversation analytic model (1982, 1986). This
was taken as an indication that in the data something other than ordinary argumentation is
occurring. The chapter then went on to show that from a structuralist perspective, one may
identify and deﬁne mutually dependent underlying units (positions, supports and disputes)
within the type of discourse that I call pseudo-argument. Furthermore it was argued that
there are two recurrent conﬁgurations of ideational units (CAls and CA28) which
themselves may serve as intermediate level units within pseudo-argument. It is claimed that
no matter how complex the argumentation may be, it will be composed of chains of
collaborative arguments (1 and 2) and ultimately of positions, supports and disputes.
Further evidence for these structural claims will be provided in the next chapter in which it
will be demonstrated that at the surface linguistic level there are signals by participants
which show their awareness of and orientation to such underlying structures. The ﬁnal

chapter will argue that these structures have a functional motivation.

Chapter 5

LINGUISTIC PROPERTIES OF PSEUDO-ARGUMENT

5.0. Introduction.

It was argued in chapter 2 that the analysis of surface linguistic features in discourse
is a valuable and necessary pursuit for discourse analysis. Speciﬁc linguistic forms are
vital resources for participants in the creation and interpretation of meaningful, coherent
discourse, and in addition are signiﬁcant clues to the analyst which help in understanding

the structure and function of a text. To quote Stubbs (1983) once again:

...at least in some discourse types... a large proportion of the talk comprises
precisely such superﬁcial indices of underlying organization... it is precisely such
items which are of interest to the discourse analyst, since they are the items which
indicate the underlying structure of the discourse or the underlying functions of
individual utterances. (178)

Just as certain surface linguistic features help analysts to identify underlying structures and
functions, they are also produced by and oriented to by participants themselves as the
discourse unfolds. It is assumed in this work that certain linguistic features are mainly
used as signals given by speakers and interpreted by hearers which point to the underlying

organization of discourse and guide the interpretations of individual utterances.

130

131

For example, Labov’s (1972) work on narrative (reviewed in chap. 2) showed that
the underlying units of narrative (orientation, complicating action, evaluation) could be
identiﬁed by participants (and by analysts) in terms of surface syntax. According to
Labov, the syntax of utterances in narratives clues listeners in to which part of the story
they are listening to and therefore is a signiﬁcant factor in the proper interpretation of the
narrative as a whole. Schiffrin’s work on discourse markers (1987) illustrates that such
surface linguistic features operate mainly in order to bracket units in discourse by selecting
and displaying the meaning and structural relations between utterances. Discourse markers
locate utterances on different discourse planes and play a crucial role in the display and
interpretation of discoursal organization and structure. Finally, the work of Gumperz
(1982) was brieﬂy discussed in chapter 2 where it was shown that certain surface linguistic
features called “contextualization cues” are used by speakers to signal the interpretive
framework (or “contextual presuppositions”) in which utterances are meant to be
understood. When interlocutors do not share such marginal (but crucial) linguistic features,
miscommunication often is the result.

In the analysis to follow in this chapter, I follow Labov, Schiffrin and Gumperz in
paying heed to certain surface linguistic features in discourse. It is assumed that such
features provide salient evidence in support of more abstract underlying structures posited
in analysis. Perhaps more importantly, such features are vital to the participants themselves
in the construction and interpretation of the ongoing talk as it unfolds (in the ways

elaborated by Labov, Schiffrin and Gumperz). Before getting to the linguistic analysis of

132
pseudo-argument however, some preliminary discussion concerning the linguistic analysis

of ordinary argument in earlier studies is appropriate.

5.1. Linguistic analysis of ordinary argumentation.

It appears that little discourse-analytic research on argumentation theory has been
devoted to analysis of low-level linguistic features (syntactic, prosodic, lexical, etc.).
Certainly the works of van Eemeren and Grootendorst (1984) and Jackson and Jacobs
(1982, 1986) make passing reference to some linguistic feature or another (e.g, rhetorical
questions by van Eemeren and Grootendorst), but their observations amount to no more
than side comments which do not strongly impinge on their main arguments. The work on
argumentation by Schiffrin (1984, 1985, 1987 , and 1994), on the other hand, has been
consistently linguistic in orientation, and therefore merits some review here. The brief
review of Schiffrin’s work to follow will cover a variety of linguistic features which are
common in ordinary argumentation (certain discourse markers, meta-talk, etc.), and should
serve to indicate the kind of analysis which will follow later in this chapter with regard to
pseudo-argument. Along the way, other preliminary points will be made which will be

germane to the linguistic analysis of PA.

5.1.1. Schiffrin (1985).
This work is mainly focused on identifying and describing characteristics and
organizational principles underlying argumentative discourse. Brieﬂy, Schiffrin ﬁrst

identiﬁes two related but distinct subtypes of argumentation; rhetorical argument and

133

oppositional argument. These subtypes appear to match O’Keefe’s distinction between

argument] and argumentz in that in rhetorical argument “. .. a speaker presents an intact

monologue supporting a disputable position,” whereas in oppositional argument “... one or
more speakers support openly disputed positions” (p.37). One of Schiffrin’s concerns is
to show the ways in which the organizational properties of rhetorical and oppositional
arguments are instantiated linguistically. This involves locating those features which are
used to bracket units within the discourse (positions, supports and disputes), and to show
how particular units are themselves fashioned linguistically.

Schiffrin notes that the location of a unit may itself serve as a bracket in a rhetorical
argument — paraphrases of positions at the beginning and end of a rhetorical argument
locate the main point of the argument at the outer boundaries, while subordinate
information (support) is enclosed within those boundaries (p.39). Often supports within a
rhetorical argument are separated by an intervening paraphrase of the main position which
“... reestablishes what is being supported by both the preceding and the upcoming
discourse” (p.40). In addition to the location of an utterance, there are structural indicators
(discourse markers — see below) which help to differentiate units within argument. For
example, coordinating conjunctions (“and,” “so,” “but”) often introduce positions, and
subordinating conjunctions (“because,” “like”) often introduce supports (p.39). Schiffrin
also mentions certain semantic indicators which help differentiate the internal units of
argument (much the same way as Labov did for narrative). For example, there may be

constellations of semantic indicators which establish a position as a generalization; i.e., the

134

indeﬁnite pronoun “you,” the stative verbs “be” and “seem,” and use of the present tense.
These features establish a generalization, according to Schiffrin, because they do not
restrict events and states to a limited time or situational frame.

Schiffrin also mentions other linguistic devices which play various roles in
rhetorical and oppositional argument. For example, intensiﬁers (like “really”) and
comparatives are often used by speakers to signal the potential disputability of some
position. That is, such devices signal the speaker’s commitment to some position, but at
the same time signal the openness of that position to attack. In oppositional argument,
individuals often make ‘ritual displays’ of cooperation just before attacking another’s
position (e.g., “all right”, or “yes, but...” (an example also discussed in Pomerantz

(1984)).

5.1.2. Schiffrin (1980)

This work has to do with the various roles of ‘meta-talk’ in discourse. Roughly
speaking, meta-talk is language which has as its object some aspect of language. (See
Schiffrin for discussion of the problems in ﬁnding the boundaries between meta-talk and
other language functions, pp. 199-202.) Schiffrin’s main claim is that meta-linguistic
clauses serve as organization brackets which indicate the external and internal boundaries of
a discourse unit. Consider the following simple example (Schiffrin, 1980, p.207) in which
a metalinguistic clause both initiates and terminates an assertion (a constituent of a larger

discourse unit):

135

(1)

PM: uh... ((T'HERE’S ONLY ONE THING» I’m prejudiced on...is
that’s when...they mingle.
Y’know marriage and all.
((THAT’S THE ONLY THING» I’m against.

In addition to separating units of discourse, metalinguistic clauses may also ‘label’ a
particular slot in the discourse (i.e., “The point is,” “One reason is that,” “for example ,”
etc.); “...meta-talk has an organizational function; that is, it acts as a discourse bracket that
initiates or terminates slots in the discourse, providing an environment in which to label the
material in the discourse” (p. 216).

To illustrate her points, Schiffrin analyses meta-talk which occurs in (data
involving) argumentation. She also examines ‘explanations,’ which appear to function the
same as rhetorical arguments do in her later work; “Explanations...typically consist of an
assertion that is backed by evidence and/or validated by reasons” (p.210). Here is a sample
of the meta-linguistic clauses which Schiffrin provides and analyses in terms of their

functions in argumentative discourse (others not necessarily related to argument are ignored

here):
(2)
That’s his opinion I’m not arguing that
I ﬁgure it this way I’m not gonna say
That’s my opinion You say that’s wrong?
I’ll tell you why I don’t agree with that
For instance, I’m not arguing...
Another thing What the hell do you mean
That is not the point That’s right
The point is That’s not true

Don’t tell me... I hate to say these things but it’s true

136

At fnst sight, such naturally occurring metalinguistic clauses serve as compelling evidence
that participants in argumentation orient to their talk as such. This is especially true in those
cases where slots are labeled (i.e., “The point is...,” “I don’t agree with that,” “I’m not
arguing that...”, etc.), since the labels indicate the participants’ own understanding of the
roles of their utterances in the discourse. This touches on a point mentioned earlier that
individuals orient themselves to the type of discourse which they are engaged in for both
organizational and interpretative purposes. Thus, meta-linguistic clauses serve as evidence
that the structures posited by the analyst are real for participants as well.

Having introduced arguments and explanations as discourse units, Schiffrin
demonstrates that such units and their internal constituents are often bracketed by
metalinguistic clauses. For example, supports (‘reasons’ in Schiffrin’s early terminology)
may be bracketed from positions (‘assertions’) by metalinguistic clauses such as “The

9’ 661’

reason for it..., 11 tell you why...,” and “For instance...” Supports for the same
position may also be bracketed from each other by metalinguistic clauses such as “Another
thing is...” Such clauses contribute to an argument’s organization by ordering the supports
and making clear that more than one bit of evidence is being offered. Positions may also be
bracketed with such metalinguistic expressions as; “The point is...,” “In my opinion...”,
etc.

When we turn from the monologic rhetorical argument (‘explanation’ for Schiffrin

(1980)) to dialogic oppositional argument, we see that in addition to bracketing their units

or constituents of units of their own talk, meta-talk may be used to bracket the talk of an

137

interlocutor. According to Schiffrin (1980, p.218), this stems from the fact that
explanations, like narratives, are units which act as requests for evaluation. It is often the
case that explanations are interrupted at critical junctures by challenging evaluations by a

listener, which function as evaluative brackets. Schiffrin provides some examples from her

data:

(3)
I don’t agree with that
That is not the point
That’s off

That’s his opinion
You say that’s wrong?

For instance, by saying “I don’t agree with that” in response to an interlocutor’s position,
the speaker brackets that position as a disputable item, and at the same time displays a
(negative) evaluation of it. Thus, metalinguistic brackets directed to an interlocutor’s talk
are evaluative, and according to Schiffrin (p.221); “...in an argument, meta-talk that is
directed to an interlocutor’s talk acts as an evaluative bracket, rather than an organizational
bracket.”

However, the line between organizational and evaluative brackets is not so clear,
because meta-talk directed to a speaker’s own talk can also be evaluative (p.224). For
example, the metalinguistic construction “That’s my opinion” not only serves as a terminal
organization bracket, but also displays the speaker’s evaluation of her own talk in that it
acknowledges the disputability of the position put forward while at the same time expresses

a personal commitment to it. In this way metalinguistic constructions may have an

138

evaluative emotive force which allows the speaker to focus on both the code itself as well
as the stance taken by the speaker towards what is said. Later in this chapter, much more
will be said about the evaluative function of certain metalinguistic expressions (speciﬁcally
those evaluative tags like “That’s right” and “Exactly,” which occur frequently in pseudo-

argument).

5.1.3. Schiffrin (1987)

Schiffrin’s work on discourse markers was discussed in chapter 2 where it was
shown that discourse markers in general bracket units in discourse by selecting and
displaying the meaning and structural relations between utterances. Schifﬁin again
illustrates her points by examining data involving argumentative discourse, which is host to
a variety of discourse markers such as “but,” “so,” “because,” “and,” etc. (These are, of
course, not limited to argumentative discourse.) She determines that these markers are used
to select and display particular structural relations between utterances and create coherence
by integrating the discourse at different planes.

For example, Schiffrin discusses the marker “but,” the main function of which is to
locate an utterance on the ideational plane of discourse by marking an upcoming unit as a
contrasting idea or action (p. 152). The utterance prefaced by “but” may contrast with either
the actual semantic content of a prior utterance, or else with implicit expectations by the
speaker or bearer. “But” may also function at the exchange structure plane in that it may be
used to mark speaker-retum to a point which has been interrupted or postponed due to

competing conversational demands.

139

The conjunction “an ” primarily coordinates idea units at the ideational level, such
that its presence signals that an upcoming unit is coordinate in structure to a prior unit
(p.141). “And” is the most frequent connector of idea units (p.129), perhaps in part
because it may occur in many of the same environments as other connectors such as “but”
and “so.” Although “and” has no semantic content, its distribution within a text may create
syntagmatic contrasts which may signal new idea units. For example, “and” and “zero” are
options in the coordination of utterances, but their distribution within a string of utterances
may serve to mark boundaries between idea units. Consider Schiffrin’s example in which
the speaker lists the attempts of anti-Serrrites to annihilate the Jewish people and then shifts

to a generalization derived from the list:

(4)
a. They threw us in the ﬁre,
b. they shot us,
c. they killed us,
d. they put is in the gas chambers,
e. and they couldn’t do it.

In this case, the events in the list are coordinated by “zero,” but when the speaker shifts to a
position, it is coordinated with “and”; “...once a textual regularity has been developed —
even if it is developed for only a short sequence — a new idea unit can be introduced by a
change from that textual norm” (Schiffrin, 1987, p. 13 1). Schiffrin goes on to illustrate the
many ways in which “and” is used as a marker of functionally differentiated idea units at
different levels of discourse structure. In terms of interaction, “and” marks speaker

continuation (simultaneously to marking idea units) in that it signals that what is being said

140

is a continuation of what preceded it.

The subordinating conjunction “because,” according to Schiffrin, is a marker of
subordinate idea units. She deﬁnes subordinate material functionally as “..that which has a
secondary role in relation to a more encompassing focus of joint attention and activity,” and
referentially as “...that which is not as relevant in and of itself, as it is to a more global
topic of talk” (p.191). In argumentation, this subordinate material will be support for some
position. Complementary to “because” is “so,” which is a marker of main idea units.

Consider an example given by Schiffrin:

(5)

Zelda: a. Well we were going up t’see uh...my - our son tonight,
b. but we’re not
c. cause the younger one’s gonna come for dinner
(1. cause he’s working in the neighborhood.
e. So that’s out.

In this example, Zelda is explaining why she and her husband are not going to see their
older son. Her ﬁrst reason (support) is prefaced by “because” which is followed an
embedded reason (for the reason in c.) which is also prefaced by “because.” The return to

the main point is prefaced by “so” (“that” refers to the ruined plan).

In sum, Schiffrin’s (1980, 1985, 1987) works have illustrated that ordinary
argument is laden with linguistic features which have an organizational function in that they

serve to bracket underlying structural units. These include certain discourse markers,

141

metalinguistic constructions, intensiﬁers, paraphrase, and constellations of semantic
indicators among others. This review was not intended to be comprehensive, but rather
was aimed to show that argumentation, like any speech activity, is host to certain surface
linguistic features which help to organize the discourse by pointing to underlying structural
units in the service of leading hearers to accurate interpretations of utterances. Such
features are clues available to the analyst who is interested in understanding the
organization of different types of discourse. To ignore the linguistic composition of a

discourse is to miss out on important evidence of its structure and function.

5.2. Linguistic realization of pseudo-argument

Having reviewed several studies (Schiffrin 1980, 1985, 1987) which treat aspects
of the linguistic makeup of what I am calling ordinary argumentation (OA), we are ready to
embark on a similar kind of linguistic analysis of pseudo—argumentative discourse. While a
comprehensive analysis cannot be undertaken here, a variety of signiﬁcant and interesting
features will be provided which together should make a convincing case that pseudo-
argument is distinct from ordinary argument on linguistic grounds. In addition, it will
become clear that such features are crucial to the participants themselves in the construction
and interpretation of the talk.

In the sections which follow, the analysis will focus on several linguistic features

which recur and which contribute to signaling and interpreting structure in pseudo-

argument.1 It will become evident that these features contribute in important ways to

142

participants’ perceptions of the speech activity that they are involved in. These features of
pseudo-argument fall into two general categories -— those which cluster around

collaborative argumentsl and signal endorsement and agreement, and those which cluster
around collaborative argumentsz and signal the importation and dispute of an absent
antagonist. The linguistic phenomena to be investigated include repetition of particular
linguistic units (e.g., pronouns, lexical items, syntactic constructions, etc.), discourse

markers and meta-talk, among others. In sum, at least the following general points will

emerge from the analysis:

1. Since the features to be analyzed are speciﬁc to PA and do not usually occur in OA,
they serve as further evidence that PA is distinct from CA.

ii. The features have an organizational function in that they are used by speakers in PA
in part to signal underlying structures, and to guide hearers in their interpretations
of utterances within those structures.

iii. The features are important contextualization cues which indicate the interpretative

framework in which utterances are to be properly interpreted.

5.2. 1. Some features of collaborative argumentl

In chapter 4 it was shown that there are two distinct recurrent conﬁgurations of

ideational units in pseudo-argument. The ﬁrst was called collaborative argumentl because

it is usually forged on behalf of two or more participants, and because it has features of

143
arguments] in that a position is given support (an argument is ‘made’). In CA1 participants
supply supports for some position even though they are in accord from the start and do not
apparently anticipate opposition from each other. It was then proposed that based on the

data, CA1 has the following structure:

(6)
A: Position 1

X: Support 1a, b, c..., n

This schematization represents the idea that any participant in collaborative argument may
contribute any number of supports for some position.

