INTERACTION MANAGEMENT IN DISPUTATIOUS SOCIAL SITUATIONS

BY

Renee Beth Stahle

A DISSERTATION

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

DOCTOR OF PHILOSOPHY

Department of Communication
Department of Sociology

1988

__ . .
u” r--*v --:=~-='""”o"‘ - -~ - -

ii
(a
O“
‘\
W

 

ABSTRACT
INTERACTION MANAGEMENT IN DISPUTATIOUS SOCIAL SITUATIONS
BY
Renee Beth Stahle

This research had three broad aims; (a) to test a coding
scheme for measuring interaction management, (b) to test
hypotheses derived from Weinstein's Theory of Interpersonal
Tactics, and (c) to investigate the use of interaction
management as a social influence attempt in divorce
mediation sessions. The coding scheme was developed to
study the linguistic and conversational features of
interaction management as the complex, simultaneous, and
sequentially organized phenomena they are. The results
showed that the features could be reliably coded by trained
coders.

No significant relationship was found for the overall
use of interaction management and mediation outcomes. A
significant relationship was found for the use of
interaction management between the participants and mediation
outcomes. The results of this research indicate that when
the participants pursue their goals in disputatious
situations, the social influence strategy of interaction

management plays a primary but not unequivocal role. In

 

Renee Beth Stahle

support of Weinstein's theory, the overall use of
interaction management is not as important to reaching a

mutual agreement as the use of interaction management between

the participants. More specifically, agreement was more

likely when the mediators, as opposed to the parents,

managed the interaction.

 

It is not
what you are
that holds you back...

but what
you think
you are not.

This dissertation
is lovingly shared
with fellow graduate students,
who let 'what they thought they were not'
stop them from completing one of their own.

iv

 

ACKNOWLEDGEMENTS

Many people have contributed in important ways to this
research; a few must be acknowledged here for the specific
marks they left on this enterprise. All of my committee
members - Drs. Frank Camilleri, Thomas Conner, William
Donohue, William Faunce, Kathy Kellerman, Gerald Miller -
gave careful and critical readings to various versions of
this manuscript. The questions and issues they raised
caused me to clarify my thinking and writing on several
points. More importantly, their comments have given me a
good sense of both my strengths and weaknesses as a scholar
and the means to improve the latter.

My dissertation advisors, Dr. Thomas Conner and Dr.
William Donohue deserve additional mention. For several
years they have worked together and taken an active role in
my professional development; heightening both my
intellectual growth and productivity. As my research team
advisor Dr. Donohue willingly shared his data base,
undergraduate coders, research time, computer time, computer
and office. Their tutelage and friendship have been, and
will continue to be, valued by me.

Finally, my love and gratitude to family - Dad, Mom,

Phillip, Tamara, Lorena, Chip and Robert - and friends -

 

Dr. Mike Allen, Dr. Nancy Burrell and Dr. Drew McGukin -

who comforted, listened, suggested, supported, prodded and

then listened some more.

vi

 

TABLE OF CONTENTS

Page
LIST OF TABLES .. ........... ............ ....... ..... viii
CHAPTER

I. REVIEW OF LITERATURE ............... ..... ..... 1
Introduction ..... . ...... ........... ..... .. 1
Significance of the Study . ................ 3
Interaction Management .................... 4
Interpersonal Tactics Theory (ITT) ........ 8

Conceptual Integration of Interpersonal
Tactics Theory and Interaction Management . 17
Hypotheses .............. ...... . ........... 38
Additional Questions ...................... 43
II. METHODS ........ ....... ...... .................. 44
Sample of Interactions ....... ....... ...... 44
Measures ........... ......... .............. 47
III. RESULTS .......... ...... ... .......... .... ....... 53
Tests of Hypotheses ............... . ....... 53
Exploratory Hypothesis .................... 58
IV. DISCUSSION ......... ............. ....... ....... 61
Interaction Management Coding Scheme ..... . 61
V. SUMMARY .............. ...... ........... ..... ... 79
FOOTNOTES ..... .............. ......... ...... .......... 82
APPENDIX A: SAMPLE MEDIATION AGREEMENT ..... ......... 83
. APPENDIX B: CODING INSTRUCTIONS ..................... 85
BIBLIOGRAPHY .......... .............. ................. 87

vii

TABLE

 

LIST OF TABLES

Page

Distribution of Linguistic and Conversational
Feature Associated with Interaction Management. 34

Inter-rater Reliabilities for the Coding of
Interaction Management ........................ 51

Mean Interaction Management for Speaker to
Addressee in Successful and Unsuccessful
Mediation Sessions 00'.........IOOIOOOOOIUOIIOO 56

Mean Interaction Management by Time for
Successful and Unsuccessful Mediation Session . 57

Mean Interaction Management by Speaker to
Addressee and Time ........... ................. 59

viii

 

CHAPTER I

REVIEW OF LITERATURE

Introduction

- Customers making deals with used car salespeople

- Politicians discussing new bills

- Lawyers appealing cases with judges

- Children fighting over toys

- Friends choosing movies to view

- Parents deciding the custody of their children

Characterizing all of these encounters is a social
influence process whereby two or more parties with different
preferences concerning an agreement to be reached attempt
to move each other toward their preferred outcome. Because
of the pervasive presence of such disputatious social
interactions where parties have opposing goals, it is
important to look at the means parties can use to manage
their disputes (Hocker 8 Wilmot, 1985). Traditionally,
researchers have focused on cooperation and competition as
more global strategies for social influence in interaction
(Putnam, 1985). More recently, both communication scholars
(e.g., Applegate G Leichty, 1985; Putnam, 1985) and social
psychologists (e.g., Grimshaw, 1980) have called for work to
determine how communication can be used to exert influence

in such social encounters.

Two different approaches to the study of communication

 

2

focus on the use of talk for social influence attempts. In
the first approach Brown and Gilman (1960), Brown and
Levinson (1978), and Weinstein (1966) proposed that
participants use talk to communicate relational identities
that promote and maintain their social influence in an
encounter. The most attention has been paid to forms of
address and reference to persons (Brown 5 Fraser, 1979).
Stahle (1985) broadened the scope of this research to
include other forms of talk - hedges, positive tag
questions, interruptions - used in the negotiation of the
relational identities holding between the participants in
disputatious social situations like divorce mediation. The
second approach has focused on the management of the
interaction itself as a social influence attempt. Grimshaw,
(1980) and Street & Cappella, (1985) proposed that
participants can exert influence through interaction
management, or the use communication tactics to determine
who talks about what and how in the encounter. There has
been, however, little research on interaction management as
a social influence attempt.

This dissertation extends Stahle's (1985) previous work
by exploring the manner in which disputing couples use
interaction management in an attempt to structure the
interaction in ways that are congruent with their goals in
divorce mediation. The goal of this dissertation is to
advance the theoretical explanation and empirical

development of interaction management as a strategy for

 

social influence attempts in disputatious social situations

like divorce mediation. To accomplish this goal this
dissertation will proceed in four sections. The first
section discusses the significance of the study. The second
section summarizes the work on interaction management. The
third section presents Weinstein's Theory of Interpersonal
Tactics as a theoretical framework to account for how social
actors go about influencing others in accordance with
actors' goals in social encounters. The last section
presents a theoretical and empirical integration of
Weinstein's theory and the work on interaction management
from which several propositions are derived. These
propositions, in turn, are applied to divorce mediation as a
disputatious interaction and the hypotheses to be tested in

this study are presented.

Significance of the Study

Understanding how participants use talk to manage
interactions in pursuit of their goals has both theoretical
and practical implications. Theoretically, little
integration or empirical deve10pment exists of the related
aspects of interaction management; the use of talk as a
social influence attempt, identification of the features of
talk that are available to structure interactions, and how
interaction management is conducted in social encounters.
This study is one attempt to provide such a theoretical
integration and empirical development.

Identifying the ways interaction management can be used

 

4

t0 Structure the interaction also has practical implications
for those involved in disputatious interactions. In the
mediation context, Donohue and his colleagues (Donohue G
Weider-Hatfield, 1986; Donohue, Allen, 5 Burrell, 1985) have
proposed that one component of mediator competence is the
ability to structure the interaction so that mediators can
direct the course of the mediation toward agreement. The
practitioners, then, need a working knowledge of and ready
access to a variety of communication tactics that can be
used to control the course and structure of disputatious
interactions (Pruitt, 1981; Rubin, 1980; Saposnek, 1985).
Once research identifies the features of talk available for
interaction management, practitioners could be trained to
move interaction in productive directions. A second aim of
this study, then, is to add to the practitioners‘
repertoires of scripted interaction management tactics, so

as to facilitate intervention speed and effectiveness.

Interaction Management

This dissertation will begin by focusing on how talk is
used to manage the structure and content of disputatious
social interactions. The idea that participants use talk to
manage the structure and content of disputatious
interactions in attempting to influence one another in
pursuit of their social goals has been discussed by a number
of different researchers (e.g., Argyle, 1969; Street &
Cappella, 1985; Wilmot, 1986). Together, these researchers'

definitions and discussions suggest several common themes

 

 

 

 

S

‘that can.be used to further develop the strategy of

interaction management.

These themes appear among the various definitions of

interaction management:

Conversational control...the ability to determine who
talks about what does constitute one type of influence.
Some common indicators of conversational control include
who talks most, who interrupts whom and who changes the
tOpic most often (Adler 8 Towne, 1987, p. 273).

Human social actors spend much of their time with others
engaged in talk... [M] uch of this talk is in some sense
manipulative with speakers attempting to cause
(influence) their hearers to alter their behaviors
(including 5 eech) or beliefs...in some way (Grimshaw,
1981, p. 205 .

The more typical connotations of the word manage implies
that a person intentionally seeks to alter the content,
tenor or events of a conversation toward some
preordained end or purpose. A less typical connotation
of the word management will also be discussed here...the
ability to control interaction depends upon the
existence of certain regularities that can be exploited
by one or the other conversational partners (Street 8
Cappella, 1985, p. 2).

Conversational control or dominance is any communicative
device (e.g., speaking len th, speaking frequency, or
interrupting one's partner which lessens the
communicative role of another (Wilmot, 1986, p. 105).

...Control is the constellation of constraints people
place on one another by the manipulation of both
interactional structure and content, which limit the
Options appropriately available subsequently to each
relational partner and the relational system as a whole
(Weimann & Kelly, 1981).

As these definitions illustrate, the first major theme

is many scholars (e.g., Grimshaw, 1980; Street 8 Cappella,

1985) conceptualize social influence to Operate during most

 

" ...‘u

n

7" .

 

 

 

 

6

Social encounters. For example, in the Grimshaw (1980)
definition presented earlier, he states that actors spend
much of their time with others attempting to influence them
into performing certain actions or in themselves being the
object of such influence attempts. This position suggests a
complementary character to social life and the goals that
social actors pursue in interaction. Social actors have
goals they want to accomplish; making a decision, escalating
relational intimacy or making a good impression (Street &
Cappella, 1985). Accomplishing many of these goals,
however, is dependent upon receiving the desired response
from the other actor in the interaction. When actor and
other interact, actor tries to elicit the desired response
from other as other becomes the object of a social influence
attempt. This reciprocal activity characterizes the
mechanism for social influence attempts.

There has also been increasing recognition of a second
theme present in the work on interaction management; actors
in social interaction spend the greater proportion of their
time with others engaged in talk. For example, as early as
1949, Morris contended that "sharing a language with other
persons provides the subtlest and most powerful of all tools
for controlling the behavior of these other persons to one's
advantage" (p. 41). This theme can also be clearly seen in
the Adler and Towne (1987) definition presented earlier when
they equate conversational control with talk. These

observations led Grimshaw (1981) and Putnam 8 Geist (1985),

 

 

 

7

among others, to conclude that communication is the essence
of social influence attempts. Unfortunately the attention
paid to the linguistic and conversational features employed
in social influence attempts has been, until quite recently,
so modest as to be practically nonexistent (Giles & St.
Clair, 1986; Grimshaw, 1980). Research across several
disciplines has identified common groups and patterns of
linguistic and conversational features that have been
associated with social influence attempts; e.g., terms of
address and reference (c.f., Brown 8 Gilman, 1960; Erwin-
Tripp, 1964; 1972), interruptions and duration of speaking
time (c.f., Donohue G Weider-Hatfield, 1986), and use of
declaratives and questions (c.f., DeVito, 1987; Erwin-Tripp,
1968). This association is useful for identifying some
similar features of talk that are available for social
influence attempts. Without clear criteria or a conceptual
rationale, however, it is difficult to identify, classify or
evaluate other features of talk as part of an interaction
management strategy for social influence attempts.

In summary, the similar themes that appear in the
definitions of interaction management are: (a) social actors
attempt to influence each other in pursuit of their goals in
social interaction, (b) management of the interaction is a
strategy that is available for these social influence
attempts, and (c) various features of talk and conversation
are the resources available to implement an interaction

management strategy. In this body of work, however, little

 

8

theoretical explanation/integration or empirical development
of these related aspects of interaction management exists.
One promising approach to this theoretical integration is
the work done by Weinstein (1966) in his Theory of

Interpersonal Tactics (ITT)}

Interpersonal Tactics Theory (ITT)

This section will begin with an explication of
Weinstein's theory to be followed by a review of some of its
limitations and suggested directions for future work.
Weinstein's ITT provides an explanation of the means by
which a social actor attempts to influence other
interactants into doing things that aid the pursuit of the
actor's goals in a social encounter.

The first major premise of ITT is the postulate from the
work of Goode (1960a,b) and Homans (1961) that human
behavior is goal directed. This premise is similar to the
first theme from the definitions of interaction management
presented above. The implication Weinstein draws from this
postulate for social interaction is that social actors bring
personal purposes into interaction. When a social actor and
another interact, these purposes can be defined in terms of
the interpersonal task or the desired response from the
other that the actor tries to elicit in the encounter. This
response may be an end in and of itself or a means to evoke
other responses more crucial to the actor's goals in the
interaction. To accomplish this interpersonal task, the

actor selects lines of action (strategies) and implements

 

9

them in some behavior (tactic) directed toward the task.
This social influence process, as presented by Weinstein,
contains the two critical elements of lines of action and
behavioral displays. For Weinstein, then, EEE central
ELE selected ehe elaborated lfl hhe actor's expressive
hehavior 13 order he elicit ehe desired response fgeh ehe
other.

