. HUIWWWHIWUIHIIIHIIIIHII"INHIIUIHIIUII

G...

,,, . 0 302043 8718

* LIBRARY

Michigan State
University

 

 

 

 

This is to certify that the

dissertation entitled

COMMUNICATION IN JAPANESE MULTINATIONAL
ORGANIZATIONS IN THE U.S.:
CONVERGENCY OF FRAMES AND OUTGROUP COMMUNICATION

presented by
Toru Kiyomiya

has been accepted towards fulﬁllment
of the requirements for

Ph . D. degree in Communication

 

m
, Major érofessor a

DateAugusiJiJﬂL

MSU is an Afﬁrmative Action/Equal Opportunity Institution 0-12771

 

 

_ w 57" ve—cr'

PLACE IN RETURN BOX to remove this checkout from your record.
TO AVOID FINES return on or before date due.
MAY BE RECALLED with earlier due date if requested.

 

DATE DUE

DATE DUE

DATE DUE

 

 

 

 

 

 

 

 

 

 

 

 

 

 

11/00 CICIRCIDatoDuopBS-p.“

 

COMMUNICATION IN
JAPANESE MULTINATIONAL ORGANIZATIONS IN THE U. s.:
CONVERGENCY OF FRAMES AND OUTGROUP COMMUNICATION
By

Tom Kiyomiya

A DISSERTATION
Submitted to
Michigan State University

in partial fulﬁllment of the requirements
for the degree of

DOCTOR OF PHILOSOPHY

Department of Communication

2000

ABSTRACT
COMMUNICATION IN
JAPANESE MULTINATIONAL ORGANIZATIONS IN THE U. S.:
CONVERGENCY OF FRAMES AND OUTGROUP COMMUNICATION
By
Toru Kiyomiya
In this study, the national cultures and the complex contexts of Japanese multinational
organizations (MNOS) are studied in the US. The present study assumes that there are
inﬂuences of national cultures which maintain signiﬁcant cultural stereotypes in MNO,
and also there are complex contexts of MNOs which might create distinctive
communication patterns among Americans and Japanese. The objectives of this study are
to quantitatively analyze the cognitive and behavioral dimensions of communication
between these two cultural groups and identify distinctive patterns of organizational
communication in terms of MNO effectiveness. The frameworks for interactions (F INT)
scale is used for measurement of individual communication frames (n=152), and ego-
networks are measured to analyze ingroup-outgroup communications (n=66). Cultural
difference is identiﬁed in communication frames and patterns of ingroup-outgroup
communication. Results about communication frames and ingroup-outgroup
communication contradict expectations of cross-cultural stereotypes. This result implies
that convergence in communication frames and communicative interactions is a key to
success in MNOs. Results also suggest that the importance of outgroup communication
is in its boundary spanning function, while increasing outgroup communication is crucial
for both Americans and Japanese to enhance their communicative convergence and

increase the effectiveness of MNOs.

Dedicated to my wife, Tomoe
and

my parents, Tsutako and Masami Kiyomiya

iii

ACKNOWLEDGMENT

I would like to ﬁrst thank all the members of my doctoral committee.
Particularly, I would like to extend special gratitude to two chair persons: Dr. J. David
Johnson and Dr. James W. Dearing. Dr. Johnson is currently Professor and Dean of the
College of Communications and Information Studies at the University of Kentucky, and
he is an original chair of my Ph.D. Guidance Committee. Dave provided me with a
number of insights and suggestions while I was a research assistant in his research team
and a teaching assistant in his classes. He has continuously helped me in my doctoral
study after he left MSU. Dr. Dearing is a current chair of my Ph.D. Committee. Jim has
generously taken care of the entire process of my doctoral study and guided me in the
right direction. I could never have completed my degree without the feedback and
suggestions from these two chairs. I would also like to thank Dr. Mary Bresnahan, Dr.
William Donohue, and Dr. Elaine Yakura. I learned a lot from them in and out of their
classes. Their warmhearted support and encouragement have been critical to my PhD.
work and my personal life.

Next, I would like to thank Dr. Linda Kay Dabbs and communication faculty at
the University of Tulsa. Linda provided me with a number of suggestions and comments
on the initial and ﬁnal draft, and together they have improved my doctoral dissertation.

I would like to thank Dr. Rie Ohashi and Mr. Yasuhiro Inoue. They helped the
process of back-translation and provided me with a number of valuable suggestions.
Thanks also go to all the respondents who participated in my research and the Japanese

Business Society of Detroit. I could not have conducted my research without their

cooperation. My appreciation goes to the National Communication Association for the
copyright permission to reproduce important information.