Given this proposed underlying structure, we may now ask what are the sorts of
surface linguistic features which normally co-occur with it. Further, how do these features
give evidence that such an underlying structure is real for participants (in the sense of being
something which they orient to and are aware of)? In the analysis which follows in this
section, various linguistic phenomena will be discussed which appear to cooccur with CA1
and which seem to provide some evidence for its structure. These phenomena serve the
purpose of bracketing and coordinating distinct supports and in addition serve as signals of
agreement and endorsement from participants. First it will be shown how the use of
agreement tags (like “exactly,”) is productive in endorsing and bracketing general positions

in CAls. Second, I will argue that the use of repetition serves similar purposes.

144

5.2.1.1. Endorsement and alliance

Not surprisingly, much of the linguistic work that goes into collaborative arguments
involves communicating a positive evaluation of the contributions of others — there is an
abundance of signals of endorsement and agreement from participants. Stubbs (1983)
deﬁnes “endorsement” as “...a move which backs up, adds weight to, approves, upholds,
chimes in with, ratiﬁes or recognizes as relevant previous talk” (p. 190). Further Stubbs
notes that the function of endorsements is “...not merely to convey propositional content,
but to take up an alignment, to make a commitment to a position, to claim fellowship or
form alliances” (p. 188). As will be discussed in chapter 6, the abundance of endorsement
markers in pseudo-argument could result from participants’ perceptions of themselves as

being ‘on the same side’ and their work to reassure and solidify their ideological alliance.

5.2.1 . 1.1. The use of “exactly” — bracketing and endorsing general positions. One sort
of linguistic signal of endorsement which occurs frequently in the data is the use of
agreement markers such as “exactly!,” “right!” and “that’s true.” The following analysis
will focus on the use of “exactly” in the data where it will be shown that this agreement
marker is used to endorse and highlight the salience of a particular contribution from
another participant in pseudo-argument. In her work on ‘meta—talk,’ Schiffrin (1980)
discusses the use of what she calls ‘evaluative tags.’ In addition to their bracketing
function, these metacommunicative tags display a speaker’s alignment towards the talk of

an interlocutor; “Phrases such as ((T'HAT’S RIGHT» are often used to indicate a positive

145

evaluation of what an interlocutor has been saying, for example, to show appreciation,
agreement, conﬁrmation” (p.225). Also, Schiffrin argues that just as evaluation in
narrative serves to tell what the story is ‘really about,’ evaluative tags like “that’s right”
may serve to bracket a more general proposition or point conveyed by an interlocutor. In
the same way, “exactly” seems to be used to indicate a positive evaluation (an endorsement)
and at the same time to bracket a more general proposition (position or support) of an
interlocutor. Consider once again an extended version of an excerpt examined earlier

involving participants K, B and C:

(7)

K: Why should they be coming to - a top ten uni- -

a big ten university, getting like ALL the money,

you know and I know people that get these checks, =
I

?: I l

K: = and don't even need everything they need. And
the thing that I really don't like is how - I mean
this is something minor, but like - they get this
government money it's like paper books, but yet
they return the books at the end of the year and
they keep that money. They don't have to return

11 it. You know what I mean? They're getting like =

1h WNH

r—wocoqcxur
O

12 ? : Yeah

13 K: = all this free money and all this stuff when
14 like - here I am taking out loans that - the
15 interest is just adding up, and I’m going to
16 eventually have to pay it all BACK. You know, not
17 saying that I’m.WORried about it, cause I know -
18 that I'll be successful with MY job, but it’s just
19 not fair that

I
20 B: But they’re getting a free ride AND =

I
2 1 —> K : exACle

22 B: = they could be successful in THEIR job TOO. AND
23 plus they get this stupid point five increase in

146

24 their GRADES, so now they're going to be graduating
25 with- I

26 C: PLUS they have a better chance
27 of getting a job, so they're - going to be - more
28 successful ( )

I
29 B: So it' s just- i- their advantages just keep

30 adding UP! Their - their advantages totally keep
31 adding UP. Their grade advantages ,

I
32 C: Yeah that' 3 true. They do have — more
33 time to relax because - they don' t have to WORK.
34 You know, like everybody else does to
35 pay off their loans.

I
36 B: They'll be under less STRESS,
37 C: They- cause they're just getting money.
38 K: I feel really BAD about all this, because
39 like we sound like RAcists or whatever,

40 and I really don't think I am,

I
41 B: I don't think I'm being- I don't know - I

42 really don't think I'm racist, I just think that

43 K: It's just a very unfair society that we're
44 living in today. Especially on THIS campus and on =
I

45 —>B: ExACle. I-

46 K: = other college campuses,

There are two instances of “exactly” in this excerpt. In the ﬁrst instance (line 21),
“exactly” is used by K. to endorse B’s assertion “But they’re getting a free ride.” That is,
we may reasonably claim that K. agrees with, approves of, and ratiﬁes B.’s contribution.
Furthermore, it may be argued that K’s use of “exactly” at this point does more than
merely endorse B.’s contribution — it also brackets and calls attention to that assertion as a

more general point. Certainly the turns by K. (lines 1-19, only interrupted by “yeah” and

147

some other inaudible minimal utterance) which occur prior to B’s turn (in line 20) are long
and present a large amount of information, the point of which is explicitly begun at the end,
but not completed by K. — “But it’s just not fair that...” Instead, it appears that B.
completes K’s general point — it is not fair because K. has to pay back her loans and
interest, whereas Black people do not (“they’re getting a free ride”). K’s use of “exactly”
is used to bracket B’s (and the beginning of K’s last) contribution as a main position.
Therefore, K.’s use of “exactly” not only endorses B.’s contribution, but also points out its
importance at the ideational level.

Now consider the second use of “exactly” at the end of the excerpt (line 45).
Again, it is reasonable to claim that B. endorses K.’s position that it is “a very unfair
society that we’re living in today.” In addition, I would argue that B. doesn’t endorse just
any position, but rather one which expresses a main point or conclusion. Considering that
the entire excerpt had to do with “unfair advantages” given to Black students (i.e., minority
scholarships, government money for books, grade-point increases, etc.), it should be
obvious that K’s position sums up the point of much of the preceding talk. B.’s use of
“exactly” therefore highlights the signiﬁcance of that position with respect to the topic at
hand. In sum, “exactly” is used in this case to bracket what is taken to be a main position
as well as to positively evaluate (endorse) that position.

The agreement marker “exactly” appears in the concluding section of another

excerpt examined earlier, involving participants A., B. and C.:

(8)
49

51
52
53
54
55

56
57

58
59
60
61
62

63

65
66
67
68
69
70
71
72
73
74

75
76

148

A: = that's why I brought up the: thing about -
AT&T last year is- i- all th— the Black students =

I
B: Oh the monkey thing?

A: = wanted - wanted to drop AT&T because of the -
because they had a monkey on th- on their
advertisement? You didn't know about that? - It's
ridiculous. It's so stupid.

B: But you know= if— if they had a kangaroo for
Austra:1ia, nobody w-
I

A: Nebody would ( ) anything

I
C: See th- tha- tha- that’ s the
funny thing though. I mean — it' s as if the Blacks
think that everything about- everything is against
them. It's like- you know what I mean?

I

A: y- well they’ re- it’ 5 like

they’ re constantly battling. If they t =
—>C: [exACle

A: = bring issues up - then - nobody would =

C: It’s like they WANT =

A: = think about it. I

C: = racism to exist. You know what I mean?

A: So that they can get ahead.

—>C: ExACle.

A: ((snicker))
B: [[ I don't know.
A: I[ It’s so stu:pid.

C. [[ IV s such a joke. I don' t — I think it’ s a
joke, I think th— it' s a big joke.

((pausel)

149

Recall that in the larger stretch of talk to which this concluding portion belongs, participants
A., B. and C. provide in sequence three structurally similar CA23, each of which presented
an instance in which Black people complained about mistreatment, and then attempted to
prove that their complaints were unfounded. The ﬁnal section summarizes and derives
further conclusions from the three CA2s. Here we ﬁnd two instances of “exactly” by C.
(in lines 65 and 71), each of which follows a concluding point. The ﬁrst instance occurs
after A.’s position (lines 63-64); “It’s like they’re constantly battling.” This position is
induced from the three cases in which Black people showed some reaction to what they
perceived to be some injustice. The use of “exactly” not only endorses this position, but
further brackets it as salient at the ideational level. The second instance of “exactly” occurs
after A.’s position (line 70); “So that they can get ahead.” Interestingly, this position is the
last one to occur in this stretch of talk, after which the topic changes, and it appears that the
participants are satisﬁed that the current topic has been exhausted (perhaps ‘resolved’ in a
sense) and that the appropriate conclusions have been drawn. One might argue then that
A’s ﬁnal position “So that they can get ahead” is the most general position that the
participants infer from the three collaborative arguments. Furthermore, it is marked by
“exactly,” which signals B.’s endorsement of the position and brackets that position as a
more general point. It is interesting to note that “exactly” occurs only in the concluding
section and does not occur at all during the three CA2s. which is further evidence that
“exactly” is used to bracket more general positions (none of which occur in the earlier

sections).

150

Another feature of “exactly” is that it used in every case in the data by a speaker
whose position was provided immediately prior to the interlocutor’s position which
“exactly” endorses. For example, in the previous excerpt, C.’s use of “exactly” in both
cases endorses A’s assertions, each of which follows an assertion by C. In the earlier
excerpt with participants K., B. and C., “exactly” is used by B. to endorse K’s position

which itself follows the interrupted beginning of a position by B (marked by “I just think”):

(9)

41 B: I don't think I'm being- I don't know — I
42 really don't think I'm racist, I just think that

43 K: It's just a very unfair society that we're living
44 in today. Especially on THIS campus and on =

I
45 —>B: ExACle. I-

46 K: = other college campuses,

It seems plausible that by using “exactly” in this case, B. is signaling to K. that K.’s
utterance is what B. would have said if she had been able to complete her turn. This
suggests that “exactly” also serves the purpose of signaling that its user not only endorses
the main position expressed by a prior turn, but also that that prior turn communicated
(perhaps in a better way) what the user of “exactly” originally meant to say or was unable

to say because of interruption.

151

5.2.1 .1.2. The use of repetition — endorsing and bracketing the supports of others.

It will be argued here that various kinds of repetition are quite common in pseudo-
argument because repetition in part serves the purposes of signaling endorsement as well as
of bracketing supports. Tannen (1989) discussed the use of repetition in discourse and
found that it serves a host of functions, including those functions having to do with the
management of information (production, comprehension, and connection) as well as those
having to do with interaction (i.e., vying for the ﬂoor, back-channeling, bounding
episodes, and ratifying the ideas of others). In terms of production, Tannen argues that the
automatic nature of repetition allows a speaker to carry on conversation with less effort and
thus to produce more ﬂuent speech. The rrrirror image of this is with comprehension where

repetition provides semantically less dense discourse which then allows the bearer to

adequately process information and keep pace with the speaker? Repetition also serves the

function of creating cohesion by tying new utterances to prior ones.

Most relevant to our purposes is the interactive function of repetition: “...repetition
not only ties parts of discourse to other parts, but it bonds participants to the discourse and
to each other, linking individual speakers in a conversation and in relationships.” (T annen,
1989, pp.51-52). In terms of interaction, Tannen argues that the repetition of the
utterances of others may serve to signal listenership, the perception of shared expertise and

acceptance. Consider the example given by Tannen to illustrate these points:

152

 

(10)
1 Steve I never saw anything wrong with those things.
2 I thought they were funny.
3 Chad Yeah.
4 Deborah I hated them.
5 Chad I agree. [i.e. with Steve]
6 Peter What. The cartoons?
7 Steve lmlcckthemsericuslx.
8 I never thought anyone was
9 Deborah I couldn’t staznd it.
[One page of transcript intervenes]
10 Steve v '
11 Peter lnmuculdtakeitsericuslx.

Peter’s repetition (in line 11) of Steve’s utterance (in line 10) does not contribute any new
information. However, Tannen argues, it contributes something signiﬁcant at the
interactional level — Peter’s participation in the conversation and his ratiﬁcation of Steve’s
utterance. Also, it is not only repetition of the content which signals ratiﬁcation, but also
repetition of form; “...it is not only what Peter says that shows that he agrees with Steve,

but also the way he says it. By repeating not only Steve’s idea, but also his words and

syntactic pattern, Peter’s contribution is a ratiﬁcation of Steve’s.” (p.61).3

Tannen distinguishes different forms of repetition. First, one must distinguish self-
repetition from repetition of others. Second, instances of repetition fall along a scale in
terms of ﬁxity of form ranging from exact repetition to paraphrase (repetition of content and
not form). There are several kinds of repetition in the data that will be analyzed in what
follows; these include instances of self-repetition, repetition of others, as well as exact

repetition and paraphrase. It will be shown that these instances of repetition serve the

153

functions of endorsement (ratiﬁcation) and in some cases the bracketing of ideational units.

Consider the use of repetition in the following short excerpt from the data (part of

the excerpt with participants K., B., and C.:

(11)

2O
21

22
23
24
25

26
27
28

29
30
31

B: But they’re getting a free ride AND =

I
K: eXACle

B: = they could be successful in THEIR job TOO. AND
plus they get this stupid point five increase in
their GRADES, so now they're going to be graduating
with- [

C: PLUS they have a better chance
of getting a job, so they're - going to be — more
successful ( )

I
B: So iV s just- i- their advantages just keep

adding UP! Their - their advantages totally keep
adding UP. Their grade advantages,

In this short excerpt there are numerous instances of repetition at different linguistic levels

(lexical, prosodic, and syntactic). For the moment, compare the second contribution by B.

(in lines 22-23, beginning with “AND plus” ) with the contribution by C. (in lines 26—28):

(12)
B: AND plus they get this stupid point ﬁve increase in their GRADES,

so now they're going to be graduating with-

C: PLUS they have a better chance of getting a job,

so they're - going to be - more successful

154

Note the lexical repetition (i.e., “plus,” “they,” “so,” “they’re going to be”), the prosodic
repetition (i.e., emphasis on the initial word, rising intonation at the end of the ﬁrst clause
of each contribution), and the obvious syntactic parallels. In addition, at the ideational level
there appears to be repetition in that in each case a cause is coupled with an effect which
together serve as support for the main position that Black students receive unfair
advantages. Thus it seems reasonable to claim that C.’s contribution mirrors B.’s
contribution in many linguistic ways. Furthermore, it may be argued that the use of
repetition in this case serves several functions. First, C.’s use of repetition signals that her
contribution is the same kind of thing as B.’s — a support. Evidence for this is that B’s
contribution is initiated by “AND plus,” which clearly marks her utterance as another item
in a list of supports (see chap. 4 for arguments that a list of supports is being provided
here). That C. also begins her contribution with “PLUS” shows that she views what she is
about to say as another item in that list. Second, C.’s use of repetition not only marks her
utterance as a support, but further marks it as parallel in content to B.’s utterance — in each
case, some “unfair” advantage leads to an “unfair” result. Third, it could be argued that
C’s use of repetition serves as an endorsement of B’s contribution. By mirroring B.’s
contribution (in form), C. shows that she agrees with what B. has said.

There is another instance of repetition in the previous excerpt which involves a

single speaker repeating her own utterance:

(13)

29 B: So it's just— 1- their advantages just keep
30 adding UP! Their - their advantages totally keep
31 adding UP. Their grade advantages,

155

This instance is interesting because it involves a seemingly unmotivated use of repetition.
That is, one rrright ask why B. bothers to repeat herself. However, in the context of the
CA1 in which the participants have been involved, B’s repetition is understandable. Her
contribution (“So it’s just- i- their advantages just keep adding UP!”) is a repetition of the
main position which was supported throughout the CA1. Repeated in this place after a
variety of supports have been given, it serves as a concluding statement (or kind of coda)
of the argument. It seems plausible that B.’s repetition (“Their - their advantages totally
keep adding UP.) functions both to bracket and intensify this conclusion. That is, the use
of repetition in this case is similar to the use of the metalinguistic tag “exactly” in that it
brackets what is seen to be a more general point of special signiﬁcance which has fallen out
of the talk.

So far, primarily the use of repetition of form has been discussed. However,
repetition of content (paraphrase) also seems to play a signiﬁcant role in CA1. Consider

the continuation of the last excerpt:

(14)

29 B: So it's just- i- their advantages just keep
30 adding UP! Their - their advantages totally keep
31 adding UP. Their grade advantages,

I
32 C: Yeah that' s true. They do have - more
33 time to relax because - they don' t have to WORK.
34 You know, like everybody else does to
35 pay off their loans.

I
36 B: They’ll be under less STRESS,

37 C: They- cause they’re just getting money.

156

After B.’s conclusion in lines 29-31, C. gives her endorsement (“Yeah that’s true”) and
then continues in lines 32-35 with a new support which is similar in structure to the
immediately prior supports provided by B. and C. in that a cause is coupled with an effect
(some “unfair” advantage leads to some “unfair” result). In this case, C. claims that the
unfairness lies in the fact that since Black people don’t have to work to repay loans, they
can relax more than White people can (Again, the poor reasoning here on C.’s part is
irrelevant to the point at hand.) At this point B. paraphrases the ﬁrst part of C.’s support
with “They’ll be under less STRESS,” (in line 36), and then C. completes the repeated

support initiated by B.’s paraphrase (in line 37):

( 15)
SUP l . C.: They do have - more time to relax
2. because - they don’t have to WORK...
SUP r 3. B.: They’ll be under less STRESS,
4. C.: They - cause they’re just getting money.