Weinstein proposed three considerations for the
selection and elaboration of a line of action by the actor:
(a) maximization of the likelihood of obtaining a desired
response from the other (b) the degree of conscious
awareness in the selection of a line of action, and (c) the
extent to which actor selects strategies based on the
meaning they have for the other. In regards to the first
consideration of maximizing goal achievement, Weinstein
(1966) proposed that the actor

has his interpersonal tasks...but he is also involved

with other actors each with their own purposes and

associated preference orderings which, in all
likelihood, are somewhat different from his...[A]ctor,
in seeking to achieve his own goals in the encounter,
must also keep the others bound in the relationship

(p. 395).

The actor is expected, then, to select a line of action
that will facilitate the pursuit of his/her goals and
promote and maintain relationships congruent with those
goals. For example, in order to have control over a

classroom interaction, a teacher would want to establish an

authority relationship and manage the structure and content

 

 

 

10

0f the interaction. According to Weinstein, any line of
action selected to maximize goal achievement must have goal
related strategic consequences and define an appropriate
relationship for the actor and the other.

Weinstein (1966) next addresses the consideration of
whether or not maximization suggests conscious rationality
in the actor's selection of lines of action:

Who cannot recall thinking ahead to an upcoming

encounter, focusing on just what might be the best

impression to convey, the best tone to strike, the best
tack to take to achieve our purposes? But a good deal
of [strategies] are not consciously selected. Many
lines of action, well designed to elicit task responses
from others are used, not because we are aware of their

[strategic] advantages but because we have learned they

are situationally appropriate. (p. 397)

Weinstein maintains that the selection of a line of action
by the actor constitutes a planned purpose which may or may
not be consciously conceived. Specifically, some strategies
may be unconsciously applied as part of a routine while
others are chosen with conscious attention to their ability
to maximize goal achievement (Brown 8 Levinson, 1978).

For any line of action multiple behavioral displays
(tactics) exist. Weinstein argues that the problem lies in
selecting behavioral displays that are most suited to
obtaining a desired response from the other. The optimum
choice is to select behavioral displays that have the
desired strategic consequence and that clearly communicate
that strategic consequence for the other. Weinstein's

solution for the implementation of lines of action was the

application of the fundamental theorem from the symbolic

 

 

 

 

11

interactionist tradition of Mead (1934) and Goffman (1959)
that, communication with significant symbols gives rise to
consensus which allows for coordination to occur. According
to Mead (1934), "significant symbols are behavioral displays
indicating the same future phases of action to actor and
other" (p. 39). According to Goffman and Mead, the actor
attempts to influence the other's responsive behaviors in
line with the actor's goal(s) in the interaction (social
influence attempt). When the actor makes a claim to pursue
certain types of interpersonal tasks (e.g., choosing the
topic of conversation) and employs particular lines of action
in pursuit of that task (e.g., interrupting or changing the
topic), the actor chooses his/her behavioral display
primarily on the basis of the shared meaning s/he believes
the other will attach to it.

The other uses a similar process of communication with
significant symbols to assign meanings to the actor's
behavioral displays. When the actor and the other
interact, then, the other first "takes the role of the
actor". Meaning is assigned by assuming that if I (myself)
as the other had said something like "Let's pursue this
other aspect.", the utterance would be the significant
symbol of a permission directive and the other would be
attempting to manage the topic of the conversation and would
expect an on-topic response from the actor. The other
further assumes that as members of the same speech

community, the actor would assign that same meaning, he

 

 

 

12

lEElEiLgm. In the accompanying example presented above, it
is necessary that the other recognizes at some level that
the actor is using a permission directive, but it is not
enough. The other, at some level, must also recognize the
strategic consequence of the action. Through this
communication with significant symbols the actor and the
other have a shared meaning as to the 1ine(s) of action
being employed in the interaction.

In summary, Weinstein's ITT provides the following
explanation of social influence attempts in interaction.
The first major premise is that actors have goals they want
to accomplish in social encounters. Given the fundamentally
interdependent character of social life, however, the
accomplishment of many of these goals is dependent upon
receiving the desired response from others in the social
encounter. Actors select a line of action that will
maximize the likelihood of obtaining that desired response
from others. Actors select a behavioral display to
communicate this chosen line of action primarily on the
basis of the shared strategic meaning they believe others
will attach to it.

Empirical work by Weinstein and his colleagues
(Weinstein 8 Deutschberger, 1963; Weiler G Weinstein, 1972)
has focused on the means of promoting and maintaining
relationships as appropriate lines of action. For example,
Weinstein and Deutschberger (1963) proposed altercasting,

which is defined as creating an identity for the other

 

 

 

13

congruent with one's goals, as a basic line of action. In
this study, students worked with a confederate to construct
a questionnaire that accounted for ten percent of their
grade in a course. Based on videotapes of the interactions,
raters coded: (a) the actions of the students relevant to
casting the confederates into goal maximizing relationships
of authority, status, social-emotional distance, support,
autonomy, and freedom; i.e., the significant dimensions of
altercasting; and (b) the frequency of pronoun usage. The
results supported the use of the altercasting dimensions
listed above, the use of the pronoun "we" with projecting
an interdependence relationship, and the use of the pronoun
"you" with attempts to impose restrictions. The hypothesis
that the high frequency of the pronoun "I" would be related
to claiming superordinate status was not supported.
Weinstein and Deutschberger concluded that altercasting is a
line of action for promoting and maintaining relationships
through selected behavioral displays.

Although the Weinstein and Deutschberger study does
provide support for altercasting as a basic line of action
for social influence attempts in social interaction, the
concepts involved in the theoretical relationships are not
clearly operationalized. First, it is unclear what the
students' goals are in the experimental situation. For
example, to get a good or passing grade, to get the
confederate to do all the work, or to impress the teacher

are all possible goals. Without clearly articulating the

 

 

 

 

 

14

goals, it is also difficult to assess if or why altercasting
is the most appropriate line of action for the students to
use to accomplish their goal(s). Finally, no clear criteria
or rationale is provided for the use of the pronouns "I",
"we," and "you" as the behavioral displays of altercasting.
These shortcomings are also shared by the Weiler and
Weinstein (1972) study. Together the work that has been
done by Weinstein and his colleagues indicates that there
has not been a rigorous test utilizing Weinstein's theory.
Theoretically, ITT provides a clear formulation of how
social actors chose lines of action and behavioral displays
in pursuit of their goals in social interaction.
Empirically, work on ITT has been hampered in two ways.
First, research utilizing the theory has focused on
relational expression as a line of action to the virtual
exclusion of other viable lines of action such as impression
management, reciprocity, or interaction management. Second,
rigorous tests of the theory have not been made. For
instance, let's consider an example similar to the
experimental situation used in the Weinstein and
Deutschberger study. Suppose two students have to conduct
and analyze an experimental study for a research methods
course. 'Student A has a very poor grasp of statistics and
would like to have Student B do all of the analytical work.
Unlike the subjects in the Weinstein and Deutschberger
study, Student A has a clear and recognizable goal. Student

A now has to select a line of action for accomplishing the

 

 

 

 

 

15

goal and behavioral displays to operationalize the line of
action. Possible lines and displays could include (a)
creating apprOpriate relational identities - convincing the
partner that the partner is more analytically competent than
the student, (b) reciprocity - offering to do all the
library research, keypunching, and typing in exchange; or
(c) interaction management - structuring the interaction to
encourage the analytical contributions of the partner.
Weinstein would expect Student A to attempt to create
appropriate relational identities but would provide no
conceptual rationale for why this would be the best line of
action for Student A to use to accomplish interactional
goals in this particular situation. Weinstein's work has
suffered from this lack of a clear conceptual rationale for
deciding which goal an actor is pursuing in a particular
situation, which line of action would best accomplish that
goal, and which behavioral displays would operationalize
that strategy. Without this conceptual rationale, it is
difficult to then develop hypotheses that would provide a
rigorous and critical test of Weinstein's ITT. Equally as
important, the generalizability of the theory is inhibited
by the adherence to one line of action. Although
Weinstein's work has come a long way in predicting how
social actors attempt to influence others into doing their
bidding in social encounters, it suffers from the lack of
clear conceptual rationales for and operationalizations of

the goals, lines of action, and behavioral displays for the

 

 

 

 

 

 

 

16

partixnalar situations in which actors find themselves.

In conclusion, both Weinstein's ITT and the work on
interaction management could be advanced by studying them in
tandem. The ITT could profit from a more rigorous test of
the theory that clearly satisfies four requirements. First,
the goals of the actor must be clearly Specified for a
particular social interaction. Second, lines of action must
be identified that will maximize the likelihood of the actor
accomplishing those goals in the specified interaction.
Next, clear criteria must be deve10ped for identifying the
behavioral displays that communicate the chosen line of
action. Last, a line of action, other than those previously
studied, must be employed to increase the generalizability
of the theory. As previously mentioned, the main problem
with the work on interaction management is a lack of
theoretical integration. The work on interaction management,
in turn, could profit from the theoretical integration and
explanation provided by ITT to explicate how social actors
strategically use interaction management for social
influence attempts in social encounters.

Combining the work on ITT with the work on interaction
management provides some clear directions in which to
proceed: (a) the goals of social actors in a specified
social interaction need to be identified, (b) interaction
management needs to be defined and a rationale provided for
selecting interaction management as a line of action

relevant to the actors' pursuit of those goals in the

 

 

 

 

 

17

SPBCified interaction, and (c) rationale and criteria need
to be identified for selecting the behavioral displays that
operationalize interaction management as a line of action.
The dual goal of this study, then, is to provide a rigorous
test of the ITT using interaction management as a line of
action that is appropriate in a specified social

interaction.

Conceptual Integration eh the Interpersonal Tactics Theory
and Interaction Management

Across and within interactions, social actors have
interaction specific goals they want to accomplish (Street 6
Cappella, 1985). The first task is to specify an
interaction. Since all or even a significant portion of the
goals can not be examined in this study, the second task is
to clearly define the dominant social goal in the specified
interaction (Clark & Delia, 1979). Lastly, the achievement
of the goal should be capable of being maximized through the
use of an interaction management strategy.

One promising situation with clearly articulated goals
is a disputatious social interaction. At the beginning of
this paper it was stated that disputatious interactions are
a fact of social encounters. As will be shortly
demonstrated, disputatious interactions can also be clearly
defined. Since, disputes exist when actors are engaged in
meeting goals that are perceived as, or actually are,
incompatible, these opposing goals also provide a clean test

of the ITT. Because of their pervasive presence and

 

 

 

 

 

18

definitional specificity, disputatious interactions are a
800d situational genre for use in this study.

The disputatious social interaction can be characterized
as one in which two or more social actors with opposing
views discuss (talk over) their differences. According to
Webster's Collegiate Dictionary, diSputatious is defined as
a discussion marked eSpecially by expression Of Opposing
views. Social interaction, in this study, implies the
presence Of two or more social actors, that
talk/conversation occupies a central place, and what Goffman
(1963) has called focused interaction: "the kind of
interaction that occurs when persons gather close together
and Openly cooperate to sustain a single focus of
attention." (p. 24)

The goal Of the actors in a disputatious interaction
can be defined as an attempt to reach a settlement Of their
differences that is more favorable tO one's own goals than
to the other's. The ability Of the actor to achieve his/her
goals is thought by many researchers (Brown 6 Levinson,
1968; Duncan, 1983; Goffman, 1969; Street & Giles, 1983) to
be dependent on the actor's knowledge/employment Of the
lines of action (strategies) and behavioral displays
(tactics) that are available for social influence attempts.

While Weinstein identified three considerations for the
selection and implementation or strategies, research in
interaction management suggests that these considerations

should be more formally Specified. The use Of lines of

 

 

 

 

19

action and behavioral displays in ITT parallels the
discussion of strategies and tactics in interaction
management but the latter set will be defined and used in
this study Since they are more commonly found in the
literature. According to Bell and Daly (1984), "the term
strategy describes an abstract category Of largely symbolic
behaviors that, when enacted, are expected tO lead to
[attainment Of a personal Objective]. Nested within a
strategy are tactics, those specific behaviors that
operationalize a strategy." (p. 93). For example, if an
actor has the social goal Of attempting to influence another
interactant by creating a favorable impression, S/he could
select a strategy of sounding intelligent and operationalize
the strategy by using big words. Researchers investigating
the strategies and tactics Of interaction management have
concluded that the strategy chosen by the social actor to
achieve his/her goal(s) must meet a number Of requirements:
(a) be a significant symbol, (b) be purposive, and (c)
maximize the achievement Of the specified goal. In other
words, Weinstein identified the issues relevant tO the
selection of strategies and tactics without taking a stand
whereas the interaction management researchers have
Specified the exact nature by which these issues must be
resolved.

According to a number Of scholars (Bell and Daly, 1984;
Schelling, 1960; Street & Cappella, 1985; Weinstein, 1966)

the chosen strategies must provide symbolic significance

 

 

 

j 1

 

   

n..-

”m ‘

 

I
l
’ o
- (
| I
‘ 1 . ' x

20

 

beyond the content Of the interaction. Symbolic
significance, according to these scholars, means both that
the strategy utilizes symbolic acts -conversational actions
performed through talk - and that this communication creates
a meaning that is broadly shared by the interactants.

The second issue identified by Weinstein in the
selection of a strategy is whether or not the strategy is or
needs to be purposively selected. The central question in
defining this requirement is, "Must an actor intentionally
and consciously implement Strategies designed tO achieve a
particular goal?" The current answer in the communication
literature (see discussions by Berger, 1980; 1986; Spitzberg
8 Cupach, 1984), as regards the intentionality issue, is
affirmative; strategies are intentionally enacted by a
social actor to fulfill Objectives relevant to the goal.
However, intentional selection does not always imply
conscious selection. There is also increasing evidence and
agreement in the current literature that it is unrealistic
to assume that actors always or typically are highly
conscious Of their communication behavior. Consciousness or
awareness is more currently viewed as existing on a
continuum from least awareness in routine, habitual or
highly stripted interactions to most awareness in
interactions where uncertainty is aroused by novel
conditions or violated expectations (Berger, 1986; Spitzberg
& Cupach, 1984). These current interaction management views

Of intentionality (i.e., selection must be intentional) and

 

 

 

 

 

 

 

 

21

consciousness (i.e., it is an irrelevant issue) are adopted
in this study as the second requirement for the selection of
Strategies. Again, this is particularly apparent in a
disputatious social interaction where strategy selection is
highly intentional (Goffman, 1969).

The last requirement for the selection of strategies,
according to interaction management researchers, is that
the selected strategies permit the actor to maximize the
likelihood Of Obtaining the dominant goal in a specified
social interaction. Thus, the chosen strategy must also
have strategic significance or consequences for the
interaction (Goffman, 1969; Smith, 1985). In order to
maximize the likelihood Of goal attainment, then, the chosen
strategy must provide a clear advantage in achieving a
specified goal.