Additionally, I would like to thank the faculty members, the staffs, and my
colleagues in Department of Communication at MSU: Ms. Barkman, Ms. Gladstone, Ms.
Ti gner, Ms. Wooten, Dr. Atkin, Dr. Boster, Dr. MaCornack, Dr. Witte, and Ms.
Morinaga. Dr. Miller and Dr. Tarnborini have guided me from the ﬁrst year of my
doctoral study. Also, I would like to express my appreciation to my ﬁ'iends and people
whose names I could not acknowledge here. I will always remember their kindness and
assistance.

Lastly, I owe thanks to my parents and my wife, for supporting my oversea study
and the completion of my Ph.D. In particular, Tomoe helped my research in data entry,
and she has always supported and encouraged me when I faced difﬁculty. Without her

my work would never have been accomplished.

PREFACE

More than ﬁve decades ago, the tragedy of World War II engulfed the United
States and Japan. Amazingly, both countries successfully overcame their losses. In
many respects, they have built great international partnerships. In business, Americans
and Japanese work collaboratively and competitively, pushing each other to innovate and
excel. I anticipate that further globalization will occur in the international economy, and
that internationalization will escalate through education and communication technologies.
The strong partnership between these two nations will become a model of cooperative
relations in global economy. Through my dissertation I intend to contribute to such

cooperative efforts among nations.

vi

TABLE OF CONTENTS

LIST OF TABLES
LIST OF FIGURES

INTRODUCTION
Globalization and Japanese Multinational Organizations in the US.

CHAPTER]: REVIEW OF THE LITERLATURE AND PERSPECTIVES ABOUT
MULTINATIONAL ORGANIZATIONS

Studies of Multinational Organizations
Growing Literatures of Multinational Organizations
Intercultural Organizational Communication
Japanese Multinational Organizations
Considering Units of Analysis
Interpersonal Context of Intercultural Organizational Communication
Cultural Variability in American and Japanese Communication
Ingroup and Outgroup Communication
Organizational Context of Intercultural Organizational Communication
Organizational Behavior in Cross-Cultural Management
Ouchi’s Organizational Theory
Diversity Management Perspective
Multicultural Organizations and Multiculturalism
Communication Accommodation Perspective
Framing Perspectives as Applied to MNOs
Frameworks for Interaction
Exchange
Formal
Normative
Sentiments
Negotiated Order
Research Questions and Hypotheses
Operationalization
Research Questions and Hypotheses

vii

CHAPTER 2: METHODS 43

Measurement 43
Framework for Interaction (FINT) Scale 43
Social Network Questionnaire 45
Demographic Questions 48
Research Design 48
Data Collection 52
Sampling 52
Procedure of Data Collection 53
CHAPTER 3: RESULTS 55
Description of Sample 55
Psychometric Properties of FINT Scale 56
Results of Phase One: Cultural Difference and F INT Subscales 59
Results of Phase Two: Cultural Difference and Ego-networks to Ingroup and
Outgroup 61
Ego-networks in J ob-oriented Communication 63
Ego-networks in Organization-related Communication 66
Ego-networks in Social Communication 68
Ego-networks in Overall Communication 69
Results of Phase Three: Divergence and Convergence in Communication of
MNOs 71
Results of Phase Four: Total F INT Score and Ingroup-Outgroup
Communication 74
CHAPTER 4: DISCUSSIONS 76
Implications in Cultural Difference and Communication Framework 76
Critiques of Cross-cultural Stereotypes 76
Hybrid Pragmatism in MNOs 78
Implications in Cultural Difference and Ingroup-Outgroup Communication 80
Implication of American Communication Patterns 80
Implication of Japanese Communication Patterns 82
Implications in Convergence of Japanese MNOs in the US. 85
Convergence in Cognitive and Behavioral Levels 85
Contextual Models of MNOs 86
Implications in Communication Framework and Outgroup Communication 91
Limitations and Future Studies 93
FOOTNOTE 98
REFERENCES 100
APPENDICES 107
Appendix A — Tables 107
Appendix B — Figures 123

Appendix C — Questionnaire 131

viii

LIST OF TABLES

TABLE 1: Japanese and Western Organizational Behavior and Protocol
TABLE 2: Description of Research Sites
TABLE 3: Demographic Information on Respondents for Each Company

TABLE 4: Exploratory Factor Analysis for FINT Items: Rotated
Component Matrix

TABLE 5: Correlation Matrix of FINT Items

TABLE 6: Items of Frame for Interaction (FINT) Scale: Reliability and
Factor Loading

TABLE 7: Means, Standard Deviations, and T-tests: FINT Subscales in
Americans and Japanese