Besides the semantic closeness of what I am calling B.’s paraphrase of C.’s assertion
(“They do have - more time to relax”), there is evidence within the talk that B.’s utterance
(3) is interpreted by C. as a paraphrase — C. apparently cuts short what she was about to
say (“They - “), and recycles the second part of her support, marked by repetition of
“because,” indicating her understanding that B’s contribution ﬁlls the same slot as does (1).
C.’s utterance (4) is itself is a paraphrase; “They’re just getting money” seems to

paraphrase “they don’t have to WORK...” It can be argued that B.’s paraphrase of the

157

initial part of C.’s support ( 1) serves minimally as an endorsement of C.’s support. In
addition, B’s paraphrase not only signals endorsement, but also adds weight to C.’s
support and strengthens their alliance.

In sum, the examples of repetition seen in this section and the examples involving
the use of “exactly” in the previous section point to the important role of endorsement in
pseudo-argument. Unlike ordinary argument in which participants in opposition only
grudgingly endorse the positions of others, participants in pseudo-argument constantly

signal that they agree with each other, and this is achieved in large part linguistically.

5.2.2. Some features of collaborative argumentz

The previous section showed that a variety of linguistic phenomena cooccur with
CAls and function in large part to signal endorsement and alliance as well as to bracket
units within the discourse (supports, main positions). In the analysis which follows in this
section, other linguistic phenomena will be discussed which cooccur with CA2 and which
seem to provide some evidence for its structure. Recall that CA25 are accomplished

ordinarily by two or more participants who collaborate in simulating an argumentz by
importing and then destroying the argument] of an absent antagonist. The structure of a

CA2 was schematized as follows:

(16)

A: POSo / (SUPo l, 2, 3..., n)
X: DIS of POSo/(SUPo)
X: SUP 1, 2, 3..., n of DIS

158

Whereas much of the linguistic work which goes into CA1 involves the social functions of
signaling agreement and alliance, much of the linguistic work in CA2 involves the use of
structural indicators which distinguish the protagonists’ from the imported antagonists’
positions and supports. That is, a variety of linguistic devices are employed in CA2 which
mark an utterance as being authored not by the speaker but rather by some absent
antagonist. In addition, other linguistic devices are used subsequently to signal dispute of
those imported positions and supports. In the sections which follow, therefore, it will be
shown how various linguistic phenomena are used by participants in PA to signal

underlying structural units such as the P080, SUPo, and the dispute of P080 and SUPo.

5.2.2.1. Importing the antagonist.

5.2.2.1.1. Direct quotation . One common way of importing the antagonist’s position in
the data is through direct quotation. Direct quotation involves the repetition of words from
a prior time and is normally framed as dialogue from the perspective of the original author
of those words. Tannen (1989) argues at length that “direct quotation” is usually
mistakenly thought of as the neutral representation of another’s words (so-called “reported”
speech), but it is better thought of as “constructed dialogue,” primarily the creation of the

speaker and not the quoted party:

I am claiming that when a speaker represents an utterance as the words of another,
what results is by no means describable as “reported speech.” Rather it is
constructed dialogue. And the construction of the dialogue represents an active,
creative, transforming move which expresses the relationship not between the
quoted party and the topic of talk but rather the quoting party and the audience to
whom the quotation is delivered. (p.109)

159

By emphasizing the role of the quoting party in representing the speech of another,
attention is shifted towards the purposes of direct quotation within the interaction. Viewing
direct quotation as “constructed dialogue” means seeing it as the active construction of a
speaker who takes alignments and evaluative stances towards what is quoted.

Consider the following excerpt examined earlier:

(17)

1 B: They totally have this- y- Exactly. And- You
2 know with Asians, you don't see Asians like =

I
3 K: I l
4 B: = pulling an attitude, and I don't think I- I
5 have like Asian friends and - I- I don't think I've
6 Ever seen an Asian - pull- you know and like- Black
7 —>people bring up like ”well — you owe it to us
8 ~4>because — you know - you had us as slaves - =

I

9 K: Yeah but

10 how long:

11->B: = thousands of years ago, and blah blah blah:,

12 —>and you OWE it to us. Blah Blah" Well what about -

13 what about when we took like during what was it

14 WOrld war Two, when we took all the Japanese, and

15 stuck them in prison camps in California. I mean

16 what about THAT? You don't see Asian people still

17 all fired up about THAT, - I mean it is totally =
I

18 K: Yeah and if that IS the

19 case,

20 B: = WRONG but- Just- I mean it was a mistake in
21 the PAST, and it shouldn't - like - I don't know,
22 it- it- YEAH it does effect us, and YEAH it was
23 WRONG, but - it wasn't ME, it wasn't even my DAD.
24 It wasn't even my GRANDpa. It was like-

I
25 C: And they' re-
26 now they' re STILL - you know basically treated:

I

27 B: ( ) ago.

28 C: = not just equal, but they get a- the advantage

160

29 of - you know in a lot of situations so they can't
30 complain that they're not treated equal now.

That B. is using direct quotation in lines 7-8 and 11-12 seems obvious — the use of the
marker “Well,” which commonly occurs at the beginning of quoted material, the use of
pronouns from the other’s perspective (deictic shift), and the quote introduction “Black
people bring up like.” Here then is the most straightforward way to import the position of
an antagonist — simply present the other’s (presumed) views in their voice from their
(presumed) perspective.

It is apparent however, that what is involved in the previous instance of quotation is
something other than the straightforward, neutral representation of the antagonist’s views;
B. is also engaging at the same time in a considerable amount of (negative) evaluation of
and commentary on the position and support she is representing. For example, B.
exaggerates by saying “you had us as slaves thousands of years ago.” It is unlikely that B.
believes that slavery took place thousands of years ago and it is equally unlikely that B.
believes that Black people believe this. It is more likely that the exaggeration hints at a
dispute which actually follows later, that slavery occurred so long ago that it is by now
irrelevant. So the use of exaggeration in this case embeds an indirect dispute within the
quote and also displays a negative evaluation of the quoted material.

In addition to the use of exaggeration, B. also engages in negative evaluation by

using the ﬁller item “blah blah blah” which seems to have the effect of downplaying or

161
dismissing the importance of the position(s) and support(s) being quoted. In the next
excerpt (seen several times so far), we ﬁnd direct quotation with evaluation in the

introduction rather than embedded within the quoted material:

(18)

18 3: = and the Black community was complaining and
19 whining- whining - "Oh we d- we don't have enough
20 ah Black vice-presidents." - whereas the three- the
21 three- the three guys probably made- made more =

I

22 I: guys
I
23 2: guys made the biggest-
24 3: = than anybody in the whole corporation:

Here again, that C. is using a direct quote is clear — the use of “o ” which commonly
introduces quotations (Schiffrin, 1987), the shift in pronominal reference, and the
introduction; “the Black community was complaining and whining- whining -.”
Notwithstanding the prosodic characteristics (whining voice), C.’s quote is fairly
straightforward compared to B’s in the previous example. It appears that any negative
evaluation or commentary by C. is carried out in the introduction rather than within the
quote; C’s characterization of the position quoted as “complaining and whining” indicates
C’s negative evaluation of that position. Also, similar to the previous example in which B.
hinted at a dispute to follow, C.’s characterization of the antagonist’s position as “whining”
hints at C.’s contention that it is unfounded, since “whining” connotes unjustiﬁed

complaint. Indeed, as was noted in chapter 4, after presenting the antagonist’s position, C.

162

(and B.) then go on to dispute that position and show that such a complaint is unfounded.

5.2.2.1.3. The discourse marker “whatever”. The two distinct sequences of moves found
in pseudo-argument begin with either POS or P080, and one would expect there to be
various kinds of linguistic devices which could be employed to signal whether a position
was adhered to by a speaker or else ascribed to an absent antagonist. We have already seen
that direct quote is an effective way to import another’s position while at the same time
signaling distance from it. This section will focus on the work done by “whatever” in
discourse, and it will be shown that one use of “whatever” functions in pseudo-argument
much the same as direct quotation does. It will be argued that ‘whatever’ is a discourse
marker which functions in at least three distinct capacities in terms of its role at various
levels of discourse. I will label the three uses WHATEVERI, WHATEVER2 and
WHATEVER3. As an initial gloss, WHATEVER] is used primarily for list and labeling
problems, where either a speaker is unable to retrieve an item from memory, or else where
the speaker is uncertain about the accuracy / appropriateness of a stated item.
WHATEVER2 is used where a speaker shuts down prototypical argumentation which is
deemed irresolvable. WHATEVER3 functions in pseudo-argument to signal speaker
orientation towards the material preceding it in the utterance. Speciﬁcally, it marks that
material as ‘other-authored’ and at the same time signals the speaker’s disafﬁliation with or

opposition to its content.

163

5.2.2.1 .3.] . WHATEVER] — listing and labeling difﬁculties . Consider the following

example from the data, seen before in section 4.2:

(19)

K: = esPECially on this campus, I really think that
- a lot of people: - like White students - ARE: -
not AS racist as the Black students. Because - I’VE
had classes where Black students are just li:ke —
oh: - I can't believe- or even like where I work.
Where I work a lot of people come in, and White

—>people Black people whatever i- a lot of Asian
people. And a lot of Black people are just li:ke -
if you don't HAVE something, for them, like - if
people need to check out basketballs if you don’t
have something for them they think that it’s
because “Oh. I'm Black. that - I can't have it."
You know?

HI—‘HI—‘WCDQCDU'IQUJNH
le-‘O

In this excerpt, K. attempts to support her position that White students are not as racist as
Black students. She seems to abort the ﬁrst argument (having to do with classes, in lines
3—4), but then begins another which is based on her work experience. K. provides some
orientation, listing the various types of people who enter the gym (in lines 5-7). She
inserts “whatever” in this case as a ﬁller item which suggests that the list could be
extended, but that perhaps further items are not available for recall at that moment. This
interpretation of “whatever” as a ﬁller item is supported by K.’s further addition of another
item to the list (‘a lot of Asian people’) after using “whatever.”

Further evidence that “whatever” signals a retrieval problem comes from instances

in the data where there is some pause before “whatever,” as in the following:

164

(20)

1 L: Um - I definitely believe that there is racism on
2 this CAMpus. Um - I can’t attest to WHERE it is, or
3 —>WHO it is, or - whatever, but I definitely believe

4 that it’s there. And um: - ((laugh)) I don’t know.

In this sample, which is taken from a conversation between three African-American
students, L. pauses after ‘or’ in line 3, seemingly desiring to but unable to provide
a third item in a list. Also note that the items being coordinated in the list need not
be noun phrases.
A related use of “whatever” is when it is used to signal speaker uncertainty about

the accuracy or appropriateness of an item. Consider the following example from the data:

(21)

1 B: I don't know but it totally- it totally separates
2 it.Like - I remember you know freshman - um.- like

3 ->orientation week or whatever:, welcome week - last

4 year, there was all these signs for you know like -

5 Black ice cream social, Black caucus ice cream social,
6 you know? Well, what if I wanted ice cream too? for

7 God’s sake. ((laughs))

In this case, “whatever” is used to express speaker uncertainty about the correctness of the
label ‘orientation week.’ Evidence for this interpretation is the two pauses separated by
‘urn’ before the item, and the repair afterwards (“welcome week”). Clearly B. has
difﬁculties at ﬁrst retrieving the correct item, and inserts “whatever” to show a lack of

commitment to the accuracy of the ﬁrst item provided (“orientation week”).

165

As with “whatever” as a list ﬁller, “whatever” with label troubles is ordinarily
preceded by “or,” and occurs clause ﬁnally. Further, this use of “whatever” serves to
downplay the weakness of the speaker’s presentation and thereby reassure the hearer that
the failure to supply the correct term does not impair the speaker’s larger discourse goals.
“Whatever” used with label problems is often followed by another clause (by the same

speaker) which is preceded by resumptive “and,” “but” or “so”:

(22)

1 H: Yeah. I have a class right now, and - there's - it's-
2 it’s an African - ah social science class type kind of

3 you know it’s a general overview of the country

4 —>whatever, and I'M amazed that- that Blacks over here

5 even want to be associated with Afric- with Africa at

6 all.

Here H. provides some background information about the class he is taking (lines 1-3), but
runs into some trouble in glossing the course’s content. Conceivably, he dimly realizes
that Africa is not a country, and inserts “whatever” to express his uncertainty about the
gloss. Here again, with “whatever” H. downplays the importance of his poorly phrased
gloss, and inserts resumptive “and” to signal a return to his main point.

These uses of WHATEVERl are treated together, since they have important
features in common. “Whatever” used for both label and list problems has to do primarily
with information management in that in both uses the speaker has retrieval difﬁculties, and
uses “whatever” to downplay the signiﬁcance of that missing or incorrect information in

terms of the speaker’s larger discourse goals.

166

5.2.2.1 .3.2. WHATEVER2 - marker of irresolvabilty . The next use of “whatever” is
attested only once in the data, but will be discussed due to my contention that it is more
common in everyday conversation than the data would lead us to believe. This use of
“whatever" ’ serves to suspend argumentation when resolution or agreement is deemed
unlikely. This device allows one to ‘bail out’ of an argument where continuation could
damage the relationship between participants. In such cases a speaker uses “whatever” to
signal that neither participant is likely to yield from their respective incompatible positions,
and thus that continuation would do more harm than good.

Consider the following excerpt from the data in which up to this point A. and T.

have been arguing over whether various instances give evidence for reverse discrimination

oncmmmm:
I23)
1 A: Okay then, tell me if THIS is reverse
2 discrimination. My freshman year:, my brother floor
3 was - very - African-American, there were only a
4 couple White guys on the floor, and I became good
5 friends with my R.A., and some of his friends, and -
6 one day I went to sit with them at the TABle, -
7 the Black table in Wilson.= He was Black, and I =
I
8 T: was your R.A. Black or White.

9 A: = went to SIT there, and people got up and left
10 because they didn’t want a White girl sitting there. -
11 What's that.

12 T: A ----- . How do you KNOW that they- that it was =
I

13 A: Because =

14 T: = because you were White.

15 A: they SAID that! They said ‘We don't want no White
16 girl sitting here!’

167

17 T: Well - again. I mean I think that you’re — that
18 you're generalizing again. ( )

I
19 -—>A: Whatever. New topic. How
20 do you feel about interracial dating on campus,

21 T: I just think th-

I
22 G: I really don' t see it happening all
23 that much? Um - a- a- in fact I've seen it
24 INfrequently. But I guess I really don' t look OUT for
25 it? and NOtice it? But I maybe have seen a ha:ndfu1 of
26 couples. Through my years.

First note that this excerpt exempliﬁes ordinary argumentation and not pseudo-argument,
since the participants are clearly in open opposition. Indeed, WHATEVER2 seems most
suited for ordinary argument and not pseudo-argument, since it turns on some
disagreement between participants. Here A. provides a narrative (lines 1-7, 9-11) which
serves as support for her claim that there is reverse discrimination on the campus. T.
shows skepticism at various points (e.g., in lines 12, 14, 17-18), and the overall effect of
his contributions is the rejection of A.’s narrative as support for her position. After having
discussed several cases such as the present one and neither party yielding from their
positions, A. ﬁnally abruptly brings the argument to an end with “whatever.” That her next
utterance is ‘new topic’ is evidence that A. (the interviewer) believes that any further
argument on the prior topic would be fruitless. A.’s overlap, intonation, and abruptness
suggest her frustration.

The use of WHATEVER2 by a participant could be considered an offer to suspend

the argument by agreeing to disagree, thus allowing the interaction to move on. Clearly

168

this offer must be accepted by all parties in order to achieve this kind of interactional (but
not ideational) resolution. In the excerpt above, it appears that T. (in line 21) may not be
ready to relinquish the prior topic (perhaps still hoping for some ideational resolution) and
therefore does not accept A.’s offer which is signaled by “whatever.” However, in line 22,
G. assents to the change in topic and even interrupts T., thus overriding T.’s bid to
continue the argument. Clearly A. and G. are ready to move into less turbulent waters.
Willard (1989) maintains that one good indication that arguments are resolution
oriented is that when arguments are ‘going nowhere,’ they are not worth continuing;
“Observations and interviews consistently reveal arguers who end arguments because they
have become repetitive or because the discussion has reached a stasis beyond which the
participants cannot progress.” (p.85) WHATEVER2 is an appropriate linguistic device to

achieve this end.

5.2.2.1 .3.3. WHATEVER3 — importing the antagonist. It will be argued in this section
that WHATEVER3 marks preceding material as ‘other-authored,’ and further expresses the
speaker’s unsympathetic relation or opposition to that material. By ‘other-authored,’ I am
referring to the ‘participation framework’ of a speaker (Goffman, 1981) whereby the
‘animator’ and ‘author’ are not one and the same person. A participation framework has to
do with the shifting positions which participants take in relation to utterances in both
reception and production. In terms of production, an animator is that aspect of self which

physically produces utterances. An author is that aspect of self which is responsible for the

169

content of talk. Schiffrin (1990) notes that “...when a speaker quotes his/her own prior
words, he/she is at once the animator and the author of that quote; but when a speaker
quotes another’s prior words, he/she is assigning the authorial role to the original source of
those words” (p.242). Another participant status involved in production and relevant to
our purposes is the‘principal,’ which is that aspect of self which expresses commitment to
an utterance; “someone whose position is established by the words that are spoken, ...,
someone who is committed to what the words say” (Goffman, 1981, p.144).

It is presumed here that shifts in participation framework are ordinarily marked
linguistically. With regard to the relation between a speaker and his or her utterance, I
follow Schiffrin (1990) who states that; “Our tacit understandings of talk include the
assumptions that what is being said is the speaker’s position, that the speaker is committed
to his words, and that the speaker does believe in what is being said — such that evidence
to the contrary is expected to be explicitly marked...” (p.242). It will be shown in what
follows that WHATEVER3 is a discourse marker which explicitly marks a shift in
participation framework whereby the speaker indicates that what is being said is not his or
her position or support (i.e., that he or she is not the author of the utterance), and further
that he or she is unsympathetic or not committed to his or her words (i.e., shift in the
display of the principal).

In terms of ideational structure, WHATEVER3 signals the importation and
representation of an antagonist’s position or support of a position. At the same time

WHATEVER3 expresses the speaker’s doubt about the validity of that position or support.