To recap the three requirements for the selection of a
strategy: (a) it should utilize symbolic acts in a way that
has Strategic consequences for the social interaction, (b)
be used intentionally but with varying degrees of awareness,
and (c) it should maximize the likelihood of Obtaining a
specified interactional goal.

One strategy that was proposed to meet these
requirements in disputatious social interactions is
interaction management. As previously stated, the goal Of
social actors in a disputatious social interaction is to
reach a settlement of their differences that is more

favorable to one's goal than to the others'. To accomplish

 

22

one's goal, an actor could attempt to influence the other's
views by making sure his/her own views are brought out in
the discussion, they are clearly understood and they receive
favorable attention. Simultaneously, the actor wants to
ensure that the other's views are not brought out in the
discussion, are not clearly understood, and receive
unfavorable attention. The actor would want tO use a
strategy of social influence that would constrain the
discussion in such a way that serious consideration Of the
actor's views are maximized while serious consideration of
the other's views are minimized. One strategy that is
available to the actor to accomplish this "maximin" goal is
to manage the structure of the interaction so that s/he is
in control of the discussion (Weimann, 1985). In managing
the discussion the actor also has some control over the
amount and type of attention various issues receive.
Strategically, then, management Of the interaction comprises
an attempt to advantageously influence the discussion such
that the outcome which is eventually reached in a
disputatious social interaction is a settlement Of the
issues which is favorable to the actor. The selection and
utilization Of a strategy of interaction management
maximizes the likelihood that the settlement of the actor's
and other's differences will be more favorable to the actor
than to the other. These social influence attempts can be
stated more formally in terms Of a strategy Of interaction

management. Adapting the definitions Of control proposed by

 

 

 

 

23
Weimann and Kelly (1981) and Wilmot (1986), interaction
management is defined as attempts to manage the interaction
structure in ways that advantageously limit the next
possible actions or statements that are available
subsequently to the other(s). Given this definition of a
Strategy of interaction management, the next task is to
provide rationale/criteria for the selection Of tactics to
implement this strategy.

Tactics were previously defined as the communication
behaviors that implement a given Strategy. Several
requirements exist that should be noted about tactics: (a)
they are observable, (b) they Operationalize a strategy, and
(c) multiple behaviors are considered. For Bell and Daly
(1984) and Kolb (1983), to meet the first requirement the
tactics must be observable in the traditional sense of being
a concrete behavior. TO Operationalize a strategy is to
choose those communication behaviors that reflect an
underlying construct which is itself meaningful to peOple's
common interpersonal motives (Street 8 Cappella, 1985). In
other words, most people would recognize a given tactic as
having a particular strategic value. According to KOlb
(1983) and Smith (1985), this approach requires that the
tactics chosen be analyzed and interpreted within the
context Of the particular strategy of which they are a part.
Street and Cappella (1985) also contend that implicit in
this approach is the suggestion that multiple behavioral

features of the strategy receive attention rather than a

 

 

 

 

24

Single behavior isolated from Others. The complex behavior
that interaction is requires the study Of a complex of
behaviors if it is to be fully understood. The tactics that
are available for a social actor to use for social influence
attempts in an interaction, then, should be Studied as
multiple, concrete behaviors that tactically implement a
given strategy.

Given the tactical requirements Specified above, the
selection of tactics to Operationalize an interaction
management strategy should focus Specifically on the
behavioral features that provide a strategic advantage for
the management of social interaction (Smith, 1985).

Street 6 Cappella (1985), among others, maintain that

the ability to manage interaction depends upon the

existence Of selected linguistic and conversational
resources that can be exploited for strategic ends by one or
the other social actors. Street and Cappella (1985)
conclude that acquiring or maintaining control over the
interactants or during interaction is generally accomplished
by behaviors such as turn-taking sequencing, gaze patterns,
speech loudness and rate, use Of gestures, facial
expressiveness, style shifting, long floor holdings, and
non-reciprocal touch. From the previous discussions on
interaction management, the use Of declaratives, questions,'
and terms Of address and reference can also be added to this
list. This brief review clearly Specifies, then, some

concrete behaviors that actors manage in their

I

 

 

25

co“\rersations. What remains unclear and confounded,
hMun-Ever, is the strategic consequence of this management;
to emerge as dominant in the interaction, to attempt to
exercise cOntrol over the interaction, or to maintain
coherent interaction (Orcutt 6 Harvey, 1985). If the
tactics which operationalize a Strategy Of interaction
management are to meet the second requirement for the
selection of tactics, these strategic consequences for
interaction management must be considered separately.
Argyle (1969) suggests that this is a possibility:
to be effective in most social situations it is
essential to be able to control the social interaction.
This does not always mean being the 'dominant' person in
the ordinary sense, but keeping the initiative, and
exercising influence over the relationship, the
emotional tone, and the content of interaction (p. 328).
Orcutt and Harvey (1985) claim that it is also necessary
Since research on the behavioral features that are available
for social influence attempts has been hampered by their
interpretation in most contexts as expressions of power or
dominance relationships or as violations Of coherent social
interaction. In conclusion, rather than viewing the use of
certain features of talk/conversation as associated with
certain relationships or as disruptions Of smooth
interaction, research Should focus in some programatic way
on these features as tactics for an interaction management
strategy for social influence attempts.
The general rationale for tactic selection that emerges

from the previous discussion, is that the actor would want

to select the linguistic and conversational features that

 

 

 

 

26

PTOVide more rather than less of a strategic advantage for

managing the interaction. Based on the previous review of

the features actors manage in their conversations, three
general classes of tactics are distinguished: (a)
conversational organizing tactics, (b) minimizing tactics,
and (c) directive tactics. Conversational organizing
tactics, in general, are used to constrain or encourage the
conversational participation of the other in a social
encounter. Encouraging or constraining the conversational
participation of the other provides a tactical advantage in
structuring both the tOpic(s) pursued in a conversation and
the attention they receive. Tactics in this category

include the following:

(1) MIRROR RESPONSES

Repeating all or part Of what the previous speaker
said.

H: Yeah, I'm gonna tell - don't stOp me.
M: OK
H: OK

Mirror Responses are used to demonstrate agreement
with the other about what is currently under
discussion (Goodwin, 1981). Their use should
encourage the other to continue the deve10pment of a
topic.

(2) POSITIVE BACK CHANNEL CUES
Positively acknowledge the current Speaker's remarks

without interrupting the current Speaker's turn at
talk.

Mm hmm, yes, true, really, yeah, Ok, right, exactly

Positive Back Channel Cues are used to reinforce
another's verbal contributions (Bruner, 1979;
Goodwin, 1981; Yngve, 1970). Their use has been
found to encourage the continuation Of a topic's
development (Duncan G Fiske, 1977).

_.-ll_lf‘

 

 

  

‘.

I
1 . ‘ x
' ' ’ x
i: . "' .» . ,1

(3)

(4)

(5)

(6)

o - f " ...
l/ ‘
a ‘ I .

 

27

NEGATIVE BACK CHANNEL CUES
Indicate disagreement with the current speaker's

remarks and do not interrupt the speaker's turn at
talk.

Un uh, no, sure, right, etc.

Negative Back Channel Cues are used to negate

the other's verbal contribution (Eakins 8 Eakins,
1978; Gumperz, 1982). Their use Should discourage
the continuation of a topic's deve10pment.

AGREEMENT RESPONSES

Current speaker positively acknowledges a previous
speaker's question or assertion at the beginning of
his/her own turn at talk.

Yes, OK, yeah, right, etc.

W: ... and that's how I see things.
M: 0K, and how about you George?

Agreement reSponses are used to diSplay continuing
interest in a tOpic's deve10pment (Bruner, 1979;
Eakins G Eakins, 1978). Their use has been found to

control the progress of a conversation (Ragan, 1983).

DISAGREEMENT RESPONSES

Current speaker negatively acknowledges a previous
speaker's question or assertion at the beginning of
his/her own turn at talk.

NO, that's not true, right, sure, etc.

Disagreement responses are used to counter the
content of a previous turn at talk. Their use
Should discourage or modify the development Of a
topic.

INTERRUPTIONS/TALKOVERS

The previous Speaker's turn at talk is cut off
before an assumed point of grammatical completeness
by the current speaker's turn.

W: Yeah well I think you're
H: Now wait a minute here, all I said was

Interruptions and Interruptive Talkovers are used tO
claim the other's turn at talk (Weimann, 1985;
Zimmerman 6 West, 1975). Their use has been found
to have a subduing effect on the conversational

  

 

 

 

 

28

participation of the other (Zimmerman 8 West, 1978)
which results in topic failure (Fishman, 1978, 1980)
or change (Hirschman, 1973; Schegloff, 1972).

(7) CEDING THE FLOOR
Current speaker is unable to complete a turn at talk
to a point of grammatical completeness.

H: I used to drink a lot

W: He thinks he's real macho
H: I used to drink a lot when I was young
W: He used to

Ceding the Floor is used to relinquish the turn at
talk (Pearson, 1986). Their use should allow
the other to regulate the topic of the conversation.

 

These conversational organizing tactics, then, provide more
or less of a tactical advantage for managing the interaction
by managing the conversational contributions of the
interactants. I

Minimizing tactics are the use of softening/forceful,
mitigating or qualifying words or phrases. The use of these
phrases suggests more or less Of a readiness to defer to any
opposing view or sudden Objections which provides more or
less of a tactical advantage in structuring the process of
conversational interaction. Tactics in this category
include the following:

(1) DISCLAIMERS

Linguistic features which soften an assertion.

I believe, I think, I mean, I suppose, I recommend,
I suggest, etc.

Disclaimers should be used to encourage confirmation
of ideas from the other. Since their use allows for
comment from the other, they should encourage the
other to manage the course of the interaction.

 

 

(2)

(3)

(4)

(5)

 

29

POSITIVE TAG QUESTIONS
One word tags which function as a means of seeking
agreement.

This is your street, right

Positive Tag Questions are used to request approval
or confirmation of ideas from the other (Eakins &
Eakins, 1978; Ragan, 1983). Since their use calls
for comment from the other, they Should allow other
to manage the course of the interaction.

NEGATIVE TAG QUESTIONS
One word tags which function as a means of gaining
information.

You did kill Ms. Jean Brody, didn't you?

Negative Tag Questions are used to challenge the
other's ideas (Ragan, 1983). Their use has been
found to gain a confirming reSponse from the other
(Eakins 8 Eakins, 1978).

HEDGES
Linguistic features marking the absence Of
certainty.

Sorta, sort Of, kind of, you know, perhaps, maybe,
hesitate, not sure, but, etc.

Hedges are used to indicate that the Speaker is not
adamant about his/her point of view but is

willing to accept the other's point of view (Eakins
8 Eakins, 1978). Their use should encourage the
expression of the other's point Of view.

QUALIFIBRS
Linguistic features which soften the assertion.

I believe, I think, I mean, I suppose, I recommend

Qualifiers are used to evade issues or dilute
expressions Of Opinion by the speaker (Ragan, 1983).
Their use Should encourage the expression Of the
other's point of view.

 

 

30

(8) POSITIVE ADVERBIALS OF DEGREE
Linguistic features that understate the degree to
which things are important.

A bit, a little, a little bit, tiny, single, only,
borrow, a sec, etc.

I have a little problem I need help with.

Positive Adverbials of Degree are used by the
speaker to note the lack of importance of the
speaker's ideas (Eakins G Eakins, 1978). Their
use should encourage the expression of the other's
point of view.

(9) NEGATIVE ADVERBIALS OF DEGREE
Linguistic features which overstate the degree to
which things occur or trivialize the degree to which
things are important.

Never, always, just, quite, tiny, little, etc.

SO what's your little problem today?
You always interrupt me.

Negative Adverbials of Degree are used by the
speaker to note the importance or lack of importance

of idea(s). Their use should encourage the
deve10pment of the speaker's own point of view.

(10) REASONS AND JUSTIFICATIONS
Reasons that the speaker gives for imposing on the
other.
Can you give me change for a dollar? I'm late.
Reasons and Justifications are used to excuse,
explain, or request understanding or forebearance
by the speaker (Eakins G Eakins, 1978). Their use
has been found to give the other a choice of how to
respond (Ragan, 1983).
These minimizing tactics, then, provide more or less of a
tactical advantage for managing the interaction by making
statements more or less tentative and absolute which
encourages or discourages conversational contributions from

the other.

 

 

31

Directive tactics vary in the degree to which they impose
belief, agreement or obedience on others. Using directive
tactics provides more or less of a tactical advantage in
structuring the interaction by varying the implied Obligation

imposed on another to respond in the directed way. Tactics

that can be included in this category are the following:

(1) NEED STATEMENTS
Statements that imply that the speaker has a right
to ask the other to comply.

I need that book.

Need Statements are used to seek a specified change
in the other's behavior (Erwin-Tripp, 1978; Fisher,
1969; Kramarae, 1981). Their use has been found to
strugture the progress of a conversation (Ragan,
1983 .

(2) IMPERATIVES
Statements that give a command or make an
accusation.

Get me that book.
Don't take that book.
Open the window.

Imperatives are used to seek a specified change

in the other's behavior (Erwin-Tripp, 1978; Fisher,
1969; Kramarae, 1981). Their use has been found to
strugture the progress of a conversation (Ragan,
1983 .

(3) IMBEDDED IMPERATIVES
Commands that are preceded by a formal addition.
Contains one of the modals can, could, will, would
Subject clause is the addressee
Predicate describes an action physically possible at
the time

Can, could, will, won't you Open the door.

Imbedded Imperatives are used to create the

appearance of giving Options while obligating
another to comply with a request (Ragan, 1983;
Shimanoff, 1983). Their use has been found to

 

(4)

(5)

(6)

 

32

contgol the structure of conversations (Ragan,
1983 .

PERMISSION DIRECTIVES
Statements that appear to ask permission, but which
require action on the part of the other.

Modal (can, can't, could, couldn't, and may) +
beneficiary + have/verb
Focus is on the Speaker's or recipient's activity

May I see that for a minute?

Permission directives are used to give the other the
choice of ignoring the implicit request (Kramarae,
1981; Treichler, Frankel, Kramarae, Zoppi, G
Beckman, 1984). Their use should make it less
likely that the speaker obtains the action or
resppnse s/he is seeking from the other (Pearson,
1986 .

QUESTION DIRECTIVES

Statements that appear to ask a question but that
can also be taken as a request for action on the
part of the other.

Is there any coffee left?
How about one last drink?
What are you laughing at?