TABLE 8: MANOVA Outcomes: Cultural Difference on Five Dimensions of
FINT Scale

TABLE 9: Means, Standard Deviations, and t-Tests: Ego-networks in
Americans and Japanese

TABLE 10: MANOVA Outcomes: Cultural Difference on the Number of
Ego-networks

TABLE 11: MANOVA Outcomes: Cultural Difference on the Percent of
Ingroup Communication

TABLE 12: MANOVA Outcomes: Cultural Difference on the Percent of
Outgroup Communication

TABLE 13: Correlation of FINT and Ingroup-Outgroup Communication
(Americans: n = 41)

TABLE 14: Correlation of FINT and Ingroup-Outgroup Communication
(Japanese: n = 25)

TABLE 15: Contextual Models of Multinational Organizations

ix

108

109

110

111

112

113

114

115

116

117

118

119

120

121

122

FIGURE 1:

FIGURE 2:

FIGURE 3:

FIGURE 4:

FIGURE 5:

FIGURE 6:

FIGURE 7:

LIST OF FIGURES

Means of Five F INT Dimensions for Americans and Japanese 124
Histogram: Number of Ego-networks in Job-oriented Communication 125
Histogram: Number of Ego-networks in Organization-related
Communication 126
Histogram: Number of Ego-networks in Social Communication 127
Histogram: Ingroup-Outgroup Communication in Job-oriented
Communication 128
Histogram: Ingroup-Outgroup Communication in Organization-related
Communication 1 29
Histogram: Ingroup-Outgroup Communication in Social

Communication 130

INTRODUCTION
Globalization and Japanese Multinational Organizations in the U.S.

The Ministry of International Trade and Industry of Japan (MITI, 1998) recently
summarized the development of globalization and Japanese overseas business. Japanese
industries mainly invested in developing countries in the 1970’s, and many Japanese
companies established and started operations in the United States and European countries
in the 1980’s. Globalization of Japanese industry rapidly escalated in the late 1980’s
because of competition in the international market and international trade policies.

In this economic and political environment, the number of Japanese companies
has grown in the U.S. For example, over two hundred companies were established in the
U.S. in every ﬁve-year period between 1970 and 1985 (Nichibei Hikaku Kenkyuukai,
1989). The total number of Japanese companies in the US was 995 in 1985 and 3,721 in
1994 (MS Research, 1999). The late 1980’s are considered a peak of globalization in
Japanese industry. Currently, the level of globalization is relatively stable; the number of
Japanese companies in the U.S. is 3,779 in 1998.

As the result of globalization, workplaces are more diversiﬁed in the U.S. In
1992, approximately 600,000 Americans worked at J apanese-owned companies in the
U.S. (Morrow, 1992). Japanese and American employees work together and
communicate with each other within these organizations. These are multinational
organizations (MNOs), “an ascendant organizational form in today’s global business
environment” (Teboul, Chen, & Fritz, 1994, p. 12). Organizational communication in
MNOS is more complex than that in homogeneous organizations (Shuter & Wiseman,

1994), and cultural diversity is considered potentially problematic. Heterogeneous

workplaces are more risky than homogeneous ones because differences oﬁen become
sources of conﬂicts, misunderstanding, and poor performance (Salk & Brannen, 2000).
Therefore, in order to create effective MNOS and improve business practices, it is critical
to consider how these MNOS overcome cross-cultural gaps between Americans and
Japanese and how these different national cultures inﬂuence organizational
communication within MNOS.

Shuter and Wiseman (1994) suggest an important research question for these
regards issuesi; “what is the linkage between national culture, corporate culture, and
organizational communication?” (p. 8). They assume that there are two major contexts,
national culture and corporate culture. National culture may have signiﬁcant inﬂuence
on communication for Japanese and Americans and they may maintain their cultural
traditions in their communication. On the other hand, there might be a complex context
of MNOS which may create unique communication patterns among Americans and
Japanese, and these patterns may be different from the contexts in which Americans work
in American companies in the U.S. and in which Japanese work in Japanese companies in
Japan. Thus, the objectives of the present dissertation study are to (1) quantitatively
analyze the cognitive and behavioral dimensions of communication between people of
these two cultures in Japanese MNOS, and (2) identify distinctive patterns of
organizational communication in terms of the effectiveness of MNOS. Toward this end,
individual communication frames are measured to analyze the cognitive aspect of
communication, and ingroup and outgroup communication is measured to analyze the
behavioral aspect of communication. Thus, the unique contexts of MNOS as well as the

inﬂuence of national cultures are studied in Japanese MNOS in the U.S.