170

Consider the third use of “whatever” in example (24):

(24)

\ImUTIbLUNH

(I)

10

ll

12

13

15
16

H: Yeah. I have a class right now, and - there's -
it’s- it's an African - ah social science class type
kind of you know it's a general overview of the
country whatever, and I’M amazed that- that Blacks
over here even want to be associated with Afric- with
Africa at all. Oh sure or whatever- it's the homeland

_>you know th- whatever.

III )
[[Why.Is it really bad?
I

Oh ma::n?

You should hear th-

- You should hear the stories my friend m- - he went
0 boarding school in Massachusetts last year?
I
D: They should be ( ) then to be
in America. They' re always BITCHing about America.

H
D
S
H: It’s SO bad. The poverty over the:re,
S
S.
1:

After presenting some orientation (lines 1-3), H. states his position that Black people have

good reason not to want to be associated with Africa (lines 4-6). In this case, H’s position

is opposed to an antagonist’s position that there are good reasons. H. then provides what

is clearly meant to be a representation of an (absent) antagonist’s support for the position he

is arguing against. That is, H. imports what (in his conception) might be a Black person’s

support for the position that African-Americans should want to be associated with Africa.

Not only does “whatever” mark this support as ‘other-authored,’ but it also has the effect

171

of distancing the speaker from the propositional content. Indeed, “whatever” in this case
(in combination with the sarcastic “oh sure,” and intonational features not depicted in the
transcription) serves to indicate the speaker’s negative orientation towards the ‘other-
authored’ support — for H., that Africa is considered a homeland by African-Americans is
not a sufﬁcient reason for wanting to associate with Africa. Hence, H. uses “whatever”
not only to mark a shift in the author, but further to mark a shift in the principal where H.
displays a lack of commitment to the propositional content of the utterance and the value of
that utterance as support for a position. H. then continues after D’s prompting by
providing an evaluation and support (“there's so much poverty”) for his own position (line
11)

It is probable that the same interpretation of H’s utterance in 6—7 would be
recoverable by participants (and the analyst) without H’s use of “whatever.” However, as
Schiffrin (1987, p.320) notes, discourse markers do not create meaning or structure, but
rather “select and display” already recoverable meaning or structural relations between
unmemdm.

Now consider another example:

(25)

A: I don't think it's being racist, I think they're
getting a lot more scholarships, and - if they need
them, fine. But maybe have some more- I don’t even
know if they have work study? Or if they all - if— a
requirement had been a scholarship? Or financial aid
( ) work study I )? Maybe they should.be working,
and not just have it all handed to them. - And they
kind of think like they - like they - desERVE it. -
Because - THEY grew up in the inner-city, and.WE grew

\DQOUlerNH

172

10 up in west Bloomfield and Farmington Hills or

11 —>whatever, So they— they deserve it. Cause our mommy
12 and daddies could afFORD it. And they don't know; And
13 I think that THEY generalize a lot too.

Here, beginning with line 7 (“And they kind of think...”) A. seems to be representing both
the position and support of an absent antagonist. She introduces the ‘other-authored
position’ (in lines 7-8) with “And they kind of think... ”, an initial indication that what
follows is a representation of someone else’s belief. This is followed by the corresponding
‘other-authored support’ (in lines 9-11); “Because THEY grew up in the inner-city....”
The use of “whatever” following the other-authored support (in line 1 1) is a signal that A.
is not committed to the validity of the support she represents. In fact, “whatever” here has
the effect of calling into question the validity of the support, and evidence that A. does not
subscribe to the support is that she indirectly disputes it afterwards (in lines 12-13) with the
criticism that such claims overgeneralize and therefore are not convincing (see chapter 4 for
further arguments that A. is “importing the antagonist”). Once again, in this example, a
speaker imports the position and support of an absent antagonist, and then argues against
it. “Whatever” is used to signal that what precedes it is ‘other-authored’ and further that the
speaker is not sympathetic to its content or is even in opposition to it.

Finally, there are instances is the data where “whatever” seems to function as

WHATEVERl and WHATEVER3 at once:

173

(26)
1 D: = said something like that. Her English - =
I
2 H: She’s a
3 gra:d student.
4 D: = was entirely — terriblez, - and she was- =
I
5 H: It's ridiculous.
6 D: = she was trying to cut somebody else down for not
7 being able to communicate, and I could barely
8 understand what SHE said. You know, it was s- TOTally
9 ridiculous, she's s- a graduate student, supposed to
10 be educa:ted, - it was retarded. - I know that =

I
11 H: We—

12 D: = people have different - you kno:w, uhh speaking
13 —>patterns, or whatever, but - you know. Once you become
14 a - college edurcat- college educated person -

15 ( )

In this example, there is evidence that D. has some labeling difﬁculty given the pause after
“different” (line 12), the use of the marker “you know” with rising intonation (line 12), and
then “uhh” (line 12). So in this case “whatever” probably marks a lack of certainty by the
speaker that she has used the appropriate phrase. On the other hand, “whatever” seems to
mark the preceding material as other-authored. Let’s consider the ﬁnal portion of D.’s
argument. D. takes the position that “educated” people should not be hard to understand
(presumably because of a non-standard dialect). Instead of providing her own support for
that position, D. provides a support for an antagonist’s position (in lines 12-13), that no
groups’ “speaking patterns” are inherently better than any other’s. Indeed, D. happens to

be a student in an introductory sociolinguistics course in which just such a notion is heavily

174

stressed. However, although D. might agree with the general proposition that some people
should be free to speak their own dialect, she evidently does not agree that this dictum
should apply to college educated people. The use of contrastive “but” in the clause which
follows “whatever” shows that D. disputes the other-authored support which she supplies
— having different speaking patterns is acceptable except for a college educated person.
The use of “whatever” in this case then serves to distance the speaker from the content of
the clause which precedes it, at least with respect to its value as a support for an
antagonist’s position. So this example gives evidence for “whatever” serving a dual
function as WHATEVER] and WHATEVER3.

In sum, it has been claimed that the analysis of linguistic devices such as
WHATEVER3 may shed light on the structure and function of pseudo-argument. We have
seen that in pseudo-argumentative discourse one would expect to ﬁnd various linguistic
devices which allow speakers to signal the importation of an antagonist’s positions and
supports. One such device is WHATEVER3, which is a discourse marker that brackets the
other-authored position or support of an antagonist. In addition, it has the effect of
downplaying or dismissing the bracketed material, and alerts the bearer to upcoming
disputation. In terms of participation framework, WHATEVER3 marks a shift in the
status of ‘author’ as well as a shift in ‘principal.’

Linguistic devices such as WHATEVER3 are useful for participants in pseudo-
argurnent since they help participants to orient to particular structural parts of the discourse

(e.g., to antagonists’ positions or supports). That is, they are important resources for

175

participants in the construction and understanding of the discourse they are involved in.
For the analyst, examination of low-level linguistic features such as “whatever” provides a
valuable tool for understanding the structure and function of a particular speech activity.
Finally, I do not want to claim that WHATEVER3 is limited to its role in pseudo-
argumentative discourse. Clearly, this use of “whatever” could conceivably be found in
genuine argumentation in which the antagonist is present. However, it seems that
WHATEVER3 is less likely to surface in this context, since the antagonist is already
present to provide the positions and supports of his/her choice — putting words into the

mouth of a real antagonist might be considered presumptuous at best.

5.3. Summary.

In this chapter is was shown that there are various linguistic phenomena which
cooccur with pseudo-argument. These surface linguistic features function in part to bracket
underlying structures of the discourse and guide the interpretations of hearers. For
example, the agreement marker “exactly” brackets a main position, and repetition (of form
and content) may be used to bracket supports. In CA23, “whatever” and direct quotation
may be used to bracket an antagonist’s imported position or support. The linguistic
phenomena analyzed also shed light on other aspects of PA. For example, the analysis of
“exactly” and of repetition showed that participants in PA are actively involved in signaling
endorsement, thereby reinforcing their alliance and the position and supports that they stake

out. Participants show concern for signaling to each other at each point that they agree

176

with, ratify and appreciate each other. The reasons for this pervasive concern will be
discussed in the next chapter. The use of WHATEVER3 and direct quotation not only
serves to bracket an antagonist’s imported position or support, but also allows the speakers
to signal disafﬁliation with the imported other’s argument and to foreshadow upcoming
disputation.

It is claimed that these linguistic phenomena provide evidence for the underlying
structures proposed for PA and that they are important resources for participants in the
construction and understanding of the discourse they are involved in. It is also claimed that
these features serve as further evidence that PA is distinct from OA, because the features
analyzed do not commonly occur in OA, where two participants playing the roles of
protagonist and antagonist are in open opposition. For example, WHATEVER3 is unlikely
to surface in OA because in this type of discourse the antagonist is present to speak for
himself or herself. To presume to represent the views of a present antagonist (marked by
WHATEVER3) could be construed as rude. Certainly rudeness may be part of OA, but the
point here is that it is not normal — even ordinary argumentation is normally a cooperative
enterprise where participants attempt to iron out differences in a rational, rule-govemed,
and polite manner (see chap. 3).

In addition, endorsements of an opponent’s positions and supports marked by

9’
1

“exactly or the use of repetition would be unusual in OA since normally participants in
OA only grudgingly endorse the positions and supports of their opponents with regard to

some issue. That is, although a participant in OA might concede some point by endorsing

177

an opponent’s utterance, this is the exception rather than the rule, whereas such
endorsement is the rule in PA. If “exactly” were to occur in OA, then it would most likely
mean something like “You are supporting MY position, and not your OWN position.”
Also, paraphrase of an opponent’s position or support in OA would be somewhat unusual
and might be interpreted as mockery or sarcasm.

Thus it is claimed here that the linguistic phenomena analyzed in this chapter help to
distinguish OA from PA. It is not claimed that these phenomena are limited to PA —
obviously repetition, endorsement, and direct quotation occur in a wide variety of
discourses with a wide variety of functions. Nor is it claimed that these are the only
interesting linguistic features of PA available for analysis. For example, one might
examine indirect quotation, or the use of 1st and 3rd person plural pronouns in terms of
how they are used to delimit “us” versus “them,” and how they signal solidarity and
alliance between participants. Or one might analyze the use of negative evaluative tags such
as “It’s ridiculous” or “It’s so stupid” in terms of how these tags bracket other-authored
positions or supports. My limited claim therefore is that the linguistic phenomena analyzed
in this chapter have the functions described when they are used in PA and that participants

are aware of these functions and competent in the use of these surface linguistic features.

Chapter 6

THE FUNCTION OF PSEUDO-ARGUMENT

6.0. Introduction.

So far I have argued for a distinct speech activity called pseudo-argument on
structural and linguistic grounds. In this ﬁnal chapter I will make the same argument on
functional grounds. As was discussed in chapter 2, those approaches committed to the
functionalist paradigm view discourse as inseparable from the purposes of language in
human life. Thus, the study of discourse requires looking beyond formal relations within
the linguistic system to consider the embeddedness of language in interaction between
individuals with particular aims in particular contexts and cultures. In this chapter the focus
will be on the embeddedness of PA in its interactive setting with special emphasis on the
purposes and motivations of participants involved in this type of talk. Earlier chapters on
the structural and linguistic properties of PA only brieﬂy touched on the purposes of the
participants in this type of interaction. That is, while the structures and surface linguistic
details were described and analyzed, there was little attempt to connect them to the purposes
and motivations of participants. For example, in chapter 4 it was noted that contrary to
what occurs in ordinary conversational argument (Jackson and Jacobs, 1986), in PA

supports are provided for positions without the speaker’s anticipation of disagreement or

178

179

doubt on the part of interlocutors. However, no account was provided for why this is so.
Further, in chapters 3 and 4 the importation of an absent antagonist’s positions and
supports in PA was described -— but there was no explicit account provided for why
participants in PA import an absent antagonist’s positions and supports. In chapter 5,
various linguistic features were analyzed which cooccur with PA and which are used to
bracket underlying structural units of the discourse. Although there was some explanation
of the use of these features in terms of certain functions (endorsement, alliance, etc.), there
was no discussion there about why such functions are part of PA in the fnst place; that is,
why do participants in PA feel the need to constantly endorse each other and to reinforce
their collaborative alliance?

More generally, the structural and linguistic analyses of earlier chapters did not
address the central (functional) question of why such structures and linguistic phenomena
occur in the ﬁrst place. That is, what is it about the context, topic, and the goals and
motives of participants which combine to produce such structures and surface linguistic
detail? The following questions, which were hinted at in previous chapters, will serve to

guide the discussion of this chapter:

1) why does PA have the structure that it has?
a) why do participants in PA bother to contribute supports for a
position that they already accept?

1b) why do participants bother to import the positions and supports

180

of an absent antagonist?

2) why call this speech activity “pseudo-argument” ?

The answers to these questions involve appeals to nonlinguistic matters, and carry us into
areas such as social psychology, rhetoric, and sociology. Nonetheless, it will be argued
that many answers to be provided to the questions above are ﬁrmly grounded in previous
discourse analytic research and are empirically conﬁrmed by the data themselves.

Before attempting to answer the questions above, let’s brieﬂy review some of the
observations about the conversations which comprise the data which were brought out in
previous chapters. To begin with, participants in the conversations of the data of this work
appear to be. involved in a collaborative activity where they work together to construct
arguments. Participants view themselves as allies, and everyone seems to participate with
some degree of enthusiasm. In their collaboration, participants constantly seek and provide
the agreement, endorsement, and ratiﬁcation of each other. Participants appear to share the
same sorts of beliefs — they do not disagree with each other, but rather enthusiastically
signal agreement at every turn. Subjects are friends in an informal setting. The content of
the conversations revolves around issues of race, most of which are relevant to campus
life. Participants are aware that the issues involved are controversial and that some of their
views might be considered “racist” by others. Argument structures (CA1 and CA2) are
employed despite the lack of anticipated disagreement. That is, there is no obvious

explanation for why argument structures are being used, given a discourse analytic account

181

of argument.

It will be shown in the following sections that the answers to the questions raised
above will take us a long way towards a better understanding of the function of PA. It will
be argued that much of the structure of PA (including the structure of CA] and CA2) is due
to identity work on the part of participants in an effort to forestall negative inferences by
others and to project an image of rationality, objectivity and fairness. That is, it will be
shown that argument structures are employed by participants in PA primarily for the
purpose of face-work and only secondarily for the instrumental purpose of persuasion (as
in OA). It will later be argued that beyond this facade of fairness and rationality is an
activity whose main purpose is to allow participants to simulate, exchange, reinforce and

practice beliefs and arguments in a nonhostile environment.

6.1. Balancing instrumental and identity goals.
Goffman (1967) observed that individuals do not just go about serving their own
instrumental goals in interaction, but at the same time maintain a vigilant concern for the

“face” of themselves and others. Goffman (1967) wrote:

The individual must rely on others to complete the picture of him of which he
himself is allowed to paint only certain parts. Each individual is responsible for the
demeanor image of himself and the deference image of others, so that for a
complete man to be expressed, individuals must hold hands in a chain of ceremony,
each giving deferentially with proper demeanor to the one on the right what will be
received deferentially from the one on the left (pp.84-85).

According to Goffman, there is a preference to maintain a positive public image (“face”) in

182

interaction, both for ourselves and others, while pursuing overt objectives. Brown and
Levinson (1978) elaborated Goffman’s notion of “face” in their theory of politeness which
embodies the notion that interactants use linguistic politeness as a form of social currency in
the carrying out of sometimes face-threatening actions. The prevalent, universal concern of
individuals for protecting their own self-image and the image of others plays into every
interaction and is a powerful force in the rrrinutia of everyday discourse.

I would like to propose that it is this concern for identity and self-image that
motivates the structures of PA and answers several of the questions posed above. It will be
argued in what follows that the structures of PA stern primarily from a strategy by
participants which allows them to achieve their instrumental goals while at the same time
forestall any negative attributions and inferences by others. This functional explanation

will appeal to the nonnorrnative nature of the content of the conversations coupled with

strong social pressures against expressing what might be considered “racist” views.1 It

will be shown that PA structures are designed to allow participants to put forward their
ideas without incurring negative inferences.

That particular discoursal strategies may be employed to satisfy at the same time
both instrumental and identity aims has been noted before. For example, with regard to the

issue of identity work in (ordinary) argumentative discourse, Kline (1987) notes that:

Since both parties simultaneously pursue both instrumental and identity aims,
agreement on situational identities is constantly negotiated and reafﬁrmed. Thus a
crucial problem for communicators becomes one of using strategies that allow for
the accomplishment of instrumental goals while simultaneously actualizing a desired
identity for oneself and other (p.243).

183

The work of Kline shows that in OA, which is inherently face-threatening, a variety of
rhetorical strategies are employed by arguers which allow them to balance their
conversational goals (to alter the adversary’s beliefs) with identity goals (to avoid affront,
to present a positive self-image, etc.). The work of Benoit (1987) points out that because
arguments may be face-threatening to participants, identity goals provide the impetus for

resolution of any 0A:

Because relationships can not withstand continued assault on relational identities,
arguments move naturally to repair so that restoration of shared relational
understandings may be possible and the conversational norm of reciprocity as
regards orientation to face may be reinstated (p.145).

In the examples which follow, it will be shown how both lower level and higher level
strategies are employed in discussions about race which allow a speaker to balance

competing instrumental and identity goals. Van Dijk (1987) frames the issue as follows:

...prejudiced talk about minorities, among other things, has the overall goal of
negative other-presentation, while at the same time preserving a positive impression
(or avoiding the loss of face). These goals are sometimes in conﬂict, for instance,
when social norms do not allow uninhibited negative talk about minority groups.
Therefore, expedient strategies are in order to reconcile these real or apparent
inconsistencies. These strategies are accomplished by sequences of moves that try
to realize both goals as effectively as possible, for instance, with a maximum of
negative other-presentation and a minimum of negative self-presentation. In other
words, speakers persistently try to manage or control the social inferences the
recipient is bound to make about them on the basis of what they say...