Question directives are used to give the other a
choice of ignoring the implicit request (Kramarae,
1981; Treichler, eh el., 1984). Their use

should make it less likely that the speaker

obtains the action or response s/he is seeking from
the other (Pearson, 1986).

HINT DIRECTIVES
Statements which require the work of filling in the
request since it is merely alluded to by the

.speaker.
It's raining out hint for a ride to school
I have a problem hint for assistance

Hint directives are used to give the other the
choice of ignoring the implicit request (Kramarae,
1981; Treichler, eh el., 1984). Their use

should make it less likely that speaker obtains the
action or response s/he is seeking from the other
(Pearson, 1986).

 

 

33

The \lse of these directive tactics by actor, then, provides
more (or less of a tactical advantage for managing the
interaction by constraining the response options available
to the other.

Each of the three categories of tactics discussed above
contain linguistic and conversational features that provide
an actor with more or less of a strategic advantage in
managing the interaction. For example, negative peg
guestions provide more of a strategic advantage than
positive peg questions because the latter allows the actor
to manage the course of the interaction while the former
relinquishes this interaction management to the other. The
strategic advantage distribution of all of the tactics
discussed above is presented in Table 1.

Based on the previous discussion it is expected that a
social actor in a disputatious social interaction will
attempt to achieve his/her goal of resolving the issues in
his/her favor by choosing a strategy of interaction
management and implementing it through the use of related
tactics. A recent investigation by Donohue and Weider-
Hatfield (1986) found that the more interaction management
tactics an actor employed the more likely s/he was to obtain
a favorable settlement of the issues. In a disputatious
social interaction, however, there are at least two actors,
each trying to Obtain the most favorable settlement of the
issues and each using interaction management tactics to

accomplish that goal. Since their goals are in Opposition

 

 

 

34
Table 1

Distribution e: Linguistic and Conversational Features

Associated with Interaction henagement

 

 

Strategic Advantage Strategic Disadvantage

DIRECTIVE TACTICS

Need statements Permission Directives

Imperatives Question Directives

Imbedded Imperatives Hints

MINIMIZING TACTICS

Negative Adverbials Degree Positive Adverbials Degree

Negative Tag Questions Positive Tag Questions
Qualifiers

Hedges

Reasons and Justifications

Disclaimers
CONVERSATIONAL ORGANIZING TACTICS

Disagreement Responses Agreement Responses

Negative Back Channel Cues Positive Back Channel Cues

Interruptions/Talkovers Ceding the Floor

Mirror Responses

 

 

 

 

35

the actor and the other should devote more of their time
and energy to trying to manage the interaction so that they
can structure it to obtain their own goals (Diez, 1987;
Moore, 1986). This focus on interaction management can
generally have one of two consequences for the outcome of a
disputatious interaction.

Two actors with opposing goals will not only attempt to
manage the interaction but will do so in a more overt and
directive way (Moore, 1986). The actor who is more active
and assertive in managing the interaction should be more
successful in obtaining a favorable settlement than the
actor who is more passive or nonassertive in managing the
interaction (Kochan E Jick, 1978; Moore, 1986). One
possible outcome in disputatious interactions is that the
actor who exercises the most interaction management will
obtain his/her preferred outcome.

Goffman (1969) and Weinstein (1966), among others,
expect that the more time the actor and other devote to
managing the interaction in pursuit of a personally
favorable settlement, the less time they can devote to
accomplishing other tasks like reaching a mutually
acceptable agreement. Young (1972) also predicts that in
such disputatious situations the use of such aggressive
tactics may lead to rigidities which impede progress toward
reaching a settlement. The findings of a study conducted by
Lewis and Fry (1977) provide further support; dyads who

reached agreement avoided disruptive tactics. Not reaching

 

 

 

36

an agreement is another possible outcome to disputatious
interactions where the focus is on interaction management.
Based on this discussion the following two general
propositions are proposed:
Proposition 1: The more interaction management tactics
employed in the disputatious interaction

the less likely the interactants are to
reach agreement.

Proposition 2: The more interaction management tactics

a social actor employs in a disputatious
interaction, the more likely s/he is to
obtain a favorable settlement when an
agreement can be reached.

These general propositions can be applied to specific

instances of disputatious interactions such as ones

requiring third party intervention.

Disputatious interactions are characterized by the
presence of two disputants, each of whom wishes and hopes to
extract an agreement that is more favorable to themselves
than the other interactant is willing to provide, and on
occasion, an inability or unwillingness to reach agreement
of their own accord (Rubin, 1980). It is this type Of
interaction that can require third party intervention and
more and more that intervention takes the form of mediation
(Pruitt 6 Kressel, 1985). The use of an interaction
management strategy as it is exhibited by both the
disputants and a mediator in a mediated disputatious social

interaction needs to be considered.

According to Moore (1986), "Mediation is the

 

 

37

intervention into a dispute or negotiation by an acceptable,
impartial, and neutral third party who has no authoritative
decision-making power to assist disputing parties in
voluntarily reaching their own mutually acceptable
settlement Of issues in a dispute" (p. 14). The goal of the
disputants in mediation is to resolve the dispute in a way
that is favorable to their own interests. The goal of the
mediator, on the other hand, is to effect a mutually
acceptable agreement (see sample agreement in Appendix A).
For the mediator to accomplish this goal requires that s/he
have both strategic flexibility and the opportunity to
intervene. In terms of strategic flexibility, the mediator
needs to be able to analyze and assess the interaction and
design effective intervention strategies to enpower the
disputants to bargain with one another and move closer to an
agreement (Donohue, Allen, & Burrell, 1985; Moore, 1986).
Equally as important, the mediator needs to be able to
regulate the interaction among the disputants so that s/he
can gain access to the interaction and implement the
selected intervention strategies (Fisher, 1983; Wall, 1981).
Some evidence in support of the position stated above has
recently been presented by Donohue & Weider-Hatfield (1985).
They found that mediator control over the course and
structure - frequency and direction of talk and the
frequency of interruptions/talkovers - of the mediation
interaction was critical in accomplishing the mediator's

goal of formulating a mutually acceptable agreement. These

 

 

38

results also provide support for the contention that the
mediators must have not only control over the interaction
but more control than the disputants. One key to reaching a
mutually acceptable agreement in mediation, then, is for the
the mediator to manage the interaction and manage it more
than the disputants. This discussion, in turn, suggests the
following general propositions:
Proposition 3: The mediator must control the
interaction through the use of
interaction management tactics if a

mutually acceptable agreement is to be
reached in mediation.

Proposition 4: The mediator must use more interaction
management tactics overall than the
disputants if a mutually acceptable
agreement is to be reached in mediation.

These propositions will now be applied to the specific

situation of divorce mediation in the form of hypotheses.

Hypotheses

Based on the previous discussion it is expected that
both the disputants and the mediator in a mediated
disputatious social interaction will attempt to achieve
their goals by reaching a final agreement on the issues that
is more favorable for themselves than for the other
interactants; for the disputants this means that the final
agreement coincides with their preferred agreement and for
the mediator this means that the final agreement is mutually
acceptable to the disputants. The disputants, then, will

want to choose a strategy of interaction management so that

 

39
they can structure the interaction to gain acceptance for
their proposals (Rubin, 1980). The mediator will also want
to manage the course of the mediation in.a way that is
productive for the mediation to accomplish it's goal of
reaching a mutually acceptable agreement (Haynes, 1981).
The disputants and the mediators will each want to manage
the interaction in pursuit of their goals. Both the
disputants and the mediator should choose a strategy of
interaction management to accomplish their goals and
activate that strategy through the use of the related
tactics. Based on previous discussion, it is expected that
the more interaction management tactics implemented in
mediations the less likely it is that the interactants will
be successful in reaching a mutually acceptable agreement.

These prOpositions can be stated more formally in the

following hypothesis:

Hypothesis 1 Those mediations which have higher
levels of interaction management across
all parties will less frequently result
in an agreement than those mediations
which have lower levels of interaction
management across all parties.

Several reseachers (e.g., Donohue and Weider-Hatfield,
1986; Barsky, 1983; Saposnek, 1983) have also proposed that
mediation is more likely to result in a mutually acceptable
agreement when the mediators have access to the interaction
and can regulate the interaction management of the

diSputants. It is more probable, then, that the mediator

will achieve his/her goal of reaching an agreement that is

 

40
acceptable to both disputants when the mediator exercises
more interaction management than the disputants, who in
turn, exercise equivalent levels of interaction
management with each other. It is also expected that the
mediation will not result in an acceptable agreement when
the mediator exercises less interaction management or vies
for interaction management with either or both of the
disputants or the diSputants vie for interaction management

with each other.
Hypothesis 2: Those mediations using lower
levels of interaction management
displayed by the mediators than by the
disputants will result less frequently in
agreement than those mediations in which
higher levels of interaction are

displayed by the mediators than by the
disputants.

Additional questions that will be addressed by this
research are the differential use of interaction management
across time in divorce mediation. Current research in
mediation (e.g., Rubin, 1980) suggests that different
behaviors are required at different times in the mediation
process but Donohue, Diez, and Weider-Hatfield (1985)

concluded that very little conceptual effort has been

devoted to determining what kinds of communication acts

prevail during any specific time period. Work on the

Stages/phases of mediation provides a starting point from
which to research both the time periods and the related

communicative acts.

Vanderkooi & Pearson (1983) in their review of the

 

41
divorce mediation models conclude that,

Most writers view mediation as a progression of stages.
Generally, these include: (1) setting the stage by
providing a neutral setting , introducing oneself as a
mediator, establishing ground rules and gaining the
disputant's commitment to mediation; (2) defining the
issues by eliciting facts and expressions of needs,
desires and feelings; (3) processing the issues by
managing emotions, encouraging empathy, narrowing
differences, exploring solutions and maintaining
positive momentum; and (4) reaching a settlement and
assuring its implementation (p. 558).

Gulliver (1979) and Bryant and Patton (1987), on a less

Optimistic note, add, respectively; that phase two is also

 

characterized by disagreements over the issues as well as
clarification and that in phase three the outcome will be
either a consensual agreement or a declaration of a major
impasse which can not be resolved through mediation.
Inclusion of these statements in Vanderkooi and Pearson's
review provides a more complete, although less positive,
description of the phases of mediation; mediation begins
with a phase where information is gathered and the issues
are clarified, moves to a phase where preliminary demands
and Offers produce an emphasis on differences, and moves to
the final phases were the narrowing of differences results
in an emphasis or stalemate on agreement.

Further support for these phases of mediation is
provided by their similarity to the phases found in the work
on small groups; an initial stage marked by defining the
situation and task, a middle phase marked by differences
and confrontation, and finally a phase marked by agreement

(see Hare, 1973 for a review). There is also support

42
(Kessler, 1978) for the contention that the Optimum
movement is to systematically progress through the phases.
Fisher (1983) in his work on small groups maintains that the
decision point is the critical bottleneck in this
progression. Gulliver (1979), coming from a mediation
perspective, also expects the transition from phase two to
phase three to be the most problematic because it involves a
Shift from opposition and hostility to coordination and even
COOperation.

Based on this discussion, it is expected that the
emphasis on differences in the middle phase Should lead to
the use of more interaction management by the disputants and
the mediator both in mediation sessions were a mutually
acceptable agreement is reached and where it is not reached.
It is further expected, however, that in the mediatiOn
sessions where a mutually acceptable agreement is reached,
the level of interaction management should decline in the
third phase as the mediator is able to manage the
interaction and move the disputants closer to agreement.

Based on this discussion the following additional hypothesis

is advanced:

° hose mediations which have higher .
HYPOtheSIS 3 Ievels of interaction management during
Phase 3 than during Phase 2 will less
frequently result in agreement than
those mediations which have lower levels
of interaction management during Phase 3

than during Phase 2.

Finally, work needs to be done to determine what kinds of

 

43
communicative acts facilitate the reaching of a mutually

acceptable agreement.

Additional Qpestions

Although the actual linguistic and conversational
tactics used by the participants in mediation should
coincide with their choice to communicate a strategy of
interaction management, it is also of value to determine ;

which of the interaction management tactics are most F

 

important to reaching an agreement in mediation. Given that
mediators can only focus on a limited number of tactics at
any one time, it is important to determine which tactics are
related to the creation of mutually acceptable agreements.
Mediators can then be taught to recognize and implement the
most useful tactics. Since there is little theoretical or
empirical rationale for predicting what pattern of directive
use tactics, minimizing tactics, or conversational

organizing tactics will be the most useful choices, the

following exploratory hypothesis is advanced:

Exploratory Hypothesis 1: What subset of interaction
management tactics are most

associated with mediations
in which agreement 15
reached?

CHAPTER II
METHOD

Sample e; Interactions

In order to examine how participants structure
interactions, criteria to determine clear examples of
appropriate interactions need to be identified and met by
any sample interaction. For the purposes of this study, it
is necessary that the sample encounters are social
interactions where the participants are involved in face-to-
face interaction with talk as the primary resource. Second,
it is important that the participants have divergent aims,
motives, or interests on which it is imperative for them to
reach an agreement. These criteria will be met in the use
of transcriptions of actual divorce mediation sessions.

In divorce mediation there are at least three
participants; two divorced or divorcing spouses and one or a
team of mediators. In the cases that are analyzed in this
study, the spouses (parents) are typically living separately
and are differing over custody arrangements for their
children (Pearson & Theonnes, 1984). More Specifically, the
parents' arrangements are mutually exclusive (i.e, both want
sole custody or the children on Christmas day). The

immediate goal of each parent, then, is to have his or her

44

 

 

 

 

4S

c“5t0dy arrangement accepted (Dibble, 1984; Pearson G
Thoennes, 1984; Saposnek, 1983; Rosanova, 1983). The goal
of the mediator is the development of an agreement that is
acceptable to each parent and best for the children
(Rosanova, 1983). In attempting to resolve their
differences, the parents and the mediator meet face-to-face
to discuss their differences. Since the ultimate authority
in mediation belongs to the parents themselves and does not
arise from the legal system, talk/conversation is the
primary resource available to the participants to use to
reach an agreement. Thus, the sample interactions meet the
first criteria.

Fulfilling the second criterion, the parents enter into
mediation because they have been unable to reach an
agreement on their own. The mediator, in turn, wants to help
the parents reach a mutually beneficial and agreeable
solution which may put his/her goals at odds with the goals
of either or both of the parents. Although the parents are
under no legal requirement to reach an agreement through
mediation, it is imperative that an agreement is eventually
reached in some manner. The decision the parents need to
make, then, is whether to settle in mediation where they
retain some control over their decisions or take their
chances in court. Kelly (1984) provides evidence to suggest
that one attraction of mediation to the disputants is their
control over issues so that reaching an agreement through

the mediation process can be considered as imperative to the

 

46

Parents. Hence, divorce mediation provides an ideal
context for the use of interaction management as a social
influence attempt.