In Chapter 1, I introduce the perspective of intercultural organizational
communication and then illustrate traditional approaches to cross-cultural studies
between the U.S. and Japan, which are typically studied in the ﬁelds of intercultural
communication and cross-cultural management. These sections conﬁrm the norms or the
expectations of cross-cultural stereotypes as well as cultural stereotypes of ingroup-
outgroup communications between and among Americans and Japanese. In order to
examine the cognitive level of communication, I illustrate communication frameworks,
and the ﬁve dimensions of the frameworks for interactions (FINT) scale (Johnson, 1997a)
are introduced as a measurement instrument. In addition, on the basis of the studies of
cultural diversity and communication accommodation perspectives, I argue that the
concept of convergence-divergence is a key indicator for the complex context of MNOS.
At the end of Chapter 1, I articulate research questions and hypotheses concerning ﬁve
major concepts: cultural difference, communication frames, ingroup-outgroup
communications, the norms of cross-cultural stereotypes, and convergence-divergence in
communication.

In Chapter 2, research design and methods are described. Development of a F INT
questionnaire, a social network questionnaire, sampling, data collection procedures, and
research sites of Japanese MNOS are discussed.

In Chapter 3, results are presented for each research question and hypothesis. T-
tests and MANOVA measure the inﬂuence of cultural difference on FINT and ingroup-
outgroup communication. In order to identify distinctive communication patterns in

MNOS, convergence and divergence is examined on the basis of these results.

In Chapter 4, the outcomes of quantitative analyses lead to the implications of
cultural inﬂuence and effectiveness of MN Os in terms of communication frames and
ingroup-outgroup communication. Distinctive organizational communication in Japanese
MNOS is discussed particularly in terms of the beneﬁts of convergent communication.
On the basis of the conceptualization of convergence-divergence, I discuss seven types of
contextual models of MNOS. Finally, the limitations of this study are clariﬁed and

directions for future research are considered.

CHAPTER 1:
REVIEW OF THE LITERATURE AND PERSPECTIVES ABOUT

MULTINATIONAL ORGANIZATIONS

Studies of Multinational Organizations
Growingéiteratures of Multinational Organizations

Globalization in the late 1980’s directed scholars’ attention to international and
cross-cultural perspectives about organizations. A tremendous amount of literature about
this topic has been produced in the past two decades. While many studies focus on cross-
cultural management, cross-cultural negotiation, leadership, and other interpersonal
relationships of intercultural communication, the study of MNOS has received less
attention, as Shuter and Wiseman (1994) point out. Particularly, the communication
domain in MNOS is little studied: “Although the body of literatures of MNOS is growing,
little of this research has focused on commmiication or, especially, intercultural
organizational communication” (Shuter and Wiseman, 1994, p. 3).

Teboul, Chen, and Fritz, (1994) suggest that MNOS have received a good deal of
attention, but mainly from the ﬁelds of economics and business management. From 1978
to 1991, Communication Abstracts reviewed only eleven articles on “multinational
corporation, three on “multinational enterprise,” ten on “transnational enterprise,” and
eight on “transnational corporations” (Teboul et al., 1994). According to Teboul, et al.,
“One explanation for this lack of research is that MNOS may be more difﬁcult to access”

(p. 15) than less complex organizations.

Interculturzﬂ Organizational Communication

Shuter (1985) initially advocated the study of “intercultural organizational
communication” (IOC), which pays attention to a different dimension from traditional
intercultural communication studies. “As organizational communication becomes
intercultural, another dimension is added to the analysis necessary to understand the
MNOS” (Shuter & Wiseman, 1994, p.7). Thus, this perspective considers IOC as the
interface between national culture and organizational communication. This perspective
of IOC is an effective approach to answer how Japanese MNOS overcome cultural gaps
between Americans and Japanese and create their unique organizational culture within
MNOS.

The present research deals with cross-cultural factors on organizational
communication in the context of Japanese MN Os in the U.S. I assume that American and
Japanese communication is inﬂuenced by both national cultures and the complex
organizational context of MN Os. National culture and corporate culture are considered
as key contexts for communication within MN Os. Shuter and Wiseman (1994) propose a
complex linkage between national culture, corporate culture, and organizational
communication. They introduce important communicative factors affected by national
culture: (1) organizational structure and communication, (2) role performance in an
organizational culture, (3) international human resources communication, (4)
communication between corporate headquarters and foreign subsidiaries, and (5)
marketing communication.