The ﬁrst section will discuss work done on the use of disclaimers in talk about race

(Schiffrin, 1980; Verkuyten et a1, 1994), and the second section will discuss work by van

184

Dijk on the use of (higher level) argument structures in talk about race.

6. 1 . 1 . Disclaimers.

One lower level strategy which is sometimes employed to balance conﬂicting
instrumental goals and identity goals is the use of disclaimers (Hewitt and Stokes, 1975).
According to Schiffrin ( 1980), disclaimers are evaluative brackets which preface a
controversial or nonnorrnative assertion and allow an individual to “brieﬂy counteract the
projection of a questionable and unfavorable self potentially associated with an
objectionable position...” (p.229). Typical examples (with regard to issues of race and
ethnicity) mentioned by Verkuyten et al. (1994) are; “I’m not prejudiced, but...,” and “I
have nothing against foreigners, but...” According to Verkuyten et al., “These kinds of
phrases are usually understood to be “rhetorical maneuvers or a double strategy of positive
self-presentation, and expressing subtle prejudicial views” (p.289). However, they go on
to argue that the use of such disclaimers does not necessarily indicate a conﬂict between
deeper prejudiced beliefs and social norms, but rather may simply indicate an awareness on
the part of the speaker that his or her assertion may trigger negative attributions on the part
of others. Schiffrin (1980) echoes this point; “This means that even speakers who do not
intend for their talk to present them as persons who are representative of stereotypical
positions, behaviors, or social categories, they are aware that this may be the effect ”
(p.230). The point to draw from this is that when speakers wish to say something

(instrumental goal), but believe saying it may lead to negative inference given its

185

controversial nature, one way to ward off the negative attribution (identity goal) is to
preface the position with a disclaimer.
Disclaimers are common in the data of this study. Consider the following example

in which C. prefaces his conclusion with a disclaimer in lines 2-3:

(1)

C: = because of how dangerous it was - because
these - these quote-unquote Blacks Ye:- you
can call me prejudiced or WHAT, but y- after
like going to school for ONE YEAR, I’ve never
felt the same about Black people since. -
they're just nzutty.

(numerator—1

Here, C. is aware that his bald pronouncement of dislike for Black people could lead to a

“racist” attribution, and he attempts to diffuse that attribution with “-you can call me

prejudiced or WHAT,”2 Here, C.’s disclaimer attempts to counteract the projection of a

negative attribution. Other disclaimers which occur in the data are; “This may sound
prejudiced, but...,” “I’m not prejudiced at all, but...” That disclaimers such as these are
so common in the data suggests that participants are aware of the controversial nature of
their views and of the content itself; they are on the lookout for things they might say which
could reﬂect badly on them, and on some occasions where such things do occur, they try to

minimize the effects with disclaimers. Here then is one low level verbal device which

contributes to the balancing of instrumental and identity goals in talk about race.3

186

6.1.2. Higher level strategies.

It will be argued here that in addition to low-level local strategies which primarily
serve to forestall negative inferences, there are high-level global strategies which serve the
same purpose. I follow van Dijk (1987) in claiming that higher level argument structures

themselves may be used to ward off negative attributions:

The expression of delicate of controversial social opinions in conversation is
routinely expected to be backed up with arguments... Within the combined
strategies of positive self-presentation and negative other-presentation, such
arguments have the fundamental function of protecting the speaker against
unwanted inferences about his or her ethnic attitudes (p.76).
In this way then, according to van Dijk, the provision of supports for positions in talk
about race is fundamentally to protect the face of the speaker.

However, van Dijk’s contention that arguments are used “fundamentally” to protect
the speaker’s face does not rule out the possibility that the speaker provides supports for
the ordinary reasons — to prevent disagreement by the bearer and to persuade the bearer.
That is, van Dijk’s point appears to preclude attributing any real substance to the arguments
provided in prejudiced talk, since in his view, arguments are provided not (fundamentally)
as justiﬁcations for positions, but rather as shields for negative attitudes. Furthermore,
given the sorts of data he is working from, there is no way to determine the “fundamental”
impetus for the provision of arguments. Van Dijk’s data consist of one-to—one interviews
between (Dutch) strangers — a White subject and a White interviewer who are meeting for

the ﬁrst time and have little idea of each other’s beliefs and attitudes. Certainly in this

setting, one might expect arguments to be given by the subject since he or she could not be

187

sure of the views of the interviewer, and therefore would be expected to take precautions to
prevent disagreement by providing supports for positions taken. Thus, in these settings
which compose van Dijk’s data, there is no way to empirically determine whether
arguments are primarily given to protect the speaker from negative inference, or to avoid
anticipated disagreement, or perhaps both. My point is that for van Dijk to ignore the latter
possibility is not justiﬁed by the data he is arguing from. Nonetheless, it will be shown in
the following section that van Dijk’s claim can be empirically tested with the right sorts of

data, and that his claim is indeed a signiﬁcant one with regard to talk about race.

6.2. The function of collaborative argumentl

Here an answer to question 1a will be proposed. Why do participants in PA bother
to contribute supports for a position that they already accept? The answer was already
alluded to in the previous section —— supports are given because they help to ward off
negative inferences about the speaker. It was established in chapter 4 that the participants
in pseudo-argumentative discourse are in accord with each other from the start with regard
to the positions they are taking. Further (linguistic) evidence for this conclusion was
provided in chapter 5. Additional evidence not mentioned so far that the participants do not
provide supports in anticipation of disagreement is that even after several supports are
given for a position, with at least one contribution from each participant involved in the
conversation, still more supports are offered (see chap. 4). This provision of additional
supports cannot be explained as being due to the anticipation of disagreement since each

participant has already signaled agreement with the position at issue (by supporting it).

188

Therefore some other explanation must be sought, and that is that by providing
supports (building CAls), the participants are eager to protect themselves and each other
from negative attributions (by each other). That is, the participants in the data are anxious
to view themselves in a positive light while at the same time express their negative beliefs
and attitudes. Therefore, they collaboratively provide supports for positions as way of ,
shielding themselves from the negative inferences which could result from their own
internalized anti-racist norms.

There is ample evidence in the data that the participants are anxious to avoid the
attribution of “racist” given the content of their talk. Consider the following exchange

which follows the particularly passionate CA] analyzed in chapters 4 and 5:

(2)

38 K: I feel really BAD about all this, because
39 like we sound like RAcists or whatever,
40 and I really don't think I am,
I
41 B: I don't think I'm.being- I don't know - I
42 really don't think I'm racist, I just think that
43 K: It's just a very unfair society that we're
44 living in today. Especially on THIS campus and on =
I
45 B: ExACle. I-

46 K: = other college campuses,

Here K. changes the topic to announce that the content of the previous talk has made her
“feel really bad,” since “we sound like RAcists or whatever,”. That K. bothers in the ﬁrst

place to step outside of the talk to explicitly discuss the negative inferences which could

189

arise from it is powerful evidence that she is aware of and oriented towards anti-racist
norms. Clearly she has become uncomfortable with the talk and seeks to directly dispel
any negative inferences which might arise among the participants by saying; “and I really
don’t think I am,” By shifting from the lst person plural pronoun “we” in “we sound like
racists or whatever,” to the singular “I,” in “I really don’t think I am,” K. perhaps
unintentionally implicates that the other participants might be racist, to which B.
immediately responds; “I really don’t think I’m racist.” Here B. joins K. in showing her
orientation towards anti-racist norms and her concern to avoid negative attribution.

It should be noted that the tendency to explicitly dissassociate oneself from potential
racist attribution is a common element of modern racism (Billig, 1988; Gaertner and
Dovidio, 1986). According to Gaertner and Dovidio, since, for modern racists, negative
feelings and beliefs about minority groups are usually excluded from awareness, “When a
situation or event threatens to make the negative portion of their attitude salient, aversive
racists are motivated to repudiate or dissociate these feelings from their self-image...”
(p.62). The statement of repudiation made by K. in lines 38-40 in the previous excerpt
occurs at a likely moment in the interaction where the prior talk appears to have become
openly hostile and pejorative.

Consider another example from another conversation in the data where G. shows

his concern that his assertion not lead to negative inference by A. and T.:

190
(3)

1 G: [[Yeah. The minority aids?= They have them right
2 next to- right— They have the residential advisor,
3 AND they have the minority ad- aids and advisors or
4 whatever, - - I can understand, It’s GOT to be =

I
5 A: Uhhuh,
6 G: = harder to adjust for them, in some ways. SOME
7 of them. I’m not saying ALL Black- I am
8 generalizing right now, and it probably comes off
9 as pretty - rude and obnoxious.

10 T: It’s because you're a racist pig!

11 ((T and A laugh loudly))

In this case, G. goes out of his way to manage the impressions of A. and T. and avoid a
“racist” attribution. His initial (somewhat inoffensive) position in lines 4 and 6 (“It’s GOT
to be harder to adjust for them”) is afterwards mitigated, qualiﬁed and apologized for to
such an extent that T. sarcastically mocks him in line 10 with “It’s because you’re a racist
pig!,” and both T. and A. break out in laughter. Perhaps the reason for their mirth is that
they are amused by G.’s overly defensive stance over a relatively innocent statement — T.-
and A. ﬁnd humor in the unnecessary extremes that G. will go to to avoid appearing racist.
Thus, it is claimed that participants in PA collectively provide supports for a
position in CAls not to forestall disagreement, but to forestall negative inference.
Therefore, the function of CA] is to balance both instrumental and identity goals — to
allow the expression of nonnorrnative positions while at the same time protecting those
involved from their own negative attributions. Note that the conclusion drawn here is not

subject to the same criticism leveled at van Dijk — the data discussed in this paper involve

191

conversations between friends in an informal setting, and so it is easier to rule out the
criticism (coupled with the other evidence mentioned above) that supports might be
provided for positions merely to avoid disagreement, as in OA. Generally, friends are
familiar with and share each other’s ideas and attitudes, whereas strangers thrown into an
idea-based conversation on a controversial issue will probably take no chances — better to
support what you say than to chance open disagreement. Other problems occasioned by

van Dijk’s data collection method will be discussed in a later section.

6.3. The function of collaborative argument;

Now we may address another of the questions posed earlier; why do participants
bother to import the positions and supports of an absent antagonist? That is, what function
or functions are served by bringing in the views of “the other side” within CA28? I would
like to suggest that a similar motive of identity management is involved for participants —
by presenting the views of an antagonist, participants project an image of fairness, balance
and objectivity and therefore shield themselves from negative inference. A CA] has the
undesirable effect of appearing one-sided, and biased since there is no opposition to
balance the uniﬁed stance of the participants. Importing the antagonist’s views thus helps
to dissolve the appearance of bias and unfairness, and so gives the impression that
participants are simply seeking objectively arrived at conclusions, tested by a fair
consideration of both sides of the issue at hand.

By presenting the ‘other side,’ participants in PA attempt to dissolve the impression

192

that they are simply airing their negative attitudes. It should be obvious from the data
considered up to this point that the participants in the conversations wish to appear rational
and objective. It is for this reason that they go to such lengths to supports their positions
although they do not anticipate disagreement from each other. But the presentation of only
one side’s views (no matter how well fortiﬁed with supports) leaves the participants
vulnerable to the impression of bias and further to the undesirable attribution of ‘racist.’
Therefore, a token attempt is made to supply the arguments of the antagonist. Clayman
(1992) notes that the practice of presenting the views of “the other side” is a staple of news
interviewers; “By counterbalancing [E’s [interviewee’s] opinions with divergent and
contrasting points of view, [Rs [interviewers] give voice to “the other side” of controversial
issues. This practice is consistent with traditional standards of fairness in broadcast
journalism...” (p. 176). In the same way, by importing the views of the antagonist,
participants in PA boost the appearance of objectivity and rationality by making it look as if
they are simply seeking the truth, considering fairly both points of view before arriving at
their conclusions.

Blair (1987) distinguishes between ordinary argumentation which arises out of
disagreement between participants and what he calls “inquiry” which is typically carried out
by one person alone who reviews and evaluates different sides of an issue with an aim to

discern whether some position is acceptable:

Inquiry so conceived still entails argumentation, for the inquirer both ﬁnds or
generates arguments for the point of view in question (and in that respect plays the
role of proponent), and also ﬁnds or generates arguments against (1) those
arguments, and (ii) that point of view (and in that respect plays the role of critic).
(p. 19 1)

193

In the instances of CA2 in the data, it seems that something akin to inquiry is occurring in

that some position is tested by an apparent consideration of both sides of the issue. In
Blair’s terms, inquiry is a “dual-role” activity in which someone plays the role of both the
protagonist and antagonist, and in this respect, CAZS resemble inquiry.

However, it is clear that the notion of inquiry in Blair’s sense is inadequate for
describing CA2 in that it does not capture the main purpose of the talk — the purpose of
the talk is not to objectively test a position against opposing arguments in order to arrive at

a fair conclusion, but rather to simulate and fabricate an argumentz in which the outcome is

determined from the start. Blair concedes that while an inquiry aims to test a certain
position against opposing arguments in order to arrive at an objective consideration of its
acceptability, it is also possible that this aim can be thwarted; “On the contrary, one person,
alone, if she is not open-minded and even vigorously self-scrutinizing, will not succeed in
conducting a very reliable inquiry.” (p. 192). The CA2s in the data hardly involve unbiased
presentations of antagonist’s arguments. Rather, the presentation of the antagonist’s
arguments amount to no more than a minimal token more aimed to serve impression
management than objectivity.

Consider once again an excerpt from the data which involves a CA2:

(4)
l B: They totally have this- y- Exactly. And- You
2 know with Asians, you don't see Asians like =
I
3 K : I 1
4 B: = pulling an attitude, and I don't think I— I
5 have like Asian friends and - I- I don't think I've
6 Ever seen an Asian - pull— you know and like- Black

194

7 people bring up like "Well - you owe it to us
8 because - you know - you had us as slaves - =

I

9 K - Yeah but

10 how long:

11 B: = thousands of years ago, and blah blah blah:,

12 and you OWE it to us. Blah Blah" Well what about -

13 what about when we took like during what was it

14 World War Two, when we took all the Japanese, and

15 stuck them in prison camps in California. I mean

16 what about THAT? You don't see Asian people still

17 all fired up about THAT, - I mean it is totally =
I

18 K: Yeah and if that IS the

19 case,

20 B: = WRONG but- Just- I mean it was a mistake in
21 the PAST, and it shouldn't - like - I don't know,
22 it- it- YEAH it does effect us, and YEAH it was
23 WRONG, but - it wasn't ME, it wasn't even my DAD.
24 It wasn't even my GRANDpa. It was like-

I
25 C: And they're-
26 now they're STILL - you know basically treated =

I
27 B: ( ) ago.

28 C: = not just equal, but they get a— the advantage
29 of - you know in a lot of situations so they can't
30 complain that they're not treated equal now.

In this excerpt, the argument of the antagonist is indeed presented (in lines 7-8 and 11-12),
in reported speech; “Well - you owe it to us because - you know - you had us as slaves -
thousands of years ago, and blah blah blah:, and you OWE it to us. Blah Blah.” It would
be difﬁcult to make a case that B.’s presentation of the antagonist’s position is unbiased
here. On the contrary, the direct quote itself is loaded with negative evaluation (see chap. 5
for discussion). So although the speaker bothers to supply an argument on behalf of the

conjured antagonist, it does not appear to be done with the aim of neutral, objective

195

consideration. Rather, it is a token gesture which functions to ward off the attribution of
one-sidedness. Note that in this case, and in all cases in the data, the provision of the
antagonist’s argument is minimal, that is, it never goes beyond giving one or two supports,
although certainly other supports are available. An objective presentation would more
likely present more than one or two supports for each position. In addition, the supports
provided are not necessarily the most persuasive or effective ones which could have been
selected; it is often the case that the imported supports are transparently weak, stereotypical
and sometimes it is questionable whether they have ever actually been provided before by
any antagonist. Thus, it seems to be the case that although the views of the antagonist are
supplied, they serve some other purpose than the pursuit of objectively arrived at
conclusions.

What other function then, besides giving the appearance of fairness, balance and
objectivity, does importing the antagonist serve? The data suggest that participants import
an antagonist’s argument in order to set up a straw man which can be collaboratively
disputed. That is, just as in CA in which one tries to destroy the antagonist’s arguments in
order to win, CA2s set up and destroy the imported antagonist’s arguments. Furthermore,
the antagonist’s arguments are presented in such a way that they are designed to be
destroyed in order to allow the participants to “win.” To put it another way, in CA2s the
imported arguments are selected for their weakness so that the participants may win the PA.
In addition, it is likely that the participants select imported arguments to which they already

have in advance some ready-made attack. For example, there are no cases in the data in

196

which an imported antagonist’s argument is given and no retort/dispute is immediately
provided.

In sum, there appear to be two main functions of CA2s The ﬁrst has to do with
identity work —-— by presenting the views of the antagonist, participants ward off the
potential impression that they are biased and simply airing negative attitudes. By providing
the point of view of the other side, the discussion appears fair, balanced and objective. As
Tannen (1989) notes, absent persons are in some sense “nonpersons,” and are therefore

treated as resources for participants:

absent persons are treated without consideration because they do not
exist in that context; in other words, in contexts in which they are absent,
they are not perceived as persons, that is, not perceived as potentially
affected by the acts of that context. Rather, absent parties are simply
resources for the facework of the immediate context. (p.109)

It should be apparent from what has been said so far that absent parties (the antagonists) in
CA2$ are treated as resources for facework by participants in PA.

The second function of CA2s has to do with an instrumental goal — setting up the
antagonist’s position in order to dispute it. Despite appearances, CA2$ are designed to
defeat the antagonist from the start; weak, stereotypical imported supports are selected with
ready-made disputes. Winning an argument (whether real or pseudo-) includes not only

presenting one’s own arguments, but also destroying the arguments of the opponent.