Twenty transcriptions of pre- and post-divorce
custody/visitation sessions (ten in which agreement was
reached and ten in which agreement was not reached) were
used in this study. The transcriptions were made from
audiotaped mediation sessions of approximately one hour.
Each session took approximately twenty-four hours to
transcribe following the rules of transcription presented by
Schenkein (1978). In addition, any references to last names
and cities were blacked out to ensure the anonymity of the
parties.

The audiotapes were selected from the corpus of 80
collected by the Divorce Mediation Research Project (1981-
1984), funded by the Children's Bureau of the U. S.
Department of Health 8 Human Services (90-CW-634) and
administered by Jessica Pearson, Director of the Research
Unit of the Association of Family and Conciliation Courts,
in Denver, Colorado. The twenty sessions used in the
present study were selected because the recordings were
intelligible for the most part and the number of
participants (husband, wife, and a mediator) was
equivalent. These particular Sessions came from the various
branches of the Los Angeles Conciliation Court. The format
of the mediation service at this court, the sources of user

satisfaction, and the user evaluations of the mediation

 

47

ontcomes are discussed in Pearson and Thoennes (1984).

Measures

The measures identified and defined to test the
hypotheses were designed to (a) examine, as a group, the
tactics that operationalize the strategy of interaction
-management and (b) better capture the 'interactional' nature
of the language and conversational features used in the
mediation sessions. Three categories of linguistic and
conversational features were specified as indicators of the
use of interaction management tactics. The three categories
of interaction management tactics that operationalize the
'minimax' strategy are directive tactics, minimizing
tactics, and conversational organizing tactics. As
previously indicated (see Table 1) the tactics that
relinquish the strategic advantage in managing the
interaction include, respectively: question imperatives and
hints; ceding of the floor when interrupted or talk is
simultaneous, positive back channel cues, and repetitions;
and hedges, positive tags, qualifiers, positive adverbials
of degree, and reasons and justifications. The tactics that
promote the strategic advantage in managing the interaction
include need statements, imbedded imperatives, and
imperatives; negative back channel cues, interruptions; and
negative tags, and negative adverbials of degree. The
tactical choices that relinquish interaction management were
coded as a -l and the tactical choices that enhance

interaction management were coded as a +1. The score across

 

48

the three subcategories of interaction management was the
sum of these values. A positive score indicated the use of
the more advantageous interaction management tactics and a
negative score the use of the less advantageous interaction
management tactics.

The second concern in coding interaction management is
how to best conceptualize and measure the structural
patterns of the language and conversational features used in
divorce mediation. Koper, Donohue, and Stahle (1985)
utilized an actional view of communication (i.e.,
communication as a transmission of a message from a sender
to a receiver) in their study on the effects of mediator
communication strategies on the disputants' levels of
language intensity in divorce mediation. A composite
measure of Ianguage intensity was used which indicated the
total amount of language intensity one participant used in
the mediation sessions. The results indicated that
successful mediators mark their utterances with
significantly less intensity in successful mediation
sessions than in unsuccessful mediation sessions. The
actional view of communication utilized by Koper, Donohue,
and Stahle (1985) suggests that the structural patterning of
the language and conversational features used in divorce
mediation is best captured by investigating how participants
communicate he eeeh ehhee. Language and conversational
features are thought to have a one way effect in the

interaction. Later research by Donohue and his colleagues

 

49
has utilized a more interactional view of communication.

An interactional view of communication would assume that
the structural pattern of the language and conversational
features used in divorce mediation is best captured by
investigating how participants communicate high eeeh ehhep.
Following this view, Donohue and Weider-Hatfield (1986)
investigated the language features the mediators and
disputants used to negotiate control issues in divorce
mediation sessions. The amount of control was

operationalized as the exchange of interruptions and

 

talkovers between the husbands and the mediators, the wives
and the mediators, and the mediators and the disputants.
Donohue and Weider-Hatfield (1985) concluded that the
mediators were less successful in reaching agreement when
they allowed one disputant (the husbands) to control the
allocation of floor time. Since the speaker-to-participant
variable was not completely crossed in the study, it is
unclear whether the mediators were also controlling the
allocation of floor time with the husbands or the husbands
with the wives. What is needed is a way of measuring
language and conversational features which is based on the
same interactional view of communication that is assumed in
the studies.

The interactional view is taken in this dissertation.
It is assumed that interaction management functions in
divorce mediation by the way in which the interactants

communicate with each other. To better capture the

 

50

‘interactional' nature of interaction management, each
tactic was assigned one of six codes for speaker to
addressee/referent; mediator to husband, mediator to wife,
husband to wife, husband to mediator, wife to mediator, and
wife to husband. When a communication code choice was
judged as addressed to two participants, the speaker

to addressee/referent code was assigned twice, once for each
of the addressee/referents.

Each numbered turn at talk in the transcripts was
content analyzed for the interaction management tactics by
speaker to addressee/referent it contained. Five teams of
coders were trained to identify each occurrence of a tactic
as well as how to assign that tactic to a category (see the
coding instructions in Appendix B). One transcript section
containing two hundred and sixty-nine turns at talk was used
to assess the unitizing and interpretive reliability of the
coders. The unitizing reliability was .87. This unit
reliability for the recognition of the linguistic and
conversational features that comprise interaction
management indicates consistent identification across
raters. The inter-rater reliability for the twenty-two
coded tactics were computed using percent agreement (Folger,
Hewes, & Poole, 1984). The inter-rater reliabilities for
the thirty tactics are presented in Table 2. Clearly, the
language and conversational features that comprise
interaction management can also reliably be categorized by

trained coders.

 

 

 

51
Table 2

Inter-rater Reliabilities for the Coding e; Interaction

Management

 

Tactics Percent

Agreement

 

Directives

Need Statements
Imperatives

Imbedded Imperatives
Permission Directives
Question Directives
Hints

Minimizing Tactics

Hedges

Qualifiers

Positive Adverbials of Degree
Negative Adverbials of Degree
Disclaimers
Reasons/Justifications
Positive Tag Questions
Negative Tag Questions

Conversational Organizing Devices

Positive Back Channel Cues

Agreement Responses

Mirror Responses

Negative Back Channel Cues

Disagreement ReSponses
Interruption/Talkovers

Gains Floor with Interruption or Talkover
Cedes Floor

 

 

 

 

 

 

 

 

 

 

 

 

 

 

52

In order to allow for meaningful comparisons between
sessions varying in length, the score for each utterance was
divided by the total number of words in the session. In
this way, each utterance is viewed as contributing a zero or
greater proportion to the distribution of interaction
management in the session. Weighting each utterance's score
proportionate to the length of the session means, then, that
the longer the session, the less weight was given to the use

of an interaction management code choice in that utterance.

 

CHAPTER III
RESULTS

Tests efi the Hypotheses
The first hypothesis predicted that in those mediations

which have higher levels of interaction management there

would less frequently be agreement. The second hypothesis

 

predicted a patterning Of the interaction management among
the participants such that those mediations, where the
mediators expressed less interaction management than the
parents, would less frequently result in agreement. The
third hypothesis predicted a relationship between outcome
and time, such that those mediations, in which there was
more interaction management in Phase 3 than Phase 2, would
less frequently result in agreement. Since both the phases
(Hypothesis 3) and the patterns (Hypothesis 2) are subsets
of the total amount of interaction management; patterns,
phases, and the total amount should be correlated. Given
that these variables are not independent each hypothesis was
tested separately using the appropriate statistics. The
results for each of the hypotheses will be addressed in
order.

Since Hypothesis 1 contained one continuous (interaction

management) and one dichotomous (success of mediation

53

 

 

 

54

535510“) Variable, it was tested using a point biserial
correlation (Cohen and Cohen, 1966). Successful mediation
sessions were coded as O and unsuccessful mediation sessions
were coded as 1. The decision rule used was if correlation
coefficient was Significantly greater than zero then the
null hypothesis was rejected. The resulting correlation
revealed a nonsignificant relationship between the observed
level of interaction management used in successful and
unsuccessful mediation sessions, I = .23, p < .55, df = 21.
In fact, the fairly equal distribution of interaction
management across mediation outcomes (successful i = .40;
unsuccessful K = .37) suggests that the global use of
interaction management to either gain or relinquish control
of the interaction is not observably related to success in
divorce mediation.

Hypothesis 2 was also tested using a positive point
biserial correlation as the rejection rule. In relation to
hypothesis 2, the results reveal a significant relationship
between the amount of interaction management used by (a)
mediators in addressing/refering to the husbands (e = .61,

p < .05, df = 21) and the wives (1 = .57, p ( .05, df = 21);
(b) husbands in addressing/refering to the wives (L = .45,

p < .05, df = 21); and (c) wives in addressing/refering to
the mediators (e a .50, p ( .05, df = 21) and husbands (e

= .69, p < .05, df = 21). A nonsignificant relationship was
revealed between the observed level of interaction
management used by the husbands in addressing/refering to

the mediators (L = .29, p < .05, df = 21). The means

 

55

presented in Table 3 indicate that (a) the mediators used
lower levels of interaction management with the husbands and
higher levels with the wives in successful than in
unsuccessful mediation sessions; (b) the husbands used lower
levels of interaction management with both the mediators and
the wives in successful than in unsuccessful mediation
sessions; and (c) the wives used higher levels of
interaction management with the mediators and lower levels
with the husbands in successful than in unsuccessful
mediation sessions. With respect to the results, the data
seem to point most interestingly to the level of interaction
management used between the mediators and the husbands.
Hypothesis 3 was also tested using the point biserial
correlation. The resulting correlation revealed a
nonsignificant relationship between the amount of
interaction management used over time in the successful (e
= .06, p ( .01, df = 21) and unsuccessful (L .08, p < .01,
df = 21) mediation sessions. The pattern of means presented
in Table 4 reveal that for both successful and unsuccessful
mediation sessions, interaction management did not diminish
over time; more interaction management is used in Phase two
than in Phase one and more in Phase three than in Phase two.
Given the results obtained for Hypothesis 2 and
Hypothesis 3, an additional analysis was run to determine if
the interaction management used by the speaker-to-
addressee/referent (Hypothesis 2) differed over time

(Hypothesis 3). To probe the relationship in more detail,

 

56

Table 3

Meeh Interaction Management for Speaker he Addressee he

 

Successful and ghsuccessful hediatieh Sessieh

 

 

Mediation Session

 

Speaker to Addressee Successful Unsuccessful
It N it N
Mediators to Husbands 1.25 393 1.35 287
Mediators to Wives 1.35 364 1.30 248
Husbands to Mediators 1.20 464 1.21 204
Husbands to Wives 1.09 307 1.12 173
Wives to Mediators 1.10 397 .98 302

Wives to Husbands 1.16 364 1.27 293

   

 

57

Table 4

fleeh Interaction Management hy Time for Successful and

 

ghsuccessful Mediatieh Sessions

 

Mediation Sessions

 

 

Time Successful Unsuccessful
it N r N
Phase 1 1.09 1142 2.18 806
Phase 2 2.10 1177 2.27 815

Phase 3 2.13 1106 2.33 844

W
- -..,“ \ -

 

 

 

 

58

the means for each of the speaker-to-addressee/referent
codes were computed across the three phases. The results
are presented in Table 5. Three clearly recognizable
patterns emerge from these results. There is less
interaction management in successful mediation sessions than
in unsuccessful mediation sessions when the mediators
address/refer to the husbands or the husbands

address/refer to the wives. There is no decrease in
interaction management over time when the mediators
address/refer to the wives or the husbands address/ refer to
the wives. There is a decrease in interaction management
over time when the mediators address/refer to the husbands.
These results add further credence to the importance of
studying interaction management by speaker-to-

addressee/referent.

Exploratory Hypothesis

The Exploratory Hypothesis investigated the relationship
between the use of particular subcategories of interaction
management and the outcome of the mediation sessions. To
test the Exploratory Hypothesis, 3 discriminate analysis was
conducted using successful and unsuccessful mediation
sessions as the discriminating groups. Ratios for each of
the subscales of interaction management (conversational
organizing devices, minimizing expressions, and
declaratives ) were computed. A stepwise procedure and the
method of minimum Wilk's lamda was employed. The resulting

equation failed to discriminate the interaction management

 

Table 5

59

2522 Interaction Management hy Speaker-to-Addressee and Time

 

 

————

Mediation Sessions

 

 

Speaker to Addressee Successful Unsuccessful
Y N i N

Mediators to Husbands

Phase 1 1.23 190 1.26 91

Phase 2 1.00 132 1.17 79

Phase 3 1.06 71 1.37 117
Mediators to Wives

Phase 1 1.33 142 1.12 94

Phase 2 1.23 137 1.22 57

Phase 3 1.19 85 1.43 97
Husbands to Mediators

Phase 1 1.10 196 1.03 70

Phase 2 1.16 155 1.09 64

Phase 3 1.08 113 1.20 70
Husbands to Wives

Phase 1 .95 74 .92 48

Phase 2 .98 126 1.13 82

Phase 3 .94 107 .79 43
Wives to Mediator

Phase 1 .98 157 1.07 118

Phase 2 1.02 142 1.15 98

Phase 3 .98 98 1.13 88
Wives to Husband

Phase 1 .98 144 1.14 107

Phase 2 1.04 124 1.22 102

Phase 3 1.02 94 1.08 85

 

60

used in successful mediation sessions from that of the
unsuccessful mediation sessions (Wilk's lamda s .991, F =

1.03, p < .94).

 

CHAPTER IV

DISCUSSION

This dissertation had three broad aims; (a) to test a
coding scheme for measuring interaction management, (b) to
test hypotheses derived from Weinstein's ITT, and (c) to
investigate the use of interaction management in divorce

mediation. Each of these aims will be discussed in turn.

Interaction Menagement Coding Scheme

The interaction management coding Scheme used in this
dissertation was constructed to (a) better capture the
'interactional' nature of the language and conversational
features used in the mediation sessions, (b) study, as a
group, the tactics that operationalize the strategy of
interaction management, and (c) provide a further test of
Hewes' claims regarding unitizing and interpretive
reliability.