In particular, the present study pays attention to the four contextual types of

MNOS that Teboul, et al. (1994) introduce: (1) ethnocentric, (2) polycentric, (3)

regiocentric, and (4) geocentric organizations. The ethnocentric type of MNO
emphasizes that authority is located at headquarters and this national culture affects the
entire management in these MNOS. The polycentric type recognizes the differences
among cultures, and local nationals hold key positions in the subsidiary. The regiocentric
MNOS emphasize that stafﬁng and personnel development are regionally based. The
geocentric MNOS more emphasize the integration of diversity and a global perspective of
decision-making. It is meaningful to study organizational communication occurring in
these complex contexts of MN Os. The present research concerns how Americans and
Japanese work collaboratively in such contexts in the MNOS. They are different from the
contexts in which Americans work in American companies in the U.S. and in which
Japanese work in Japanese companies in Japan. The context of MNOS might create
unique communication patterns and communication perspectives among Americans and
Japanese. Therefore, in the later sections, I reconsider this taxonomy of MNOS and
reconstruct contextual types of MNOS in order to systematically investigate their
idiosyncratic patterns of communication.
J_ap_anese Multinational Organizations

The present research focuses on Japanese MN Os. There are many advantages to
studying Japanese MNOS. The study of Japanese organization is something of a special
case, but the quantity of generally reliable and useful research on Japanese management
and organizations has risen sharply in the last two decades (Lincoln, 1990). According to
Shuter (1992), the most-researched countries outside the U.S. for management and
communication are Japan (66 studies) and the People’s Republic of China (22 studies).

Country and region speciﬁc research is abundant about East Asian organizations,

particularly Japan (Shuter & Wiseman, 1994). The rich data on Japanese management
and communication practices provide a solid base for researchers.

There are several reasons why Japanese organizations are more studied than other
nations. Japan is the second largest country in the world economy and it is considered a
model of successful management practices. Indeed, American management has
attempted to learn Japanese management and Japanese production systems, such as the
j ust-in—time system in Toyota. In addition, a number of Japanese establishments
currently operate in the U.S. and European nations, so there are more chances to study
Japanese organizations. Another important reason is that Japanese management and
communication practices are highly heuristic for scholars in the West because they see
the cultures in Japan and East Asia as opposite to Western cultures. A number of studies
focus on the uniqueness of Japanese management and stress that Japanese culture and
homogeneity inﬂuences management practices and organizational culture.

For example, there are important discussions about Japanese management systems
in the ﬁeld of comparative industrial relations. The major question is whether Japanese
industrial relations systems are different from the other nations’ systems. Shimada
(1983) reviews that many works by English-language literature afﬁrm that Japanese
systems are different from the West’s, and it is attributed to the uniqueness of Japanese
cultures. Shimada (1983) classiﬁes these studies into four approaches: classical cultural
approach, descriptive institutional approach, functional analysis approach, and
neoculturalist approach. His classiﬁcation also reﬂects the chronological order of
theoretical development. The classical cultural approach emerged in the 1950’s and

1960’s. Abegglen (195 8) is a representative work. He characterized Japanese labor

relations as being analogous to “family relations” and a reﬂection or outgrth of
Japan’s pre-industrial social organization. This view emphasizes the unique Japanese
industrial relations as the “three pillars”: life-time employment, seniority wage, and
enterprise unionism. The descriptive institutional approach emerged in the 1960’s and is
recognized as a problem-centered approach. The functional analysis approach, which
came out after 1960’s, emphasizes “market competition in economics, conﬂict and
conﬂict resolution in sociology, or power in political science” (Shimada, 1983, p. 13).
The common perspective among functional analysts is to reject the Japanese uniqueness
of industrial relations systems. They believe that Japanese systems are not signiﬁcantly
different but more similar to the West’s in many respects. Also, the characteristics of
Japanese systems do not stem from pre-industrial or feudal relations but result from
rational reactions Or optimal choices in economic decision-making. Thus, proponents of
the cultural approach and the functional analysis approach have had a major debate about
the uniqueness of Japanese management and commonalities between Japanese and
Western management.

From a communication point of view, it is also important to consider this
problem. Shuter and Wiseman (1994) point out two major thrusts identiﬁed in research
and theory about MNOS; an approach stressing organizational universals and an approach
focussing on national cultural differences. The former posits “organizational principle
and behavior that are endemic to organizations regardless of culture” (Shuter &
Wiseman, 1994, p. 4). This. approach is very similar to the perspective of the functional
analysis approach in the industrial relations study. The latter is the literature on the

inﬂuences of national culture on organizational behavior, and it is similar to

neoculturalist approaches. Many studies in intercultural communication are predominant
over organizational universals, and they assume signiﬁcant cultural inﬂuences on
interpersonal communication. The present research will investigate the inﬂuence of
national cultures and organizational universals of MNOS.
Considering Units of Analysis