197

6.4. Summary — why call it “pseudo-argument”?

Now we may summarize the functions of PA by accounting for the name chosen
for the speech activity described and analyzed in this work. There are two senses of the
word “pseudo” which make it an appropriate term to use in describing the kind of talk
analyzed in this work; ﬁrst of all, “pseudo” may mean “false” or “deceptive,” that is,
something intended to do something other than what it pretends to do. On the other hand,
“pseudo” may also mean “simula ”, or “imitative,” that is, something which attempts to
take on the appearance of something else. It is claimed here that the name “pseudo-
argument” is appropriate because the discourse it describes incorporates both of these
senses.

We have seen throughout chapters 4-6 that PA in many ways simulates OA, but is
defective or at least distinct in some structural, linguistic or functional way. First, in CA1
participants jointly build an argumentl in support of some position. This joint effort
simulates CA in that some position is supported. It is defective or distinct in that in OA one
and the same individual must present both the position and its corresponding support(s); in

CA1 this need not be the case In addition, in ordinary argumentl supports are only
provided when a speaker anticipates disagreement on the part of the hearer (Jackson and
Jacobs, 1982, 1986); in CA] supports are not provided in anticipation of disagreement, but
rather to forestall negative inference. In ordinary argumentl, the provision of support for a

position leads to the irnplicature (for the bearer) that the speaker anticipates or wishes to

forestall disagreement on the part of the bearer (Jackson and Jacobs, 1982). It might be

198

argued then that this same irnplicature serves the purpose in PA of lending the appearance
of rationality and skepticism among the participants even though it is plain to all involved

that no disagreement is likely. Finally, in ordinary argumentl the provision of supports for
a position is primarily an attempt to persuade a presumably unpersuaded other; in CA1,
persuasion is not the primary goal. Thus the simulation of ordinary argumentl in CA1

allows participants to borrow elements (and the effects of those elements) from ordinary

argumentl while at the same time accomplish goals quite distinct from those of OA.

In addition, PA earns its name in a more obvious and signiﬁcant way — it involves

the simulation of ordinary argumentz. That is, in CA2$ we ﬁnd participants jointly creating
and sustaining the appearance of an argumentz by importing the positions and supports
(argumentsl) of an absent imagined antagonist and subsequently disputing those
argumentsl. Of course, in ordinary argumentz the antagonist is present and is free to
supply his or her own arguments] or to dispute the arguments] of the protagonist. The
simulation of an argument; in PA allows participants the beneﬁt of having considered the

views of the antagonist and therefore of shielding themselves from the appearance of bias.

The goal of ordinary argumentz is for participants to come to some resolution, where each

participant’s goal is to persuade the other side to change his or her view. On the other
hand, the goal of each participant in CA2 is not to change the opponent’s view (the
opponent is absent), but rather to reinforce the ‘rightness’ of their own side’s views. This

is accomplished by jointly disputing each and every argument presented on behalf of the

199

antagonist — there are no instances in the data where the absent antagonist ‘wins’ in PA.

Also, unlike in ordinary argumentz in which very often no resolution is reached (Willard,

1987), in PA resolution is reached in every case, with the victory of the protagonists and
the vanquishment of the antagonist, often in the form of some concluding statement.

The data analde which involve CA2s also show why they are ‘pseudo’ in the
other sense of “false” or “deceitful.” This is because although the participants do import
the antagonists in an effort to appear fair, the cards are stacked against the antagonist from
the start. It is unlikely that the protagonists will argue forcefully or at any length on behalf
of the antagonist, and the positions and supports supplied are typically weak and
stereotypical. In addition, the antagonist gets no chance for rebuttal. After the subsequent

dispute of the imported argumentl, the CA2 is magically resolved. So although an

argumentg may be simulated in CA2, the simulation has a falseness about it in that it does

not actually attempt to recreate a fair argumentz, appearances to the contrary.4

In sum, PA earns its name because it simulates GA in various ways in order to
borrow certain appearances which derive from ordinary argumentative discourse. In other
words, PA looks like OA and is meant to, since the use of argument structures (including

argumentl and argumentz) lends the appearance of fairness, objectivity and rationality, and

wards off negative inferences about the participants. In addition, PA is “false” in the sense
that despite these appearances, the discourse is anything but fair and objective. It has been
argued that participants engage in PA not primarily for the purposes of persuasion, but

rather in order to reinforce their mutually held nonnonnative beliefs and perhaps more

200
importantly to reassure each other of their (perhaps threatened) ideological alliance and

solidarity. The use of argument structures in pseudo-argumentative discourse then is false
in the sense that it accomplishes something other than what it purports to accomplish on the
surface. Rather than actually providing fairness and objectivity, it mainly functions to
control negative inference in an essentially one-sided, attitude-laden discussion. PA is, in
sum, a type of discourse which allows participants to secure their alliance and strengthen

mutually held beliefs in the face of strong social pressures which threaten those beliefs.

6.5 Conclusion

This concluding section will be divided into three parts. The ﬁrst part will attempt
to address several possible problems and objections. The second part will attempt to
evaluate the contribution of this work to the developing area of discourse analysis. It will
be pointed out that there are speech activities which borrow structures from other activities
for speciﬁc purposes, e.g., to appear objective, rational, etc. It will also discuss how an
analysis of a larger discourse type can focus separately on underlying structure, linguistic
form, and function, while at the same time viewing these aspects of the discourse as
integrated in practice. Finally I will take the position that the structure and linguistic
features of a discourse stem from the functions to which they are put by providing
examples from the analysis of PA.

The third part of this concluding section will attempt to show how this work makes

a contribution to our understanding of talk about race. Clearly, this applied aim was

201

implicit throughout this work, and in this ﬁnal section I would like to make several
suggestions about how a discourse analysis such as the one carried out here can deepen our
understanding of the discoursal processes involved in the transmission, reinforcement and
practice of beliefs and arguments concerning the topic of race in everyday face-to-face

interaction.

6.5. 1. Possible objections.

One might object that by including both CA] and CA2 in the larger speech activity
of PA, I make the same mistake as early argument theorists who failed to distinguish the
“two senses of argument” (O’Keefe, 1977). That is, just as early theorists confused issues
by talking about argumentation rather than argumentl and argumentz, it could be claimed
that PA elides and confuses the distinct patterns of CA] and CA2. Thus, just as one might
reasonably ask whether “argumentation” is deﬁned by argumentl or argumentz, or both,
one might also ask whether PA is deﬁned by CA] or CA2, or both.

I take the position here that one may appeal to the CA treatment of argumentation to
resolve this issue. Recall that the CA approach of Jackson and Jacobs (1982) attempts to
tie the distinct senses of argument into a single framework — argumentation in
conversation arises either when one party objects to what another has said (argumentz), or

else when one party anticipates disagreement and seeks to forestall it by providing

unsolicited arguments (argumentl). Under this approach, the two types of argument

involve the suppression of disagreement, whether that disagreement has been expressed or

202

is only anticipated. Thus, while the two senses of argument must be kept conceptually
distinct, they are related in an integral way. Further, based on this connection, it seems
justiﬁed to speak of “argumentation” as that type of activity which involves either
argumentl, argument;, or both, since both kinds of argument involve the suppression of
(actual or potential) disagreement.

I would like to claim that in a similar way, one is justiﬁed in speaking of PA as that
type of activity which involves either CA1, CA2, or both. However, what ties together
CA] and CA2 under the label of “pseudo-argument” is not the suppression of disagreement
(as in ordinary argumentation), but rather the suppression of negative attribution. Recall
that CA] S involve the collaborative supply of supports, whereas CA28 involve the supply
of an antagonist’s positions and supports which are then collaboratively disputed. It was
argued that both CA] and CA2 employ argument structures for the same purpose — to
forestall the impression that participants are biased, racist, etc.

Thus, just as one may speak loosely of “argumentation” while keeping in mind the
two types of argument, one may also speak loosely of “pseudo-argument” while keeping in
mind the two types of PA: CA] and CA2. And just as argumentation may be said to have

occurred in cases where either argument] or argument; or both have occurred, it may be

said that PA has occurred where CA] or CA2 or both have occurred. One might further

object that CA2 is somehow a better example of PA than CA1, just as argument; seems to

be a better example of ordinary argument, given the intuitive notion that argumentation

requires some kind of expressed opposition. And thus, it might be claimed that CA1 alone

203

does not qualify as PA in the same way as argumentl might not qualify as “real”

argumentation. However, given the CA framework adopted here, there does not seem to
be any justiﬁcation for making one type of argument primary or more essential than the
other, since both spring from same functional source — the suppression of actual or
potential disagreement. Similarly, there does not seem to be justiﬁcation for setting either
type of collaborative argument above the other, since both arise in response to the same
functional pressure -— the suppression of negative attribution. Thus, the term “pseudo-
argument” does not confuse two distinct types (CA1 and CA2), but rather is a term which
may be loosely applied when either CA] or CA2 or both have occurred, in the same way as

the term “argumentation” may be used when either argument] or argument; have occurred.

Next, one might object that PA is not very different from the case of academic
writing in that in both types of discourse one or more parties put forward arguments and
very often import and then attempt to destroy the opposing views of some absent
antagonist. Indeed, it is often the case that academic writing, like PA, involves a
collaborative effort on the part of more than one party. In addition, as in PA, academic
writing often involves the second function of CA28 described above — the positions and
supports of the antagonist are almost always refuted and sometimes are crafted with such a
goal in rrrind.

However, I would argue that PA and academic writing are distinct activities on
several grounds. First and most obviously, PA is different from academic writing in that it

occurs spontaneously in conversation, whereas academic writing is preformed and written.

204

Second, and most signiﬁcantly, the two types clearly differ on functional grounds. The
primary goal of academic writing is to persuade one’s readers, whereas we have seen that
the primary goal of PA is not persuasion, but rather the reinforcement and practice of
beliefs which are already acceptable to all participants. Academic writing assumes the
skepticism of its audience and therefore is akin to what Willard (1989) calls “substantive
argument,” where arguments arise naturally due to the general inquisitorial nature of the
discourse. Thus, in academic writing, one presumes skepticism and so one must be
maximally explicit, providing arguments even where disagreement is not anticipated. In
PA, on the other hand, one presumes the agreement of the audience, which continually gets
reinforced in the give and take of the conversation.

PA is distinct from academic writing in another way — in PA, it is often the case
that the imported positions and supports of antagonists are stereotypical and very easily
refutable. For participants in PA, this is easy to get away with, since they are aligned in the
ﬁrst place and are unlikely to question the accuracy or representativeness of the imported
views. On the other hand, since academic writing presupposes the skepticism of its
audience, authors may less easily misrepresent the views of their antagonists. So although
setting up straw men in academic writing is not uncommon, it can be said that authors in
this type of discourse are more accountable than in PA for how they represent the views of
an antagonist.

Finally, one might object that given the lirrrited amount of data analysed in this

work, there is no assurance that PA is not just a local phenomenon, limited to talk between

205

college students about issues of race on a particular college campus. Indeed, it does not
seem justiﬁed at this early stage of research to make any general claims about how often,
where, and with regard to which topics PA takes place. Future studies should attempt to
reduplicate the conditions under which PA may arise, in different locations, between
different kinds of participants, and with a variety of topics. In addition, an investigation of
where PA does not occur could also lead to signiﬁcant results. For example, one might
predict that PA structures would not occur in conversations where participants share anti-
racist beliefs, and such conversations could be analysed to test this prediction.

Thus, while this work has attempted to establish the existence of a type of discourse
called “pseudo-argument,” it is admitted that much more work needs to be done to conﬁrm
it and to determine its details and limits. It is hoped that this work has at least pointed out

that these avenues of research of worthy of pursuit in the future.

6.5.2. Contribution to discourse studies

6.5.2. 1. Pseudo-discourse. It is improbable that PA is the only speech activity which
borrows structures from another activity in order to accomplish some end. That is, it is
likely that there are other types of talk which achieve their distinct functional ends by
appearing like another type of talk. We saw that PA borrows structures from ordinary
argumentative discourse in order to create the appearance of rationality, objectivity and
fairness, and to ward off negative inferences. If this phenomenon is not limited to PA,

then we might proﬁtably explore the ways in which other types of (pseudo-) discourse

206

simulate other speech activities for some effect. It seems plausible that this kind of
exploration could yield valuable insights about how speech activities may be related, and
about how some speech activities may serve as resources for others.

One ﬁrst interesting example of pseudo—discourse is the “infomercial,” which in
some ways seems similar to PA. An infomercial is a commercial advertisement on
television (popular several years ago, but perhaps now less commonly used) which
simulates a news broadcast. It usually involves a well-dressed anchorperson (actor)
announcing in news-like fashion some newly arrived technological marvel which will
undoubtedly beneﬁt mankind in some way. Often the infomercial will involve other
characters playing the role of reporter, person on the street, etc. In this way, an
advertisement masquerades as a news broadcast even though it is clear to all involved (the
actors, producers and audience) that some product is being sold, and that a news broadcast

is being simulated. As in PA, one speech activity (the infomercial) borrows the structures

and overall organization of another speech activity (a news broadcast).5 In addition,

although this has not been studied, there appear to be linguistic features which are
borrowed from the news broadcast speech activity as well (prosodic, syntactic, lexical,
etc.).

Of course, the borrowing of structures from the news broadcast has a function, and
that is to generate the appearance of objectivity, which is normally associated with news
reporting. By simulating a type of talk which is commonly believed to be nonbiased and

neutral, the infomercial appears to take on these features, and in doing so attempts to

207

convince potential customers that what is being sold is more than snake oil; that is, that the
company behind the product is providing it for the beneﬁt of the customer and not merely
for its own ﬁnancial gain.

Thus, the infomercial, like PA, is a kind of ‘pseudo-discourse’ which simulates
some speech activity in order to create a certain appearance which is normally associated
with the simulated activity. In both cases, the aim is to achieve the appearance of
objectivity and neutrality and to ward off the appearance of bias. Further, as in PA, the
function of the infomercial is distinct from the speech activity it simulates; the function of a
news broadcast is to give the news, not to sell a product.

The next example of what I am calling pseudo-discourse is different from the ones
seen so far in that the appearance aimed for in simulating another type of discourse in this
case is not that of objectivity and neutrality, but rather authority and expertise. This is
when in everyday speech one person “lectures” another. Often this type of talk occurs in
asymmetrical relationships between parents and children, or between bosses and
employees. In this case, the speech activity which is simulated is that of a lecture given by
a teacher to a student in a school setting. In the pseudo-lecture (for lack of a better term) a
person borrows the organizational structures and linguistic features of a formal lecture in
order to appear authoritative and competent with regard to some issue. As with the other
kinds of pseudo-discourse discussed so far, the actual function of the pseudo-type is
distinct from the function of the simulated type; the aim of a pseudo-lecture might be to

scold, complain or insult, whereas the aim of a genuine lecture is to instruct, inform or

208

teach. That people are aware of this kind of discourse is evident in common expressions
such as “Don’t lecture me!,” or “I’m sick of your lecturesl.” It seems plausible that
borrowing features of a lecture provides an air of authority and competence on the part of
the speaker, an air which might be useful in reprimanding someone or inﬂuencing their
future behavior.

Another example of pseudo-discourse is the use of extended question-answer
sequences in conversation. In this case, the talk resembles talk which occurs in certain

institutional contexts such as the courtroom or the police station:

those ﬁnding themselves on the receiving end of a series of more than a
few questions are likely to respond by asking why they are being
“interrogat ” or cross-examined.” ...extended question-answer sequences
are so strongly associated with legal contexts that these latter may be
invoked to describe and complain about such sequences when they occur in
everyday conversational settings. (Atkinson, 1992, pp.207-208)

In the previous examples, it was shown that the function of the pseudo-discourse diverged
from the function of the discourse which is simulated. In the present example, however,
the function of extended question-answer sequences in ordinary conversation is arguably
much the same as in institutional contexts — to bring to light certain information which is
presumably kept hidden by the party being questioned. Nonetheless, in such cases the
association between extended question-answer sequences and the institutional contexts of
cross-examination or interrogation may invoke such contexts in ordinary conversation, a

result to the beneﬁt of the questioner who stands to gain all the rights and privileges ceded

to questioners in such settings (Atkinson, 1992; Drew, 1992).5 Thus, recasting the

209

activity as cross-exarnination or interrogation is a potent strategy for questioners in
conversation. It is for this reason that potential answerers-defendents in ordinary
conversation commonly reject the bid of the questioner to recast the activity by resisting
series of questions and answers, since they ﬁnd themselves rather unarmed and on the
defensive.

There are other examples of pseudo-discourse which could have been explored
here, such as the use of ‘preaching’ style in arguments or political speeches to take on the
appearance of righteousness or grace. One point to emerge from these examples, and
which should be pursued in future research, is that there are stereotypical associations
which people make between speech activities and particular features such as ‘objectivity,’
‘neutrality,’ ‘authority,’ or ‘competence,’ and pseudo-discourse allows individuals to proﬁt
in various ways from such associations. This point is reminiscent of Gumperz’ (1982)
notion of “metaphorical code-switching” in which meaning is transmitted by way of

stereotypical associations with the code employed.

6.5.2.2. Discourse analysis of structure, ﬁtnction and linguistic features. The analysis
carried out in this work has remained atheoretical, not subscribing to any particular
approach to discourse, but following a set of assumptions which are acceptable to most of
the approaches. Given the early stage of development of the vast area of discourse
analysis, this approach to analysis seems justiﬁed. In this work, I hope to have shown

how the separate analyses of the underlying structure, function and linguistic features of a

210

larger discourse unit may be combined to deepen our understanding of a particular speech
activity. In addition, since these aspects are in practice integrated within a discourse, the
analyses of structure, function and surface linguistic detail are interrelated and mutually
supporting. This allows the analyst to test and conﬁrm or disconfrrm hypotheses, for
example, about function by looking to surface linguistic features, or about structure by
looking to function. For instance, the hypothesis that a function of PA is to allow subjects
to reinforce their agreement could easily be disconﬁrrned by looking to the surface
linguistic detail for counterevidence; is there linguistic evidence of dispute or opposition
between the participants? In the same way, there may be functional or linguistic evidence
in support of underlying structures posited in analysis. For instance, the functionalist view
of PA as a collaborative enterprise is consistent with the structures of CA] and CA2,
posited on independent grounds (reconstruction), which accounts for particular
conﬁgurations of underlying units (positions, supports, and disputes) in relation to the
participants.