The coding scheme developed in this dissertation is
based on the assumption that the linguistic and
conversational features comprising interaction are best
studied as the complex, simultaneous, and sequentially
organized phenomena they are. The first goal was to make
coherent the complex array of linguistic and conversational
features that comprise interaction by studying together

61

 

62

those communication behaviors that serve the similar
function of interaction management. Linguistic and
conversational features were chosen that had an empirically
strong and conceptually significant relationship to
interaction management. In this way the measure of
interaction management reflects the single combined potency
of the linguistic and conversational features. With the
employment of this functional grouping, the power of a
summary statement by functions is far greater than one based
on individual features. This power could be further
increased if the measure of interaction management was
comprised of differentially weighted scale values for the
linguistic and conversational features rather than the equal
weightings that were used in the present coding scheme.

The interaction management coding scheme used in this
dissertation assigns the same equal weight (1 for the use of
more advantageous features and -1 for the use of less
advantageous features) to every linguistic and
conversational feature. For example, a question imperative
like 'May I interrupt here?‘ would receive a score of -1
while 'May I interrupt, please?’ would receive a score of -2
(-1 for the question imperative and -1 for the politeness
form please). While these weighted scores are an
appropriate first step in the measurement of interaction
management, work on the politeness function of directives
suggests an even more promising direction (c.f., Carrell a

Konnecker, 1981; James, 1976).

 

63

James (1976) developed a coding scheme to measure the
politeness of directives using the paired comparison
procedure. Using his politeness coding scheme the question
imperative, 'May I interrupt here?’ would receive a score of
3.34 while 'May I interrupt, please' would receive a 4.070
and a need statement like 'I need to interrupt here' would
receive a score of 0.000 and 'I need to interrupt here
because...‘ would receive a score of 1.081. Although the
interaction management function of the directives would
probably be the inverse of the politeness function, using
the paired comparison procedure would be the first step in
improving the measurement of the linguistic and
conversational features of interaction management. One way
to proceed would be to use the paired comparison procedure
to develop scales for each of the three subcategories of
interaction management.

Even an interaction management coding scheme produced by
the paired comparison procedure would still have validity
problems when measuring an utterance such as 'I think you
have to be quiet'. 'You have to be quiet' would be coded as
an imperative and might receive a higher directive score
like 4.070 and 'I think' a lower minimizing expression score
like 1.980. The utterance would receive a composite score
of 6.050. A coding scheme using the paired comparison
procedure is still not sensitive to the occurrence of
linguistic and conversational features in a stream of talk.

Is the relationship between the two features simply additive

 

 

 

64

or <ioes the occurrence of 'I think' before the directive
reduce the degree of interaction management of the
directive? It is a long term (perhaps lifetime?) goal of
the researcher to work on a measurement scheme for
interaction management that is sensitive to the interaction
management value each linguistic and conversational feature
would receive in a stream of talk.

In this dissertation an interaction view of
communication is assumed in which the interaction management
used by interactants is the result of an exchange of
interaction management tactics. To more accurately capture
the 'sequentially interactional' aspect of interaction
management, the interaction management score that an
interactant received was measured as a function of the
interaction management between the interactant and a given
addressee/referent in the interaction rather than a
composite score which reflected his or her interaction
management with all the addressee/referents in the
interaction. As will be seen in the discussion of
Hypothesis Two, measuring interaction management by speaker-
to-addressee/referent adds critically to understanding the
effect of interaction management in disputatious
interactions.

Hewes (1985) suggested that useful, nonarbitrary
interpretation requires consistency in both the
identification of the units to be categorized or rated

(unitizing reliability) and consistency in the labels or

 

65

TatiJags assigned to those units (interpretive reliability).
Unitizing reliability answers the question; 'Did the coders
all reliably unitize the transcripts into some specific
number of interaction management units?'. It provides
quantitative verification that we are identifying
consistently phenomenon in a text across coders. The
unitizing reliability in this dissertation was .87 which is
reasonably high. Thus, coders can reliably unitize the
transcripts into some specific number of interaction
management units. Once we have established that coders can
reliably unitize, we want some quantitative assurance that
common labels are attached consistently to the units
(interpretive reliability).

Hewes (1985) suggests that category by category indices
of interpretive reliability be used rather than global
indices like Guetzkow's P (Guetzkow, 1950). He claims
that the'former global indices miss a potential source of
systematic bias; the possibility that certain categories may
be harder to use or more ambiguous, thus possessing lower
reliabilities.

To further test Hewes' claim both Guetzkow's P and
category-by-category reliabilities were computed for the
data coded in this dissertation. Using the Guetzkow P, the
global reliability is .92. Based on this high reliability
one would conclude that the overall interpretive reliability
for the interaction management coding scheme is quite good.

This global rating could Obscure, however, lower

 

 

 

 

66

reliabilities for particular categories that are in need of
more work.

To determine if the conditions that promote category-by-
category bias obtain in this data, category-by-category
reliabilities were computed using percentage agreement
(Folger, Hewes, 6 Poole, 1984). The resulting reliabilities
(as presented previously in Table 2) ranged from .89 to .99.
According to Folger et a1. (1984), if the number of
categories is reasonably large (more than 3) and the
reliabilities are reasonably high (.9 or higher), it is very
unlikely that there is a serious problem with bias. Since
all but one of the categories (Positive Adverbial of Degree
, .89) had reliabilities of .9 or higher, it is unlikely
that there was a serious problem with bias in the measure of
interaction management used in this dissertation.

The results of the more particularistic measure Of
reliability does, however, point out that there is still
room for improvement with particular categories in the
interaction management coding scheme; the positive (.89)
and negative (.90) adverbials of degree tactics that have
the lowest reliabilities. In computing the unitizing
reliability, it appeared to the author that one possible
problem with coding these tactics was identifying their use
in context. Take the following transcript segment as an
example:

001 M: Would you think of sharing um the raising of

your children where you might live near one

another and they would have your two
households. Have you considered that?

 

67

002A W: Well, I have a little problem with that.
002B W: Well, I don't know
003 H: So what's your little problem with THIS plan?

If 'little' is coded as an adverbial of degree without
considering the context than it should be coded as a
positive adverbial of degree since it functions to minimize
the impact of what is being said. If 'little' is coded,
taking the context into consideration, in utterance 002A it
still functions as a positive adverbial of degree. In
utterance 003, however, 'little‘ should be coded as a
negative adverbial of degree Since it functions to
trivialize the importance of the wife's objection. To
ensure higher reliabilities for these categories in the
future, coders should be trained to be (a) less global or
automatic in assigning adverbials like 'little' to the
positive category and responses like 'really' to the
negative category and (b) more sensitive to the context in
which the adverbials are used.

Given that this research is the first in a series of
studies dealing with interaction management in disputatious
social situations, future research is indicated using this
construct. Most immediately, further work needs to focus on
improving the coding reliability for some of the more
contextually sensitive categories. Future research will be
directed to refining the interaction management scores
assigned to the linguistic and conversational features that

comprise the interaction management coding scheme.

 

     

 

68

validation of a revised interaction management coding scheme
might be accomplished through its application in other
disputatious communication contexts such as labor -
management negotiations.

A second goal of this dissertation has been to provide
general validation for Weinstein's ITT by providing a more
rigorous test of the theory then has previously been
conducted; identifying a specific interactional goal and
related strategy/tactics to accomplish the goal. It was
hypothesized that the goal Of actors in disputatious social
interaction is to reach a settlement of their differences
that is more favorable to one's goal(s) than to the other's
goal(s). One way actors can accomplish their goals is to
manage the interaction in such a way that the discussion is
favorable to their own position and unfavorable to the
other's position. The strategy of interaction management is
an attempt to influence the discussion in such a way that
the outcome that is eventually reached is a settlement of
the issues more favorable to the actor than to the others.
More specifically, the effect of interaction management on
the outcome of the divorce mediation sessions, the
disputatious social interaction investigated in this
dissertation, was hypothesized to be a function of the
overall level of interaction management, the interactants'
levels Of interaction management, and the level of
interaction management used over time. An examination of

the results revealed some interesting insights and avenues

 

69

£°T future research into interaction management and ITT.

The results from Hypothesis One showed that the level of
interaction management is not a significant predictor of the
success of divorce mediation sessions. These results
suggest that between the mediation sessions, the
interactants are using no particular level Of interaction
management to either gain or relinquish management of the
interaction. Indirectly, these results provide support for
the ITT postulate that social actors attempt to successfully
achieve their own purposes he relation he others.

Management of the interaction is not merely a function of
the level of interaction management used but the negotiation
of the management of the interaction by the interactants in
relation to each other.

The picture of the interactants' interaction management
that emerges from the results for Hypothesis Two is that (a)
in the successful sessions the mediators and the wives are
more actively involved in managing the interaction than the
husbands and (b) in the unsuccessful sessions the husbands
are more actively involved in managing the interaction than
the wives and mediators. In the successful sessions
(agreement reached) the mediators use a higher level of
interaction management with the wives and lower with the
husbands. In the successful sessions it appears that the
mediators are able to use interaction management to provide
the necessary structure to move the mediation toward a

mutually acceptable agreement. An agreement is not reached

   

70

in the sessions where the level of interaction management
used by the participants is mixed. The parents' attempts at
management of the interaction, and the mediator's attempts
to manage the interaction appears to sabotage agreement
making. These conclusions are also partially supported by
the Donohue and Weider-Hatfield (1986) test of their
mediator competence model. They conclude that "When the
mediator loses control by allowing the disputants to
differentially talkover him/her or differentially interrupts
the disputants, the mediator is less capable of providing
the structure that is needed to reduce conflict intensity."
(p.35). Both the results of this dissertation and the
Donohue and Weider-Hatfield study suggest that the
mediators' control of the interaction is critical to the'
success of the divorce mediation sessions.

Perhaps the most intriguing question about these
findings concerns the levels of interaction management used
between the interactants in divorce mediation sessions. In
the successful mediation sessions the highest levels of
interaction management are used by the wife and the
mediator; in the unsuccessful, by the mediator and the
husband. The additional results from the post hoc analysis
Show that over time there is (a) a decrease in the level of
interaction management used by the wives to the mediators
and the mediators to the wives.

These results suggest that in the successful mediation

sessions, the wives and the mediators are negotiating the

 

 

 

E31

ta‘
cc

I)?

 

71

management of the interaction and the mediators are able to
take control of the interaction. Once the mediators have
control of the interaction their level of interaction
management decreases as they focus on other mediation
concerns that move the parents closer to agreement. In the
unsuccessful mediation sessions, all the participants are
negotiating the management of the interaction and no one is
able to take control Of the interaction, including the
mediator. The increase of the mediators' levels of
interaction management in the unsuccessful sessions may be
due tO poor mediation skills on the part of the mediators or
the mediators attempting to reach an agreement until it is
clear that there is no realistic hope of reaching a mutually
acceptable agreement.

As for the parents, the present dissertation finds that
in both successful and unsuccessful sessions the wives use
higher levels of interaction management than the husbands
but the wives use higher levels of interaction management
in the successful as opposed to the unsuccessful mediation
sessions. The additional results from the post hoc analysis
shows that over time there is no decrease in the level of
interaction management used by the mediators and the wives
to the hosbands.

The results are exactly the opposite of those found for
the wives and mediators in this dissertation but supportive
of those found in the Donohue and Weider-Hatfield study.

They found that the wives interrupt and talkover the

 

 

-. - = _

 

 

72

mediator significantly less than their husbands in the
unsuccessful sessions but the wives used significantly more
language intensity. The picture of the wives' involvement
in the unsuccessful sessions that emerges from both sets of
results is seemingly contradictory. Donohue and Weider-
Hatfield (1986) conclude that the wives may be responding to
a lack of control over the interaction in the unsuccessful
sessions by using more intense language when they do secure
a chance to speak. It is also possible that both the level
of language intensity and interaction management are a
reaction by the husbands and wives to the importance of a
particular custody issue under discussion.

In the successful sessions the husbands' use of lower
levels of interaction management suggests that the husbands
are (a) being actively crowded out of the discussion or
passively left out as the mediators and the wives vie for
management of the interaction, (b) capitulating, or (c)
allowing the mediators to champion their preferred
agreement. The first alternative is an unlikely explanation
since Donohue and Weider-Hatfield (1986) found that nearly
equivalent amounts of talk were being used by all the
interactants across the mediation sessions.

One way to address the other two alternative
explanations is to investigate the sequential organization
of both the use of interaction management and the movement
toward agreement across the mediation sessions. The

movement toward agreement could be measured by first

 

73

tallying the number of issues Open at the outset of the
mediation and then tracking movement off the issues, the
narrowing of the differences, and the resolution of the
issues (Kochan & Jick, 1978). This would indicate what the
issues are in the sessions, who is involved in the
discussion of the issues, and the outcome of the issues. An
accompanying sequential analysis of the interaction
management tactics would indicate when certain types of
interaction management tactics are followed by other
tactics. By comparing the two analyses some judgment could
be made as to which issues are generating the most use of
interaction management along with the pattern of interaction
management being used and the final disposition of the
issues. If the husbands raise particular issues and they
are not actively considered as potential parts of the
custody agreement then the husbands' purposed custody
arrangements are being dropped from the discussion. If the
husbands raise particular issues that are then championed by
the mediators in opposition to the wives, then the other
alternative explanation is supported. More importantly, if
any of these alternative explanations are supported by later
research, then the mediators are not doing their jobs. An
agreement is reached in the successful sessions but not a
mutual agreement in which each interactant is actively
engaged in the process.

Based on Vanderkooi and Pearson's (1984) review of the

divorce mediation models Hypothesis Three predicted that the

 

 

74

level of interaction management would follow a Specific
pattern across a progression of stages; setting the stage
and defining the issues (less interaction management),
processing the issues (more interaction management) and
reaching (less interaction management) or not reaching (more
interaction management) an agreement. The results Showed an
increase in the level of interaction management from Phase
one to Phase two and from Phase two to Phase three in both
the successful and the unsuccessful sessions. The most
plausible explanation for this finding is that a phasic
model is not apprOpriate for the study of phenomena like
bargaining, negotiating and mediation.2
In selecting a phasic model for mediation, Vanderkooi
and Pearson (1983) write, "Although some writers
characterize mediation as involving cyclical or contingent
strategies, ... most writers view mediation as a progression
of stages ... This stage/phase approach has been adapted by
all the practitioners who have deve10ped and published model
approaches to divorce mediation." (p. 558). In reviewing
the work Vanderkooi and Pearson cite in support of a phasic
model (i.e., Black 8 Joffee, 1978; Coogler, 1978; Haynes,
1981; Kessler, 1978; Milne, 1978) and more recent work
(Blades, 1984) it was evident that the proposed models were
based on the writers' own mediation experiences rather than
on empirical tests. Combining the absence of empirical
support with the lack of empirical support found in this

dissertation, leads to the conclusion that mediation does

 

75

net follow a phasic model. The questions that still remain
are, 'Why?" and 'What model does characterize mediation?‘
One possible answer is contained in the statement that
begins Vanderkooi and Pearson's's (1983) review, "...some
writers characterize mediation as involving cyclical or
contingent strategies." (p. 558).