The organizational structures of MN Os are more complex than those of
conventional, homogeneous organizations (Teboul et al., 1994). Therefore, it is
beneﬁcial to consider the unit of analysis of the MNOS. The focus of the present study is
on Japanese MN Os in the U.S., and a group of Japanese employees and that of American
employees are the major units of analysis. Suzuki (1998) identiﬁes that Americans and
Japanese form distinct groups within Japanese MNOS, and they are recognized as culture
identity groups based on their nationality (Cox, 1993). Thus, intergroup relationships
(i.e., ingroup and outgroup communication) between two culture groups are the critical
point of study in the Japanese MNO in the U.S. The major question is how Japanese and
American groups interact with each other in order to reduce their communication gaps
and integrate cultural diversity. For this end, the studies of interpersonal relationships
between Americans and Japanese are literature-reviewed in the ﬁeld of intercultural
communication. Another level of context is the organizational dimension of MNOS,
which is typically discussed in management literature, such as similarities and differences
between Japanese and American organizations and management styles. It is also
necessary to consider the issues of workplace diversity and multiculturalism in
organizational settings. Thus, the present study reviews the literature of three levels of

contexts: interpersonal (dyad) context, group context, and organizational context, and it

10

seeks to analyze communication inﬂuenced by two national cultures and also identify

distinctive organizational communication in Japanese MNOS.

Interpersonal Context of Intercultural Organizational Communication
Communication and culture reciprocally inﬂuence each other. The culture in which
individuals are socialized inﬂuences the way they communicate, and the way that
individuals communicate can change the culture they share over time. (Gudykunst &
Ting-Toomey, 1996, p. 3)

As Gudykunst and Ting-Toomey (1996) mention above, most intercultural
communication studies assume that cultural factors are closely related to individuals’
communication styles. Many empirical studies have focused on the different
communication styles in the East and the West, and they examine either consistency or
inconsistency with stereotypical views of American and Japanese communication styles
(Hasegawa & Gudykunst, 1998; Neuliep, 1997; Stephan, Stephan, Saito, & Barnett,
1998; Sugimoto, 1997; Sullivan & Taylor, 1991)2. In general, American and Japanese
communication styles are assumed to be different on the basis of national cultures. In
order to conﬁrm the norms of cross-cultural stereotypes, it is necessary to review the
cultural variability of interpersonal relationships between Americans and Japanese.
Cultural Variability in American and Japanese Communication

In the literature of intercultural communication, typically, culture can be
considered as “a set of fundamental ideas, practices, and experiences of a group of people
that are symbolically transmitted generation to generation through a learning process”

(Chen & Starosta, 1998, p. 25). Therefore, cultural variability is often studied on the

11

basis of value orientation and communication patterns among different national cultures
(Hall, 1976; Hofstede, 1980;K1uckhohn & Strodbeck, 1961).

Individualism-collectivism is the major dimension of cultural variability isolated
by theorists across disciplines and across cultures (Gudykunst, 1997 ; Gudykunst & Ting-
Toomey, 1988). Individualistic cultures are described as self—oriented, placing personal
goals over the group’s goal, while collectivistic cultures are described as emphasizing
group-orientation, placing the group’s goals over individual goals. Although there have
been discussions of oversimpliﬁcation in this dichotomy, individualism-collectivism is
the most studied dimension in cross-cultural research. For example, it is used in
Hofstede’s (1980) well-known research. He developed the individualism index (IDV)
and measured the degree of individualism. According to his work, Americans and
Japanese are clearly characterized by this index: USA = 91 in IDV and Japan = 46 in IDV
(Hofstede, 1997). The U.S. ranked ﬁrst and Japan is rated 22nd out of 50 countries. This
indicates that the U.S. is the most individualistic country, while Japan is a relatively low
individualistic culture and often recognized as a collectivistic culture. Although both
individualism and collectivism exist in all cultures, one pattern tends to predominate
(Gudykunst & Matsumoto, 1996). Thus, it is commonly accepted that individualism
predominates in Americans and collectivism predominates in Japanese. In order to
examine American-Japanese communication in the MNOS, the present research basically
acknowledges the following claim as the norm of cross-cultural stereotypes: Americans
will be high on individualism and Japanese will be high on collectivism or low on

individualism.

12

However, recent intercultural communication studies have criticized the
oversimpliﬁcation of individualism-collectivism and sophisticated the study of the
cultural orientations in Americans and Japanese (Gudykunst & Matsumoto, 1996;
Nishida, 1996). One major criticism comes from Japanese scholars because the scheme
of individualism—collectivism is biased by Western perspectives. Nishida (1996)
introduces the reactions from Japanese scholars, and he argues that collectivism is not an
accurate description for Japanese culture3 . Instead of collectivism, Japanese scholars
consider that Japanese communication may be better considered by the terms,
“relativism” and “contextualism.” Another criticism from American scholars is that the
cultural-level of individualism-collectivism is mediated by individual level factors, such
as values, personality orientation (idiocentrism and allocentrism)4, self-construals5
(independent and interdependent), and so on (Gudykunst, 1997; Gudykunst &
Matsumoto, 1996). These critiques have improved the study of American and Japanese
cultural orientation.