In addition to this three-sided analysis, this work has shown that any approach to
discourse should take into consideration the interactional framework in which the talk
occurs. If for example, the sequences of allowable contributions by participants in PA
were ignored, then it would be more difﬁcult to distinguish this speech activity from OA
since both types involve the same underlying units (positions, supports and disputes).
What distinguishes PA from OA is not the underlying units but rather how those units may

be combined and the constraints placed on participants concerning who may contribute

211

what and when.

6.5.3. On the applied side —— contribution to our understanding of race talk.

Although talk about race was used in this work to exemplify PA, it was not
proposed that PA was lirrrited to this content. On the contrary, it appears that PA may
involve any “delicate” issue in conversation concerning which participants share a similar
set of nonnorrnative beliefs. Thus while data involving talk about race in this work
provided an illustration of PA, the content of the discourse was not focused on for its own
sake. In this ﬁnal section, I would like to take a more applied view of the work carried out
here by focusing brieﬂy on discoursal processes in talk about race. Speciﬁcally, I would
like to suggest a few ways in which the analysis carried out in this work might make a
contribution to our understanding of the discoursal processes involved in the transmission,
reinforcement and reproduction of modern racism. It will be argued that earlier works
which pioneered the study of the reproduction of racism in everyday discourse (van Dijk,
1987; Potter and Wetherell, 1987) based their conclusions on data which unfortunately
provided a distorted and incomplete view of such discoursal processes because they did not
reﬂect the actual interactive settings in which such processes normally occur. In addition, it
will be argued that because of their data (which involved interviews between strangers),
they missed an opportunity to recognize PA as a signiﬁcant, common component in the

reproduction of racism.

2 12
6.5.3.1. Van Dijk’s work on the reproduction of ethnic prejudice in discourse. Van Dijk’s

pioneering work demonstrates convincingly how discourse analysis can be used as a tool
for examining how ethnic prejudice is reproduced in everyday conversation (1987) and in
the media (1993). It is van Dijk’s contention that ethnic attitudes and beliefs are “social
cognitions” which are acquired and diffused predominantly but not exclusively through
face-to-face interaction: “...social cognitions, in general, and ethnic attitudes, in particular,
are acquired, shared, validated, normalized, and communicated primarily through talk...”
(p.31). Because everyday talk is the primary medium through which such social
cognitions are communicated, van Dijk places discourse analysis at the heart of the study of
the reproduction of ethnic prejudice: “...discourse is in many respects the central element
in the processes of the interpersonal communication of prejudice, and discourse analysis is
a key method for the study of the cognitive and social structures and strategies that
characterize these processes” (p.30).

Therefore, van Dijk’s work is not just interested in the cognitive dimensions of
prejudice (e.g., how it is represented in memory), but also how prejudice becomes
manifest in face-to-face discourse; that is, much of van Dijk’s work concentrates on how
prejudiced discourse is organized at different levels of analysis, and how it is “...also
interpersonally and socially constrained in the communicative context” (p.12). This latter
aim necessarily involves an examination of interaction strategies such as positive self-
presentation (discussed earlier in this chapter) and persuasion, in addition to social goals,

norms and values. While a review of van Dijk’s fascinating and complex work goes well

213

beyond the scope of this paper, and while I subscribe to most of his aims, assumptions,
and conclusions, I would like to point out what appears to be a problematic ﬂaw in van
Dijk’s methodology.

Van Dijk’s work rests on 180 one-to-one interviews which took place either in
Amsterdam or San Diego. Each interview was between a student of van Dijk’s and a
resident of one of these two cities. In each case, both the interviewer and the interviewee
were White. The interviewers in Amsterdam mostly went door to door asking for the
interviews which took place within the homes of the interviewees. Most of the San Diego
interviews were carried out in public places due to the difﬁculties of being allowed to enter
a stranger’s home in the United States. The interviewees were all strangers to the
interviewers and were asked to participate in an anonymous, tape-recorded interview about
their opinions and experiences with regard to “this city” or “this neighborhood.” After
some time, interviewers normally introduced questions having to do with people in the
neighborhood and in many cases this led to some discussion of “ethnic” topics.

Van Dijk is careful to note that efforts were made to make the interviews as
spontaneous and as conversation-like as possible. For example, the interviewer avoided
the use of notes or written questions, and allowed the interviewees to take the conversation
in whichever direction they wanted. The motivation for attempting to make the interviews
as informal and natural as possible was so that the discourse of these data would not be
markedly different from natural conversations which actually occur between ethnic in-

group members; ‘...we have reason to assume that they [the interviews] are sufﬁciently

214

similar to spontaneous conversations to warrant conclusions about the nature of everyday
talk about ethnic minority groups” (p18).

In sum, van Dijk’s data consist of recorded one-to one interviews between a White
social scientist (student) and a White interviewee who are strangers to each other. It is then
assumed that the data are similar enough to natural conversations to warrant the conclusions
arrived at concerning everyday talk about minority groups by dominant ethnic group
members. I would like to take issue with this assumption — it is not the case that
interviews between strangers about a delicate topic sufﬁciently resemble naturally occurring
conversations to warrant conclusions about everyday talk involving this topic. This is
mainly because people do not speak to strangers the same way they speak to those they
know and are close to, especially when the topic at hand is a delicate one — van Dijk’s
assumption ignores the obvious sociolinguistic truism that talk to strangers is very different
from talk to people we know.

This difference bears directly on conclusions van Dijk arrives at concerning
strategies of positive self-presentation and the use of argument structures. As was
discussed earlier in this chapter, the realization of these interactional face-related strategies
might not be the same between strangers as between people who are involved in an
ongoing relationship. For example, that one would buttress one’s opinions with arguments
in talk with a stranger in order to avoid a negative attribution does not necessarily mean that
one would do the same in talk with someone one knows. However, van Dijk is committed

to this assumption.

215

A second problem with van Dijk’s assumption is that a one-to-one interview is a
very different kind of speech activity from a natural conversation. First, natural
conversations are not limited to two people. Indeed, there is no reason to expect that
everyday talk about minority groups should only involve two people. Second, interviews
are speech activities in which the interviewer asks questions and the interviewee answers
them. This is extremely different from natural conversation in which questions and
answers are normally more evenly distributed between participants.

In sum, the assumption that one-to-one interviews are similar enough to natural
conversations about minority groups to warrant conclusions about the reproduction of
racism in everyday talk is unjustiﬁed. Therefore, van Dijk’s conclusions are suspect on
methodological grounds. To put it another way, there is no way to be sure that the
discoursal processes involved in one-to-one interviews between strangers are even
remotely similar to those involved in everyday talk about ethnic out-groups between
friends, family members or acquaintances. Van Dijk’s conclusions could simply be an
artifact of his methodology. Clearly then, what is needed to conﬁrm van Dijk’s
conclusions and to rectify this methodological problem is to collect data which is not
subject to the same criticisms; that is, to collect data which more closely resemble ‘real’
conversations that dominant group members have about minority groups.

Verkuyten et a] (1994) were aware of this methodological problem:

There are relatively few studies that focus on people’s actual speech or
everyday discourse (but see Billig et a1, 1988). Studies that do analyze
people’s talk have been mainly concerned with material collected during
interviews whereby it is often claimed that these interviews were as informal

 

216

and spontaneous as possible (Essed, 1991; Van Dijk. 1987; Wetherell &

Potter, 1992). However, an interview context might have an effect on the
output, because part of what people say is probably linked to the fact that
they are communicating with social scientists.

Verkuyten et al therefore attempt to collect data which overcomes these problems by
creating a situation where more than one respondent discussed their views not with a social
scientist but with each other. Their data are based on eight taped 2 and one-half hour
sessions at a local community center in the Dutch city of Rotterdam. Although the
participants did not know each other beforehand, the two discussion groups that were
studied met for one evening each week for 4 weeks. During the discussions, two of the
researchers were present and one of the two initiated the discussion by raising certain
issues related to living in the neighborhood, education, crime, etc. By minimizing the
participation of the researchers and by allowing participants to meet beforehand and to
argue and discuss issues among themselves, Verkuyten et a1 hoped to create a more realistic
situation which would more closely resemble everyday talk; “In this way, we wanted to
create a situation that represented a “real-life” situation with everyday talk as much as
possible” (p.281).

While Verkuyten et al’s method of data collection is an improvement over van
Dijk’s, I would argue that it still has two signiﬁcant ﬂaws. First, the participants in these
discussions had only recently met and therefore had little opportunity to interact and learn
about each other. Their situation is not that much different than that of the strangers in van

Dijk’s interviews — the participants do not have an ongoing relationship with each other,

217

and for this reason, both situations are very much unlike “real-life” situations in which talk
about minorities occurs. Second, Verkuyten et al’s method still involved the presence of a
social scientist. Although the researchers were not actively involved in the discussion, it
cannot safely be assumed that their presence did not inﬂuence the talk. Again, it is
sociolinguistically naive to think that participants ignored the researchers — they knew that
they were assembled in order to be studied by the researchers there in their presence.

Thus, Verkuyten et al’s method was an improvement over van Dijk’s in that- it
attempted to gather talk between participants rather than between a single participant and a
researcher. Also, there was an effort made to allow participants to get acquainted with each
other. On the other hand, participants were not involved in an ongoing relationship (either
friend, family, neighbor, etc.) and researchers were still present during the discussion.
Therefore, it seems unjustiﬁed for Verkuyten et al to assume, as did van Dijk, that their
data represents “real life” situations in which talk about minorities occurs.

Of course, there is little chance of ethically capturing actual conversations about
minority groups beyond the eye of the researcher (and the ear of the tape recorder), and this
is why researchers must rely on more indirect evidence from situations which aim to at least
closely resemble such actual conversations. It has been claimed so far that previous
research has not adequately dealt with this methodological impediment in that the data
collected for these studies do not sufﬁciently resemble such actual conversations to warrant
conclusions about the nature of everyday talk about minority groups. I would like to argue

that the data used in the present work overcome at least two of the problems faced by the

218

earlier studies. First, the conversations which compose these data occur between friends,
and second, they occur outside the presence of a researcher. Based on these conditions, it
will be argued that the data for this work sufﬁciently resemble spontaneous, unobserved
conversations, and therefore do warrant the conclusions arrived at with regard to everyday
talk about minority groups.

The data upon which this work rests is made up of conversations between friends.
This means that the participants have interacted with each other before and probably know a
good deal about each other’s backgrounds and beliefs. In addition, given that they are
friends, it can probably be safely assumed that they have certain things in common with
each other. I would argue that actual, everyday talk about minority groups occurs between
people who have this sort of relationship. Certainly it is doubtful that such talk commonly
occurs between strangers. In American culture at least, it can be said that talking about
delicate issues such as religion, politics and race with a stranger or even an acquaintance is
somewhat taboo. Thus, such talk ordinarily is reserved for conversations between people
who know and are close to each other. Conversations between friends satisfy this
criterion.

Second, the conversations analyzed in this study did not occur in the presence of a
researcher. Obviously, actual talk about minority groups does not ordinarily occur in the
presence of a researcher, let alone any outsider. Thus, the powerful inﬂuence of (even a
passive) researcher was avoided in these data. One might object that the tape-recorder

indirectly represented the researcher for participants and therefore inﬂuenced their talk.

219

However, there is no indication in the data that the participants oriented themselves towards
either the recorder or the absent researcher. Whatever the effect of the recorder, it can
minimally be claimed that it’s inﬂuence was less than what would be the inﬂuence of an
actual researcher present in these conversations — so this represents a methodological
improvement over earlier studies.

Another difference between the data of the present work and the data of earlier
studies is that in these data the line between interview and conversation was blurred.
Although one of the participants in each conversation selected and introduced issues, that
same individual also fully participated in the discussions, so that once issues were raised,
the talk immediately became conversation rather than interview. Thus, by urging the data
collector to actively participate in the discussions, a more conversational kind of talk
emerged.

One might object that the requirement of “ﬁlling the tape” with 15 minutes of
conversation was a functional pressure which contributed to the shape of the discourse.
However, the data seen in chapters 2-6 make clear that participants in the data are more than
willing to discuss these issues raised by the interviewer/data gatherer. Participants in many
of the conversations in the corpus often went well beyond the 15 minute requirement and
appeared to be eager to discuss the topic of race on campus. Thus it is unlikely that the need
to ﬁll the tape was a signiﬁcant factor in the discourse.

In sum, although it is not claimed that the data collected for this work is authentic

everyday talk about minority groups (since it was elicited, recorded, etc.), it is claimed that

220

it closely resembles such talk and represents an improvement over the data of earlier
studies. This change (improvement) in data is signiﬁcant because it suggests the possibility
that the data used in earlier studies may not have been able to capture important
characteristics about discourse on race. Indeed, I would argue that the nature of the data of
earlier studies (e.g., van Dijk, 1987; Verkuyten, 1994) precluded any observation of
pseudo-argumentative discourse. This is mainly because one would not expect to ﬁnd PA
operating in interview situations between strangers (van Dijk, 1987), nor in situations
where talk occurs between near-strangers in the presence of researchers (Verkuyten, 1994).
PA ﬂourishes in discussions (not interviews) between people who maintain an ongoing
relationship and share certain beliefs, values, etc. (and not between strangers or near-
strangers).

The value of the method of data collection employed in the present work is that it
not only provides for a kind of talk which more closely resembles everyday talk about race
or ethnicity, but also that it provides for a setting in which PA may surface. Perhaps the
reason why PA has not been noted previously in the discourse literature is the difficulty of
providing the appropriate methods which allow it to be captured on tape. Also, it may have
escaped attention since it is a type of talk which presumably occurs only in private places
between people who are close to each other.

In any event, I believe that the study of PA could make a signiﬁcant contribution to
our understanding of discoursal processes having to do with the diffusion and reproduction

of attitudes and beliefs about minority groups. The data presented in this work involved

22]

content which embodies what has been called the modern racist ideology. Participants
return to the same themes again and again, including the beliefs that African-Americans are
unjustiﬁed in their demands and complaints, that racism and discrimination no longer exist,
and that advantages given to African-Americans are unfair (to Whites).

I contend that pseudo-argumentative discourse plays an important role in the
reproduction of modern racism in everyday talk, since it provides for its users a suitable
staging ground for the practice, diffusion, and reinforcement of beliefs and arguments
about minority groups, and allows for the balancing of both instrumental and identity
goals. That is, it is a type of discourse which allows participants to exercise nonnonnative
beliefs with relative impunity. Modern racist beliefs and arguments are rarely carried out in
real-life oppositional situations. Rather, more often they are rehearsed between people who
are close to and similar to each other in a nonhostile environment, where positions and

supports can be shared, honed, evaluated and reinforced.

FOOTNOTES

 

FOOTNOTES

Chapter 1

1

This is not to say that “old-fashioned” racism is no longer a force in the United States,
but rather it seems more accurate to claim that it is no longer the rule or norm in
mainstream society.

Chapter 2

l

The content of the data to be presented and discussed throughout this paper may be
offensive to readers. I would like to say on a personal note that I do not subscribe to
the views put forward in many of the excerpts which follow, and I ﬁnd the tone and
content of much of the talk abhorrent, disturbing and alarming. Nevertheless, with the
sincere aim of preserving objectivity in this work, I refrain from expressing my own
feelings towards the participants and the content of their talk. Instead, I prefer to allow
their words to speak for themselves. Unless otherwise noted, all of the excerpts in this
paper involve only European-American participants.

Narratives which function as arguments are very common in the data of this study, and
this is expected given the high degree of contact between Black and White students on
the campus. '

Although even narratives are not purely monologic (see Jefferson 1978, Schegloff,
1992)

The bid to make speech act theory the cornerstone of a theory of discourse is extremely
problematic (see the critiques of Levinson, (1983), and Geis, (1995)).

This review of CA will leave untouched other kinds of conversational structures such
as tum-taking, and other kinds of overall organization such as openings and closings
(see Ievinson 1983 for review)

An exception to this is disagreeing with someone’s self-deprecation, which seems to be
the preferred response

Chapter 3

l

Argumentation is a narrower term than persuasion since, the latter may not involve the
former — I can persuade you by literally twisting your arm.

222

223

O’Keefe acknowledges that the distinction between the two senses of argument is only
a starting point for analysis because it is not always clear how to delimit argumentsl

and arguments; For instance, does an argument] take place if there is no argument;?
Also, can arguments; occur in the absence of argumentsl?

Preston (1993, p.208) points out that disputes are not always explicitly stated and are
often implied from their supports.

It is claimed here that these particular (speech act and conversation analytic) models
may serve as useful heuristic tools for the investigation of the data of this work in order
to demonstrate simply that the data involve something other than what is considered by
some discourse and conversation analysts to be prototypical argumentation. There may
be other speech act or CA models which adequately account for the data, although I
have no knowledge of these.

 

Searle and Austin regarded argumentation as an illocutionary act, but without
recognizing the problems this presented to their theories.

Van Eemeren et al's (1993) conditions treat the standpoint as part of the argument,
whereas van Eemeren and Grootendorst (1984) treat only the constellation of assertives
as part of the argument. Van Eemeren et al's treatment is given here.

For van Eemeren and Grootendorst (1984), there is a distinction made between pro-
argumentation and contra-argumentation, each of which have a different set of
conditions (except for the propositional content condition, which is the same for both).
Pro-argumentation involves the attempt to justify an expressed opinion, whereas
contra-argumentation involves the attempt to refute an expressed opinion. Only the
felicity conditions for pro-argumentation are presented here.