Empirical support is available for both a contingent and
a cyclical model of development in mediation. Kochan and
Jick (1978) developed and tested a model of the mediation
process using actual public sector labor mediations. Their
results support a contingent model of mediation. Putnam and
Jones (1982), using a grievance case role play, found that
sessions where no agreement was reached, as opposed to
sessions where agreement was reached, exhibited a tightly
structured, reciprocal pattern of the use of attack-attack
or defend-defend strategies. Putnam and Jones (1982)
concluded that the bargaining task (grievance case) was
particularly conducive to the creation of attack and defend
patterns. Based on her research and a review of the
bargaining literature, Putnam (1985) concluded that in small
groups the dominant interaction pattern is one of equivalent
symmetry whereas in negotiation the dominant interaction
pattern is one of competitive symmetry. The conclusion
that can be reached from these studies is that very little
conceptual effort has been devoted to determining what kinds
of communication acts prevail during specific time periods

in situations characterized by negotiation and bargaining

 

76

(Donohue, Diez, & Weider-Hatfield, 1985) and that research
on the interaction phases needs more scrutiny (Putnam,
1985).

Poole (1983ab; 1985) adds direction to the needed
conceptual and research effort by calling for studies of how
the issues and decisions involved in a task effect the
sequencing of interaction rather than more studies on the
phases groups must pass through to arrive at decision
points. Poole (1983b) views group development as a set of
parallel activity strands that evolve simultaneously and
interlock in different patterns over time. The activities
he proposes are task, relational, topical, control
strategies, and conflict management. These activities are
best studied by tracing each 'strand' over time and
observing the differing degrees of association and
coordination between the strands. It appears then that the
way to proceed is to begin where the small group decision
making theorists began.

For years small group researchers have attempted to
discover or create a classification scheme capable of
accounting for systematic differences in the communication
behaviors in groups (e.g., Bales 1951, 1979; Fisher 1980).
These research efforts began with the systematic observation
and classification of the communication acts that occurred
during small group meetings. Mediation researchers, then,
should begin where the small group researchers began;

observation and classification of the communication acts

 

_m
.m’fi' - -I l

 

 

77

that occur during mediation sessions.

The clearest practical implication of this research for
the practitioners of divorce mediation is that the mediators
who eventually use higher levels of interaction management
than the parents are able to structure the discussions in a
more productive direction. Two remaining questions are 'What
aspects of the discussion is interaction management used to
Structure?' and 'What are the best linguistic and
conversational forms for the mediators to use?"

The differences in the use of interaction management
when mediations result in agreement or disagreement, could
be due to the mediators' use of interaction management to
refocus the discussion on the issue Of custody or what is
best for the children. When the husbands and wives come to
mediation intent on pushing for their own disparate goals
they are more likely to use higher levels of interaction
management to try to gain a tactical advantage in managing
the interaction. If the mediators can gain control of the
interaction, they can refocus the discussion on an
overriding goal like reaching an agreement that is best for
the children. The mediators can devote their time to
gathering information, clarifying issues and other tasks
that help move the couple closer to agreement.

Although the mediators must rely heavily on interaction
management techniques, there are also pressures operating on
the mediators to keep the exchanges civil and provide an

example for the parents of COOperative language use (Blades,

m— 4': '-' £

 

 

78

1984). It may be that the practitioners, who are more
successful, then, are those who have learned to do things
like interrupting while using a directive form that signals
the right of the disputant to refuse; for example, "May I
interrupt here?" More work would need to be done, then, to
discover how practitioners can balance these two demands
successfully and the effects of this balance on the outcomes
of disputatious situations. Again, the further development
and use of an interaction management coding scheme that is
more sensitive to the occurrence of the linguistic and
conversational features within a stream of talk should
provide more useful results.

These results would also suggest that the difference
between successful and unsuccessful outcomes lies partly in
the amount of time the practitioners must spend in managing
the interaction as compared to time Spent on other tasks -
clarifying, gathering information - that help move the
disputants toward an agreement. More importantly, when the
mediators must spend an inordinate amount of time and energy
managing the interaction, an agreement is less likely to be
reached. Further investigation needs to be conducted, then,
into the use of particular behaviors by the disputants

across time in the interaction.

 

CHAPTER V

SUMMARY

The primary purpose of this research was to test certain
hypotheses about the interrelated effects of Weinstein's
Theory of Interpersonal Tactics and the strategy of
interaction management in disputatious social situations
like divorce mediation. These results provide support for
the interaction theory presented in this study and also have
practical implications for practitioners of divorce
mediation.

In support of Weinstein's ITT, the results of this study
indicate that when the participants pursue their goals of
reaching a favorable agreement in disputatious situations,
the strategy of interaction management is used to manage the
structure of the interaction so the discussion is supportive
of the participants' goals. In support of interaction
management, its use was found to play a primary but not
unequivocal role in disputatious social interactions like
divorce mediation. More Specifically, the overall use of
interaction management is not as important to the success of
the mediation sessions as the use of the interaction
management between the participants. Successful mediation

sessions are more likely when the mediators, as opposed to

79

 

80

the parents, eventually manage the interaction. The
mediators and the wives vie for management of the
interaction when the mediation sessions are successful and
all the participants vie for management of the interaction
when the mediation sessions are unsuccessful. The one clear
recommendation that can be made to mediation practitioners,
based on these results, is that mediators must manage the
interaction in order to provide the structure that moves the
parents toward agreement.

Since this is an initial investigation it is not
surprising that this research raises as many questions as it
answers. The most puzzling results of this dissertation are
that the husbands use lower levels of interaction management
in the successful as opposed to the unsuccessful mediation
sessions. One possible explanation for these results is
that in successful mediation sessions the husbands have
given up trying to manage the interaction because they have
problems participating or the discussion is favorable to
their position. Participation problems can be discounted as
an explanation since Donohue and Weider-Hatfield (1986)
found that the husbands, wives, and mediators all talked
equivalent amounts across the divorce mediation sessions.
Further work needs to be done to determine if the discussion
of the issues is favorable to the husbands' preferred
agreements}

The other equivocal result in this dissertation is that

the use of interaction management did not follow the phasic

 

81

model of development as predicted. It was suggested that a
phasic model might not be as appropriate for negotiating or
bargaining groups as it was for decision-making groups.
More work needs to be done at a descriptive level to
discover the sequential patterns of communication in
negotiation and bargaining contexts.

Future research on the strategic use of interaction
management in disputatious social interactions should
proceed in three clear directions. First, several
refinements are needed in the interaction management coding
scheme; closer attention to context and the scaling of
linguistic and conversational features within the stream of
talk. Second, several aspects of the issues that are raised
in the mediation sessions need to be investigated;
identification of the issues and whose agreement they
support, how the issues are dealt with and by whom, and the
eventual fate of the issues. This investigation should shed
light on the patterning of the levels of interaction
management used between the participants.

Lastly, work needs to begin on the observation and
classification of the communication acts that occur in

mediation sessions so that the interaction sequence can be

investigated.

 

FOOTNOTES

'1Weinstein's Theory of Interpersonal Tactics was
abbreviated as ITT because the use of the abbreviations
which result from the original title (TIT) produced more
titters and snickers than serious consideration of the
theory.

2There is a relationship between negotiation, bargaining,
and mediation such that the work in one of the areas can
inform on work in the other areas. According to Rubin and
Brown (1975) negotiating and bargaining are synonymous .
terms; process whereby two or more parties with divergent
aims, motives, or interests attempt to settle what each
shall give and take or perform and receive, in a transaction
between them. Divorce mediation, in turn, is a dispute
resolution process for parting spouses in which they
negotiate the terms of their divorce settlement with the
assistance of an impartial third party, the mediator
(Bishop, 1984).

3Research in both of these areas has already begun.
Stahle and Burrell (1987) have investigated the subtle ways
that the mediator can use legislative episodes to encourage
or discourage the consideration of an issue in divorce
mediation. Donohue and his colleagues are currently at work
analyzing the issues that occur in divorce mediation.

82

 

APPENDIX A

MEDIATION AGREEMENT 12/10/81

Robert and Joan Smith having separated for eighteen months

now wish to arrange their affairs and to file for divorce.

They therefore agree:

1.

That they shall continue to act together as parents to
their three children: Ann (17), Roger (16), and Susan
(15).
Both parents shall continue to be involved in all major
decisions regarding their children's lives and
wellbeing; these decisions shall include, but not be
limited to, decisions regarding health and dental care,
education, holidays and vacations.
The children shall continue to reside as they currently
live: the two girls with their mother and the son with
the father.
The children shall continue to visit freely with both
parents at their, the children's discretion. The
children, being of sufficient age, will be responsible
for letting their parents know where they are.
Both parents will continue to carry the children on
their work related health insurance policies.
Both parents will be pay equally for the children's

83

 

 

 

 

APPENDIX B

CODING INSTRUCTIONS

COLUMN 1 Enter the number of the transcript on the first
line of every page. It's not necessary to repeat
it for every line on the same page.

COLUMN 2 Enter the utterance number, even if there is no
interaction management tactic used in a given
utterance.

001 M: I'll begin by explaining why you're here

COLUMNS 3,11,21,30 At the beginning of each major group of
language characteristics is a column marked A/R. For

each item that is coded, record the number that
corresponds with the speaker to addressee/referent of

the item as follows:

to/about
to/about
to/about
tO/about
to/about
to/about

O‘U‘bMNl—i
22:13:12:
2:321:22:

You may need to use several lines - one per
addressee or referent - for each utterance you code.

For example:
002 H: Shelly, 1 think Mr. Jones can

You would need one line to record the wife's name
(Shelly) and the (I think) directed to her

and another to record the reference to the mediator
(Mr. Jones).

OTHER COLUMNS Under each of the major language
characteristics, the specific characteristics
to be coded are abbreviated and listed
following the order in the coding manual.

For each utterance record the number of times

85

 

86

a given characteristic occurred in addressing
or in reference to one of the parties in the
interaction.
Select the characteristics to be coded in the
transcript. TREAT EACH CHARACTERISTIC AS A
SEPARATE CHARACTERISTIC TO BE CODED.
A. For example,
003 H: Excuse he, 1 think I can handle this.
(Excuse me) would be coded as one characteristic
(1 think) would be coded as one characteristic
B. For example,
004 M: Let me just interject here
(Just) would be coded as a separate
characteristic. AND

The whole sentence would be coded as a separate
characteristic.

Some of the conversational organizing devices are
Interruptions and gaining the floor. They are
indicated by lines of utterances enclosed in a
bracket. TREAT EACH BRACKET AS A SEPARATE UNIT TO
BE CODED SKIPPING THE FIRST LINE IN THE BRACKET.

005 M: All right now What kind of um, controls
would you uh put on it what uh

006 W ecified hours and specified days

:((PAUSE)) and a telephone call to

007 M: How frequent would that be

008 W: confirm

009 M: What days what hours

BEFORE YOU BEGIN TO CODE: Read through the transcript and

become familiar with the (H)usband, (W)ife, and (M)ediator
and their case. This is the only way you'll be able to
decide if a reference to (George) is the H's name and gets

recorded as such or the M's name.

 

 

BIBLIOGRAPHY

Adler, R. B., 6 Towne, N. (1987). Looking eeh looking he;
Interpersonal communication. New York: Holt, Rinehart,
and Winston.

Applegate, J., G Leichty, G. (1985). Managing interpersonal
relationships: Social cognitive and strategic
determinants of competence. In R. Bostrom (Ed.),
Communicatieh yearbook, yehh 1 (pp. 33-55).

Beverly Hills, CA: Sage.

Argyle, M. (1969). Social interactieh. New York:
Atherton Press.

Bales, R. F. (1951). Interactieh process analysis.
Cambridge, MA: Addison-Wesley.

Bales, R. F. 5 Cohen, S. P. (1979). SYMLOG. New York:
Free Press.

Barsky, M. (1983). Emotional needs and dysfunctional
communication as blocks to mediation. Mediation
Quarterly, ;, 55-66.

Bell, R. A., & n51y, J. A. (1984). The affinity-seeking
function of communication. Communication Monographs,

£1, 91-115.

87

 

   

r ./ '
s _\7 . ; __\
1...,” r _ . . x- . " 7.

88
'Berger, C. R. (1980). Power and the family. In M. E.
Roloff G G. R. Miller (Eds.), Persuasion: New directions
he theogy and research (pp. 197-224). Beverly Hills,
CA: Sage.

Berger, C. R. (1986). Social power and interpersonal
communication. In M. L. Knapp 8 G. R. Miller (Eds.),
Handbook eh interpersonal communication (pp. 439-499).
Beverly Hills, CA: Sage.

Bishop, T. A. (1984). Family and divorce mediation issue.
Dispute resolution forum. Washington, DC: National
Institute for Dispute Resolution.

Black, M., G Joffe, W. (1978). A lawyer/therapist team
approach to divorce. Conciliation Courts Review, he,
1-5.

Blades, J. (1984). Mediation: An old art revitalized.
Mediation Quarterly, e, 59-98.

Brown, P., 8 Fraser, C. (1979). Speech as a marker of
situation. In K. Scherer 5 H. Giles (Eds.), Social
markers he Speech (pp. 33-62). Cambridge, MA:
Cambridge University Press.

Brown, P., & Levinson, S. (1978). Universals in language

usage: Politeness phenomena. In E. Goody (Ed.), Questions

 

and politeness: Strategies he social interaction

 

(pp. 56-288). Cambridge, MA: Cambridge University Press.

Brown, R., 8 Gilman, A. (1960). The pronouns of power and
solidarity. In T. A. Sebeok (Ed.), §hyhe he language
(pp. 253-276). Cambridge, MA: Free Press.

 

 

89

Bruner, J. S. (1979). Smiles can be back channels. Journal
eh Personality eee Social Psychology, e1, 728-734.

Bryant, S. E Patton, B. R. (1987). Communication strategies
and divorce mediation. In L. P. Stewart 8 S. Ting-Toomey
interaction eeetexts (pp. 49-61). Norwood, NJ: Ablex.

Carrell, P. L. 5 Konneker, B. H. (1981). Politeness:
Comparing native and nonnative judgments. Language
Learning, e, 17-31.