In short, American cultures are basically characterized by individualism but
mediated by individual-level factors, the strength of idiocentrism and independent self-
construal. Japanese cultures are stereotypically characterized by collectivism but more
accurately described as relativism and contextualism mediated by allocentrism and
interdependent self-construal.

In addition to individualism-collectivism, Hofstede’s (1980) study describes
American and Japanese culture orientations in the other three dimensions. American
culture is characterized as low uncertainty avoidance, low masculinity, and low power

distance. Japanese culture is oriented in high uncertainty avoidance, the highest score in

13

a masculinity dimension, and slightly higher power distance than American’s. The other
important variable is communication patterns, such as hi gh- and low-context
communication (Hall, 1976). Americans tend to have low-context communication, and
they emphasize verbal and explicit expressions. Japanese tend to have high-context
communication, and they often use implicit communication and equivocal expressionsb.

Thus, as the intercultural communication literature illustrates, Americans and
Japanese are signiﬁcantly contrasting in cultural orientations and communication styles,
so Japanese MNOS in the U.S. can be expected to contain two different types of
personnel within an organization. If both ethnic groups maintain their own cultures and
communication styles, there are potential conﬂicts within an organization. Improving
communication between two culture groups requires that individuals become aware of
how they communicate (Gudykunst & Nishida, 1994). In this sense, an important point
of study is to identify whether Americans and Japanese are either divergent or convergent
in communicative behaviors and values under the context of multicultural organizations.
Inggoup and Outggoup Communication

As Suzuki (1998) ﬁnds distinct groups between American and Japanese
employees, in the present study, I assumed that they shape their own cultural groups in
the Japanese MN Os in the U.S. Cox (1993) more precisely categorizes and classiﬁes two
types of group identity: “phenotype identity group” and “culture identity group.” The
former stems from physical and virtual difference while the latter stems from different
values and norms. In this sense, cultural identity groups are not necessarily deﬁned by
national cultures only. They may be deﬁned by social and organizational factors, such as

gender, social class, jobs, and so on. However, because cross-cultural difference must be

14

the most primary issue handled by management (Cutcher—Gershenfeld et al., 1998).

Since national cultures are deeply rooted in the socialization of organizational members,
it is plausible that nationality continues to affect members’ preference and behavior (Salk
& Brannen, 2000). Therefore, it is meaningful to examine how organizational
communication differs in two culture identity groups based on nationality in multicultural
context. American employee groups and Japanese employee groups are the units of
analysis in terms of ingroup and outgroup relationships.

“Cultural differences become especially salient in the Japan-U.S. business
setting” (McDaniel & Quasha, 2000, p. 312). Within Japanese MNOS in the U.S.,
ingroup communication and outgroup communication are the most critical features in
American groups and Japanese groups because individualistic and collectivistic cultures
differentiate the communication patterns in intergroup relationships. For instance,
collectivistic cultures emphasizes social norms of ingroup rather than individual pleasure,
sharing ingroup beliefs rather than unique individual beliefs, and cooperation with
ingroup members rather than maximization of individual outcomes (Gudykunst, Yoon, &
Nishida, 1987; Triandis, 1986). Therefore, it is generally considered that the members of
collectivistic cultures have clear distinctions between ingroup relationships and outgroup
relationships (Chen & Starosta, 1998; Gudykunst & Ting-Toomey, 1988). In contrast,
“since individualistic cultures have many speciﬁc ingroups, they exert less inﬂuence on
individuals than ingroups do in collectivistic cultures where there are few general
ingroups” (Gudykunst et al., 1987, p. 296). Therefore, it can be considered that members
of individualistic cultures do not have clear distinctions between ingroup and outgroup

relationships.