Chapter 4

1

Greatbatch (1992) makes a similar observation with respect to the lack of such markers
in news interviews where disputes erupt between interviewees who are representing
different sides of an issue — since the disputes are mediated through a third party (the
interviewer), they are automatically mitigated and thus preference markers become
“redundant.” In addition, disputes commonly escalate in such contexts because
disputants are aware that the interviewer will eventually intervene to provide an exit
from the dispute.

This seems to be a common feature of written academic discourse where the arguments
of others must be presented and then disputed.

The media is an important player in discourse on race. Van Dijk (1987) notes that;
“Because the media provide the daily discourse and attitude input for most adult
citizens, their role as a prevailing discourse and attitude context for thought and talk
about ethnic groups is probably unsurpassed by any other institutional or public source
of communication” (p.41).

224

Chapter 5

1

By linguistic features, I am refering more generally to any linguistic phenomena,
including repetition, meta-talk, the sequential location of an utterance in relation to
others, etc.

This explains why denser discourse with less repetition can be frustratingly difﬁcult to
keep up with (e. g., at academic conferences).

In some cases, partial repetition may be used in conversation to signal doubt and to
allow the recipient to “think again” (Drew, 1992, p.476). However, repetition which
serves this purpose will be accompanied by linguistic features (mainly prosodic) which
make clear the intent of the speaker. That such a function of partial repetition is absent
in the data of this study is evident in the lack of any attempt to reforrnulate contributions
on the part of recipients.

Chapter 6

l

Pseudo-argument is not limited to discourses on race. But most likely it is limited to
discussions in which non-normative ideas are exchanged between people who are allied
in their ideologies and who wish to present themselves to each other in a positive light.

In line 2, “quote-unquote” is used as a device to call into question what C. apparently
takes to be the appropriateness of the label “Blacks.” This device may more generally
be used by speakers to signal a lack of committrnent to some term or idea. The use of
this device is probably related to the use of direct quotation to represent content which
one attributes to someone else (see chapter ﬁve, section 5.2.2.1.1.). Unlike direct
quotation, however, in which the speaker may either afﬁliate or disafﬁliate with the
quoted material, the prefatory use of “quote-unquote” appears to function only in a
disafﬁlative way. -

Certainly there are other low level verbal devices and semantic moves which have been
discussed (see van Dijk, 1992; Potter and Wetherell, 1987) which serve the strategy of
warding off potential negative inferences in response to some controversial assertion.

It should be noted, however, that the label “pseudo” is not intended to refer to the
quality, logicality, truth or falsity of the arguments put forward in this particular kind of
discourse. Rather, the term “pseudo” is meant to refer to the use of certain (argument)
structures which accomplish something other than what they purport to accomplish on
the surface.

There are other types of non-verbal simulation as well (dress, kinesis, reproduction of a
television news stage, etc.).

These include the exclusive right to ask questions, and topic control.

APPENDICES

APPENDIX A

APPENDIX A

Extra Credit for IAH231C

14.21;: For this project you will serve as an interviewer/participant for a study which will
examine students’ ideas, feelings and perceptions about race relations on campus (relations
between African-American and European-American students).

W To carry out this task, you will need to have a good tape-recorder, a
new standard sized tape, several hours to spare, and several friends of the same race as you
who are willing to participate in an audio recorded discussion of sensitive issues.

W You must ﬁnd two or three ﬁiends of the same race as you who you
know are motivated and enthusiastic. Reassure them that whatever they say will be strictly
conﬁdential, and so they should be as open and honest as possible. Your role will be to
guide the conversation so that topics stay within the general domain of 'race issues' on
campus. If participants begin to stray from these themes, you will gently bring them back.
In addition to initiating topics and guiding the conversation, you will also contribute your
own ideas as a participant. The conversation should ﬂow naturally and the presence of the
recorder should be played down.

You should aim at recording at least 15 minutes of conversation. 15 minutes is a
long time, and so you need to prepare in advance just in case the conversation begins to die
out. The list of topics on the next page may be referred to during the interview, but this
should be done discretely, if at all.

Before recording, be sure that all participants are relaxed and in a casual
environment (preferably in a dorm room or some other informal quiet setting). Everyone
should be sober and ready to tackle important issues. Have all of your participants,
including yourself, read and sign a consent form. Note that there is a speciﬁc consent form
for you as a ‘student collector’ and another form for everyone else (‘student volunteers’).

A good quality recording is essential. This means that noisy settings are to be
avoided (parties, bars, music in background, etc.). Please use a new, standard size tape,
and label the tape with your name and the date. Just before recording, do a test to be sure
the recorder is working. When you do this test, state your name and the date clearly. Then
go back and see if your voice was recorded. leave this information on the tape. Be sure
subjects are close enough to the microphone.

You might inform your participants the day before that the conversation will be
conducted at such and such a time, and that it will involve the discussion of race issues on

225

226

campus. This will give them the chance to think some beforehand.

levies;

Here are some topics / questions to choose from in your interview I conversation.
Don't feel limited to these issues. Any issue related to race relations on campus is
acceptable. Use your judgment to ﬁgure out which topics will result in the most talk (what
will get people ﬁred up?). let other participants initiate new topics too. Let topics emerge
naturally. Initiate a new topic only when no one has anything more to say about an old
topic.

1) Do you believe that prejudice / racism exists on campus?

2) Does reverse-discrimination exist on campus?

3) How do you feel about interracial dating on campus?

4) Is there too little or too much interaction between black and white students?

5) How do you feel issues of race are handled in the media ( in the State News)?

6) Do you think it‘s important to make friends outside of your own race at
school?

7) Do African-American students receive differential treatment from the
university?

8) Does the university do enough to ensure racial harmony?

9) Do you support 'minority only’ scholarships?

10) Are courses exclusively devoted to African-American studies appropriate?

11) Are white students insensitive to racial prejudice?

12) Are black students overly sensitive to racial prejudice?

13) Do you think either African-American or European-American students feel
physically threatened by the other group on campus?

You might begin your conversation by saying to your friends that you want to get
their ideas about race issues and problems on campus, and that the ideas collected will be
studied by a professor at the university with the aim of understanding how black and white
students feel about such issues. Encourage them to say what they really feel and think.
Use your intuition in getting things started, and keeping things going. Remember that we
want your ideas too -- you will be both moderator and participant.

APPENDIX B

APPENDIX B

Key to transcription conventions

“ . ” Falling intonation followed by pause

“ , ” Continuing intonation

“ ? ” Rising intonation followed by pause

“ — ” Pause

“ ! ” Exclamation

“ CAPS ” Emphatic stress

“ : ” Lengthening

“- ” A hyphen marks an abrupt cut-off or self-interruption

“ = ” marks unbroken continuity between a speaker’s talk which has been
separated due to limits on space.

“. [ ” A lefthand bracket marks the starting point of some overlap

“ ] ” A righthand bracket marks the ending point of some overlap

“ ( ) ” Unintelligible material

227

LIST OF REFERENCES

LIST OF REFERENCES

Alexandrova, D. (1987). Rhetoric and the theory of argumentation. In F. H. van Eemeren,
R. Grootendorst, J. A. Blair, & C. A. Willard (Eds)
Wes (pp. 266-273). Dordrecht-Holland: Foris Publications.

Atkinson, J .M. (1992). Displaying neutrality: formal aspects of informal court
proceedings. In-P. Drew & J. Heritage (Eds). mm (pp. 199-21 1). Cambridge:
Cambridge University Press.

Austin, J. (1976). WWW. Edited by J .O. Urmson and M. Sbisa.
Oxford University Press (2nd rev. ed., 1st ed., 1962).

Baker. GP. (1895). MW Ginn

Benoit, P. J. (1987). Orientation to face 1n everday argument. In F. H. van Eemeren, R.
Grootendorst, J. A. Blair, & C. A. Willard (Eds. ), ° v
W. Dordrecht-Holland: Foris Publications. 144-154.

Billig, M. (1988). The notion of ‘prejudice’: some rhetorical and ideological
aspects. Int, 8, 91-110.

Blair, J. A. (1987). Argumentation, inquiry and speech act theory. In F. van Eemeren, R.
Grootendorst, J. A. Blair & C. A. Willard (Eds. )
discipline (pp. 189-200). Dordrecht—Holland: Foris Publications.

Brown, R, & Levinson, S. (1987). Politeness. Cambridge: Cambridge University Press.

Clayman, SE. (1992). Footing in the achievement of neutrality: the case of news-interview
discourse. In P. Drew & J. Heritage (Eds). mm (pp.163-198). Cambridge:
Cambridge University Press.

Cox, J. R., & Willard, C. A. (1982). Introduction: The ﬁeld of argumentation. In R. J. Cox
&C.A.Willard(Eds.:c) .. - -
xlvii). Carbondale: Southern Illinois University Press.

 

Crosby, F., Bromley, S, & Saxe, L. (1980). Recent unobtrusive studies of black and
white discrimination and prejudice: a literature review. W. 87, 546-
563.

Dovidio, J. F. & Gaertner, S. L. (1986). Prejudice, discrimination, and racism: historical
trends and contemporary approaches. In S ..L Gaertner & J. F. Dovidio (Eds)
(pp. 1-33). Academic Press.

Drew, P. (1992). Contested evidence in courtroom examination: the case of a trial for rape.
In P. Drew & J. Heritage (Eds). truism (P11470620). Cambridge: Cambridge
University Press.

228

Essed, P. (1991).
Sage.

 

Gaertner, S .L, & Dovidio, J. F. (1986). The aversive form of racism. In S ..L Gaertner &
J F Dovidio (Eds ) Prejudicediscrimiuaticnandracism (pp. 61 -89) Academic Press

Garﬁnkel, H. (1967). W. Englewood Cliff, NJ: Prentice Hall.

Geis, M. L. (1995) - -- -.
W. Cambridge: Cambridge rUniversityW Press.

Goffman, E. (1967). The nature of deference and demeanor. In Intensetigmnljinia] (pp.49-
95). New York: Anchor Books.

 

Goffman, E. (1981). Footing. In Equnspﬂalk (pp.124—159.) Philadelphia: University of
Pennsylvania Press.

Greatbatch, D. (1992). On the management of disagreement between news interviewees. In
P. Drew & J. Heritage (Eds). mm (Pp.268-301). Cambridge: Cambridge
University Press.

Grice, H. P. (1975). Logic and conversation. In P. Cole & J. L. Morgan (Eds. ) mm
semantics Warts (pp 41-58) New York: Academic Press.

Gumperz, J. (1982). W. Cambridge: Cambridge University Press.
Hanks, W.F. (1990). W. Chicago: University of Chicago Press.

Heritage, J. (1988). Explanations as accounts. a conversation analytic perspective. In C.
._ .7 __ -_ .. . 210...! (pp. 127-144).

 

London: Sage . I

Hewitt, J ., & Stokes, R. (1975). Disclaimers. W. 40, 1-11.

Hymes, D. (1972). Models of the interaction of language and social life In J. Gumperz
and D. Hymes (Eds) , _ - , , ,
(pp. 35-71). New York: Holt, Rinehart and Winston.

 

Hymes. D. (1974). Towards ethnographies of communication. In W
(pp. 3-28). Philadelphia: University of
Pennsylvania Press.

Jackson, S. (1986). Rational and pragmatic aspects of argument. In F. van Eemeren, R.
Grootendorst, J. A. Blair & C. A. Willard (Eds. )
disginline (pp. 217-228). Dordrecht-Holland: Foris Publications.

Jackson, S. & Jacobs, S. (1982). Conversational argument: a discourse analytic approach.
lnR.J.:cCox&C..AWillard(Eds.) -
(pp. 205-237). Carbondale: Southern Illinois University Press.

 

Jacobs, S. (1986). The management of disagreement 1n conversation. In F. van Eemeren,
R. Grootendorst, J. A. Blair & C. A. Willard (Eds)
Qﬁdisginline (pp. 229- -.240) Dordrecht—Holland: Foris Publications.

230

Jefferson, G. (1973). A case of precision timing in ordinary conversation. Semiotics. 9,
47- 96.
Jefferson, G. (1978). Sequential aspects of storytelling 1n conversation. In J. Schenkein

(Ed) Smdimuﬂmanizaticncfscnversaticnalinteraction (pp 219-248) New
York: Academic Press.

Katz, 1., Wachenhut, J. and Hass, R. G. (1986). Racial ambivalence, value duality, and
behavior. In S. L. Gaertner and J. F. Dovidio (Eds. )
racism (pp. 35-59). Academic Press.

Kinder, D. R., and Sears, D. O. (1981). Symbolic racism versus racial threats to ‘the good
life’19:11:11alcf.Personalit1Lan51.8.09idllsirchszlrrsirr40 414-431

Kline, S. (1987). Identity management in argumentative discourse. In F. van Eemeren, R.
Grootendorst, J. A. Blair, & C. A. Willard (Eds)
discipline (pp. 241-252). Dordrecht-Holland: Foris Publications.

Labov, W. (1972). The transformation of experience in narrative syntax. In Languagoin
theinncuiﬂ (pp. 70-109). Philadelphia: University of Pennsylvania Press.

Labov, W., & Fanshel, D. (1977). Wise. New York: Academic Press.
Ievinson, S. (1983). Pragmatics. Cambridge: Cambridge University Press.

Levinson, S. (1992). Activity types and language. In P. Drew and J. Heritage (Eds). Talk
mils (pp.470-520). Cambridge: Cambridge University Press.

Linde, C. (1981). The organization of discourse. In T. Shopen and J. M. Williams (Eds)
WW (pp. 84-] 14). Cambridge, MA: Winthrop Publishers.

Linde, C., & Labov, W. (1975). Spacial networks as a site for the study of language and
thought. Language 51, 924-939.

McConahay, J. B. (1986). Modern racism, ambivalence, and the modern racism scale. In

S L Gaertner and J F Dovidio (Eds )Errrudrcetdrscnmmamnandmsm (pp 91-
125). Academic Press.

O'Keefe, D. J. (1982). The concepts of argument and arguing. In R. J. Cox & C. A.
Willard (Eds ) WWW (pp 343)
Carbondale: Southern Illinois University Press.

O'Keefe, D. J. (1977). Two concepts of argument. Wm
Association. 13,121-128.

Perelman C & OlbrechtS-Tyteca. L (1969) WI].
argumentation. Notre Dame, IN: University of Notre Dame Press. (paperback edition,
197 1).

Pomerantz, A. (1984). Agreeing and disagreeing with assessments: some features of
preferred/dispreferred. turn shapes. In J. M. Atkinson & J. Heritage (Eds. ) Inc;
(pp. 57- 101). Cambridge:
Cambridge University Press.

Potter, J., & Wetherell, M. (1987). WWW

231

Maximo. London: Sage Publications.

Preston, D. (1993). The uses of folk linguistics. WWW
Linguistics, 3, 181-.259

Schegloff, E. (1972). Sequencing in conversational openings. In J. Gumperz and D.
Hymes (eds), Directions in sociolinguistics. New York: Holt, Rinehart and Winston,
346—80.

Schegloff, E. (1982). Discourse as an interactional achievement. Some uses of 'uh huh'
and other things that come between sentences. In D. Tannen (Ed. ),
(pp. 7 1-93). (Georgetown University Round Table).
Washington, DC: Georgetown University Press.

Schegloff, E. (1992). In another context. In A. Duranti & C. Goodwin (Eds), Rcmnndng
WWW (PP 191-228) Cambridge: Cambridge
University Press.

Schegloff, E. & Sacks, H. (1973). Opening up closings. Semiotics. 7, 289-327.

Schiffrin, D. (1980). Meta-talk: Organizational and evaluative brackets 1n discourse.
Smiclcaicallnquirx 50 199-236

Schiffrin, D. (1985). Everyday argument the organization of diversity 1n talk. In T. van
j - .. .
46). London: Academic Press.
Schiffrin, D. (1987). Disooursoinaikcijs. Cambridge: Cambridge University Press.

 

Schiffrin, D. (1990). The management of a co-operative self during argument. the role of
opinions and stories. In A. D. Grimshaw,

W
Wm (pp. 241-259). Cambridge: Cambridge University Press.
Schiffrin, D. (1994). Appioochcsiodiscoum. Cambridge, Mass: Blackwell.

Searle J R (1970) Wﬂmmmmum Cambridge:
Cambridge University Press (2nd ed.;1sted.,1969).

Searle, J .R. (1976). The classiﬁcation of illocutionary acts. Languagcinjocicu. 5, 1-24.

Stubbs, M. (1983). .J ' -
Chicago. University of Chicago Press.

Tannen, D. (1989). - ' ~ ' " ..
discourse Cambridge: Cambridge University Press.

Taylor, T., & Cameron, D. (1987). Analyzingconyczsation. New York: Pergarnon Press.

Toulmin, S. E. (1958). W. Cambridge:

van Dijk, T. A. (1987).
London: Sage.

 

 

 

232

van Dijk, T. A. (1993). Elitedisccurseandracism. London: Sage Publications.

van Eemeren, F.& Grootendorst, R. (1984). W.
Dordrecht-Holland: Foris Publications.

van Eemeren, F. Grootendorst, R. Jackson, S, & Jacobs, S. (1993). Esconstiuciing
W. Tuscaloosa: University of Alabama Press.

Verkuyten, M, de Jong, W., & Masson, K. (1994). Racial discourse, attitude, and

rhetorical manoeuvres: race talk 1n the Netherlands. WNW
Psychology. 13, 278- 298.

Wetherell, M. & Potter, J. (1992). ' '
W- New York: Columbia University Press.

Willard, C. A. (1989). WM- Tuscaloosa: The University of Alabama
Press.

Wittgenstein, L. (1958). Philosophiciiinuostiganons. Oxford: Blackwell.

.rti; I‘-

"11111111111111.1111"!IIIII