Clark, R. A., G Delia, J. G. (1979). Topoi and rhetorical
competence. Quarterly Journal eh Speech, eh, 187-206.
Cohen, J., 6 Cohen, P. (1975). Applied eeltiple regressieeh
Correlation eeelysis heh hhe behavioral sciences.

Hillsdale, NJ: Lawrence Erlbaum.

Coogler, O. J. (1978). Structured hediation he divorce
settlement. Lexington, MA: Lexington Books.

DeVito, J. (1987). heeeamentals eh interpersqeeh
communication. New York: Random House.

Dibble, C. (1984). Bargaining in family mediation: Ethical
considerations. Mediation Quarterly, 1, 75-92.

Diez, M. (1987). A conceptualization of the rules of
negotiation interaction. In D. Ellis, 5 W. A. Donohue
(Eds.), Contemporary issues he language discourse
processes (pp. 223-238). Hillsdale, NJ: Lawrence

Erlbaum.

 

   

 

90

Donohue, W. A., Allen, M., 8 Burrell, N. (1985).
Communication strategies in mediation. Mediation
Quarterly, he, 75-89.

Donohue, W. A., G Weider-Hatfield, D. (1986). Communication
strategies in mediation. In A. Milne and J. Folberg
(Eds.), h mediation handbook heh practitioners
(pp. 147-198).

Donohue, W. A., Diez, M., G Weider-Hatfield, D. (1985).
Skills for successful bargainers: A valence theory of
competent mediation. In R. N. Bostrom (Ed.),

Competence he communication (pp. 219-258). Beverly
Hills, CA: Sage.

Duncan, S. (1983). Speaking turns: Studies of structure and
individual differences. In J. M. Weimann & R. P.
Harrison (Eds.), heeverbal interaction (pp. 149 -l78).
Beverly Hills, CA: Sage.

Duncan, S. 5 Fiske, D. W. (1977). Face-to-face interaction.
Hillsdale, NJ: Erlbaum.

Eakins, B. W., G Eakins, G. E. (1978). §ee differences he
heeee communication. Boston: Houghton Mifflin.

Erwin-Tripp, S. (1964). An analysis of the interaction of
language, topic and listener. American Anthropologist,
ee, 86-94.

Erwin-Tripp, S. (1968). Sociolinguistics. In L. Berkowitz
(Ed.), Advances he experieeeheh social psychology, Xeh A
(pp. 91-165). New York: Academic Press.

 

91

Erwin-Tripp, S. (1972). On sociolinguistic rules:
Alternation and co-occurrence. In J. J. Gumperz G D,
Hymes (Eds.), Directions he sociolinguistics,

(pp. 213-250). New York: Holt, Rinehart a Winston.

Erwin-Tripp, S. (1978). Children's verbal turn-taking. In
E. Ochs G B. Schieffelin (Eds.), Developmental pragmatics
(pp. 391-412). New York: Academic Press.

Fisher, B. A. (1980). §hehh gheep decision-making. New
York: McGraw-Hill.

Fisher, R. J. (1969). Intereetional eeeflict heh beginner .
New York: Harper 6 Row.

Fisher, R. J. (1983). Third party consultation revisited as
a method of intergroup conflict resolution: A review of
studies. Journal eh Conflict Resolution, £1, 301-334;

Fishman, P. M. (1978). Interaction: The work women do.
Social Problems, he, 397-406.

Fishman, P. M. (1980). Conversational insecurity. In H.
Giles, W., P. Robinson, a P. Smith (Eds.), Language:
Social psychological perspectives (PP. 127-142).

New York: Pergamon Press.

Folger, J., Hewes, D., & Poole, S. I. (1984). Coding
social interaction. In B. Dervin & M. Voight (Eds.),
Progress he communication sciences (pp. 115-161).
Norwood, NJ: Ablex.

Giles, H., E St. Clair, W. (1986). Recent advances he
language, communication eee social psychology. London:

Lawrence Erlbaum.

 

92

Geffman, E. (1959). The presentation _h self he everyday

 

hhhe. Garden City, NY: Anchor Books.

Goffman, E. (1963). Behavior he public places: hehee ee
hhe social organizatiee eh gatherings. New York:
Free Press.

Goffman, E. (1969). Strategic interactiqe. Philadelphia,
MA: University of Philadelphia Press.

Goode, W. J. (1960a). A theory of role strain. American
Sociological Review, he, 483-496.

Goode, W. J. (1960b). Norm commitment and conformity to
role-status obligations. heerican Journal eh Sociology,
6_6_, 246-258.

Goodwin, C. (1981). Conversational organization: Interaction
between speakers eee hearers. New York: Academic Press.

Grimshaw, A. D. (1980). Selection and labeling of
instrumentalities of verbal manipulation. Discourse
Processes, e, 203-229.

Grimshaw, A. D. (1981). Talk and social control. In M.
Rosenberg G R. H. Turner (Eds.), Social psychology:
Sociological perspectives (pp. 200-232). New York: Basic
Books.

Guetzkow, H. (1950). Unitizing and categorizing problems in
coding qualitative data. Journal eh Clinical Psychology,
e, 47-58.

Gulliver, P. H. (1979). Disputes eee negotiations: A

cross cultural perspective. New York: Academic Press.

 

93

GumPerz, J. J. (1982). Discourse strategies. Cambridge, MA:
Cambridge University Press.

Hare, P. (1973). Theories of small group deve10pment and
categories for interaction analysis. eeehh £5222
Behavior, 1: 259-304.

Haynes, J. M. (1981). Divorce mediation: A practical gehee
heh therapists eee counselors. New York: Springer.

Hewes, D. (1985). Systematic biases in coded social
interaction data. heeee Communication ReSearch, hh,
554-575.

Hirschman, L. (1973). Analysis eh supportive eee assertive
behavior he conversatieee. Paper presented at the
meeting of the Linguistic Society of America, San
Francisco.

Hocker, J. G Wilmot, W. W. (1985). Interpersonal conflict.
Dubuque, Iowa: Wm C. Brown.

Homans, G. (1961). Social behavior: hhe elementary hehee.
New York: Harcourt, Brace and World.

James, S. (1978). Effect of listener age and situation on
the politeness of children's directives. Journal eh
Psycholieguistic Research, 1, 307-317.

Kelly, J. (1984). Mediation and psychotherapy:
Distinguishing the differences. hediatiqe Qearterly, h,
33-45.

Kessler, S. (1978). Counselor as mediator. Personnel eee

Guidance JOurnal, pg, 91-106.

 

94

Kochan, T. A., G Jick, T. (1978). The public sector
mediation process. Journal eh Conflict Resolutiee, hh,
209-240.

Kolb, D. (1983). Strategy and the tactics of mediation.
heeee Relations, he, 247-268.

KOper, R. J., Donohue, W. A., G Stahle, R. (1985). Language
intensity he divorce mediatiee. Paper presented at the
Speech Communication Association meeting, Denver.

Kramarae, C. (1981). heeee eee eee Speaking: Frameworks
heh analysi . Rowley, MA: Newbury House.

Kressel, K. (1971). Labor hediation: he exploratopy survey.
Albany, NY: Association of Labor Mediation Agencies.

Lewis, S. A. 6 Fry, W. R. (1977). Effects of visual access
an orientation on the discovery of integrative bargaining
alternatives. Organizational Behavior eee heeee
Performance, he, 75-92.

Mead, G. (1934). hheeh eehhh eee society. Chicago, IL:
University of Chicago Press.

Milne, A. (1978). Custody of children in a divorce process:
A family self-determination model. Conciliatiee Courts
Review, he, 1-10.

Moore, C. W. (1986). The eediation process: Practical

 

Strategies for resolving conflict. San Francisco, CA:

Jossey-Bass.

 

95

Morris, C. W. (1949). Individual differences and cultural
patterns. In Conference ee science, philosophy eee
religious conflict eh peeeh he modern culture,

(pp. 40-50). New York: Atherton Press.

Orcutt, J. D. 8 Harvey, L. K. (1985). Deviance, rule-
breaking and male dominance in conversation. Symbolic
Interactionism, h, 15-32.

Pearson, J., 8 Thoennes, N. (1984). The preliminary portrait
of client reactions to three court mediation programs.
Mediation Qearterly, e, 21-40.

Pearson, J. C. (1986). Communication between men and women.
interpersonal communication (pp. 281-306). New York:
Random House.

Phear, W. P. (1981) Involving children within the divorce
mediation process. Conciliatiee Courts Review, he, 14-22.

Poole, M. S. (1983a). Decision development in small groups,
II: A study of multiple sequences in group development.
Communication Monographs, he, 206-32.

Poole, M. S. (1983b). Decision development in small groups,
111: A multiple sequence model of group decision
deve10pment. Communication Monographs, he, 321-341.

Poole, M. S. (1985). Tasks and interaction sequences: A
theory of coherence in group decision-making interaction.
In R. Street 8 J. Cappella (Eds.), Seguence eee
pattern he eeeeenicative hehavior (pp. 206-224).
Baltimore, MY: Edward Arnold.

 

96

Pruitt, D. (1981). Negotiation heheyior. New York: Academic
Press.

Pruitt, D., E Kressel, K. (1985). The mediation of social
conflict: An introduction. Journal eh Social Issues, eh,
1985, 1-10.

Putnam, L. (1985). Bargaining as task and process:

Multiple functions of interaction sequences. In R.

Street E J. Cappella (Eds.), Sequence and pattern he

 

heeeenicative hehavior (pp. 225-242). Baltimore, ML:
Edward Arnold.

Putnam, L. G Geist, P. (1985). Argument in bargaining: An
analysis of the reasoning process. hhe Southern Speech

Communication Journal, he, 225-245.

 

Putnam, L. 8 Jones, T. (1982). The role of communication in

bargaining. Human Communication Research, e, 262-280.

 

Ragan, S. (1983). Alignment and conversational coherence.
In R. Craig 8 K. Tracy (Eds.), Conversational coherence:
Eeheh structure eee strategy (pp. 157-173). Beverly
Hills, CA: Sage.

Rosanova, M. (1983). Mediation: Professional dynamics.
Mediation Quarterly, h, 63-74.

Rubin, J. Z. (1980). Experimental research on third-party
intervention in conflict: Toward some generalizations.
Psychological Bulletin, eh, 379-391.

Rubin, J. 2. G Brown, B. R. (1975). The social psychology

of bargaining and negotiation. New York: Academic Press.

 

97

Saposnek, D. (1983). Strategies in child custody mediation:
A family systems approach. hedhetion Qearterly, h,
29-54.

Saposnek, D. (1985). hediating child custody disputes. San
Francisco, CA: Jossey-Bass.

Schegloff, E. (1972). Notes on conversational practice:

Formulating place. In D. Sudnow (Ed.), Studies he social

 

interaction (pp. 26-41). New York: Free Press.
Schelling, T. C. (1960). hhe strategy eh conflict. New
Harvard University Press.
Schenkein, J. (1978). Studies h_ hhe organization _h

conversational interactiee. New York: Academic Press.

 

Shimanoff, S. (1983). Communication rules: Theory

 

 

eee research. Beverly Hills, CA: Sage.

Smith, P. M. (1985). Language, hhe eeeee eee society.
New York: Basil Blackwell, Inc.

Spitzberg, B. H. G Cupach, W. R. (1984). Interpersonal
communication competence. Beverly Hills, CA: Sage.

Stahle, R. (1985). hhe impact eh hhe strategic pee eh
mediation. Unpublished manuscript. Michigan State
University, Department of Communication, East Lansing.

Stahle, R. 8 Burrell, N. (1987). Argumentation in divorce
mediation: What counts as decisional premises in child
custody decisions?. In E. Wenzsel (Ed.) Proceedings
eh hhe heehh annual ehhe conference ee argumentation,

(pp. 178-198).

   

 

98

Street, R. L., a Cappella, J. N. (1985). Seguence and
pattern he communicative behavior. Baltimore, ML:

Edward Arnold.

 

Street, R. L., G Giles, H. (1983). Speech accommodation
theory: A social cognitive approach to language and
speech behavior. In Roloff, M. E. G C. R. Berger,
(Eds.), Social cognition eee communication
(pp. 199 - 226). Beverly Hills, CA: Sage.

Treichler, P. A., Frankel, P. M., Kramarae, C., Zoppi, K. G
Beckman, H. B. (1984). Problems and problems: Power
relations in a medical encounter. In C. Kramarae, M.
Schultz, G W. O'Barr (Eds.), Language eee peeeh, (pp. 62-
88). Beverly Hills, CA: Sage.

Vanderkooi, L., 8 Pearson, J. (1983). Mediating divorce
disputes: Mediator behaviors, styles and roles. Family
Relations, eh, 557-566.

Wall, J. (1981). Mediation: An analysis, review and
proposed research. Journal eh Conflict Resolution, 3;,
157-180.

Weiler, J., 6 Weinstein, E. (1972). Honesty, fabrication,
and the enhancement of credibility. Sociometry, §§,
316-313.

Weimann, J. (1985). Interpersonal control and regulation of
conversation. In R. Street 8 J. Cappella (Eds.),
Seguence eee pattern he eeeeeeheetive hehavior
(pp. 85-102). Baltimore, MY: Edward Arnold.

 

   

 

99

Weimann, J. M., 8 Kelly, C. W. (1981). Pragmatics of
interpersonal competence. In C. Wilder-Mott & J. H.
Weakland (Eds.), hhgeh eee imagination: Essays hhee hhe
legacy eh Gregory Bateson (pp. 283-297). New York:
Praeger.

Weinstein, E. (1966). Toward a theory of interpersonal
tactics. In C. W. Backman 8 P. F. Secord (Eds.),
Problems he social psychology (pp. 394-398). New York:
McGraw-Hill. ‘

Weinstein, B., & Deutschberger, P. (1963). Some dimensions
of altercasting. Sociometry, A, 454-466.

Wilmot, W. W. (1986). Dyadic communication. New York:
Random House.

Young, 0. (1972). Intermediaries: Additional thoughts on
third parties. Journal eh Conflict Resolution, he,
51-65.

Yngve, V. H. (1970). On getting a word in edgewise. In
M. A. Campbell et al. (Eds.) Papers hhee hhe ehehh
regional meeting (pp. 567-578). Chicago, IL: Chicago
Linguistics Society a Department of Linguistics,
University of Chicago.

Zimmerman, D. 8 West, C. (1975). Sex roles, interruptions
and silences in conversation. In B. Thorne & N. Henley
(Eds.), Language and sex: Difference and dominance,

(pp. 105-129). Rowley, MA: Newbury House.

0

 

 

 

 

11

 

 

 

 

 

 

N STQTE

U

29300

V

nrcwrcn
WWI
31

mm