15

There have been few studies about cross-cultural perspectives of ingroup-
outgroup communication (Salk & Brannen, 2000; Gudykunst et al., 1987; Oetzel, 1998;
Suzuki, 1997, 1998). The impact of individualism on the social penetration process in
ingroup and outgroup relationships in Japan, Korea, and the U.S. was examined by
Gudykunst, Yoon, and Nishida (1987). Their study revealed that “the greater the degree
of collectivism present in a culture, the greater the amount of personalization and
synchronization, but the less the difﬁculty perceived in communication with classmates”
(p. 301). The members of collectivistic cultures (Japan and Korea) perceive more
personalization and synchronization (greater social penetration) in ingroup relationships
than members of individualistic cultures (the U.S.). The data about communication in
ingroup relationships clearly support the predictions derived from the conceptualization
of individualism-collectivism. However, “the results for communication with outgroup
members were not as clear-cut” (p. 302). For instance, there are strong situational
demands on the behavior in collectivistic cultures, so the amount of personalization in
communication with members of outgroups is speciﬁed by the situation (Gudykunst &
Ting-Toomey, 1988). Gudykunst et al. (1987) attempt to show plausible explanations
using Triandis’s (1986) distinction of collectivism, such as Japanese culture as
‘contextual collectivism’ and Korean culture as ‘simple collectivism.’ In sum, the study
indicates that “individualism-collectivism was related systematically to perceptions of
communication in ingroup relationships, but its relationship to perceptions of
communication in outgroup relationship was more complicated” (Gudykunst et al., 1987,
p. 295). In this sense, outgroup communication is not well explained from the traditional

perspective of individualism-collectivism. However, outgroup communication might be

16

critical in the context of MN Os because two culture groups may attempt to enhance their
outgroup relationships and reduce their communication gaps in the MNOS. The MNO
needs to integrate diverse workforces, so strong ingroup relationships within Japanese
and American groups become barriers to developing effective MNOS. Therefore,
increase of outgroup communication might be a distinctive feature in MNOS, and this
general hypothesis will be examined in the present study.

Salk and Brannen (2000) and Suzuki (1997, 1998) approach ingroup and outgroup
relationships in Japanese MN Os, which is a similar context to the present study. While
Gudykunst et al. (1987) measure the differences of ingroup and outgroup relationships in
terms of the perception of social penetration, their studies have a great advantage of
directly measuring communication contacts in ingroup and outgroup relationships. They
uses the methods of social network analysis. Suzuki (1997, 1998) measures the
frequencies of ego—networks in terms of ingroup and outgroup relationships in several
Japanese MNOS in Chicago. She classiﬁed communication ties in terms of three types:

( 1) task speciﬁc, (2) task general, and (3) non-task. In contrast, Salk and Brannen (2000)
have social network data of 15 managers from an international joint venture between
Germany and Japan. Similar to Suzuki, they classiﬁed three network types based on
Ibarra (1993): (1) task-related, (2) advice-related, and (3) private. They commonly
assume that communication contacts in ingroup and outgroup relationships are different
in task-oriented (i.e. formal) and nontask or social-oriented (informal) context. Three
network types are conceived as a useful approach to measure ingroup and outgroup

relationship.

17

Salk and Brannen (2000) and Suzuki (1997, 1998) commonly underlie the social
identity theory that affects communication ties in ingroup and outgroup. Suzuki’s (1997,
1998) shows the evidence of two culture identity groups based on the nationality among
Americans and Japanese employees in Japanese MNOS in the U.S. Another evidence
found in her Japanese sample is that greater identiﬁcation is associated with less social
distance in ingroup whereas the American sample was somewhat weaker (Suzuki, 1998).
These results are consistent with the expectation of individualism-collectivism.

Similarly, Salk and Brannen (2000) ﬁnd strong ingroup density in their Japanese
sample while the German data shows almost no signiﬁcant difference in their propensity
to use ingroup and outgroup members for advice-related and private communication, not
for task related communication. They conclude that there are a general pattern of
ingroup-preference and strikingly different patterns in Japanese and German advice—
seeking. Also Salk and Brannen (2000) conclude “that culture is important but that its
role is far more complex than past research and theory suggest” (p. 199).

There is an important ﬁnding for outgroup communication in Suzuki’s (1998)
study. “Task-speciﬁc communication associated with greater social distance in outgroup
communication was found for high- and low-identiﬁcation groups in both the U.S. and
Japanese samples” (Suzuki, 1998, p. 175). It can be considered that outgroup
communication relies more on the task-oriented context than the norms of individualism-
collectivism. In other words, outgroup communication is related with organizational
contexts in MNOS, and this is an important implication for the present study.

In sum, “the in-group and out-group distinction in terms of the national cultural

labels is meaningful to the organizational members” (Suzuki, 1998, p. 167). Similar to

18

the study of Gudykunst, et al (1987), the conceptualization of individualism-collectivism
can generally explain the ingroup-outgroup relationship in the studies of Salk and
Brannen (2000) and Suzuki (1997, 1998). However, according to Suzuki (1998),
outgroup communication is more complicated than anticipated: an individual’s social
identity partly explains her or his intergroup relationships. Suzuki (1998) found that
outgroup relationship is more inﬂuenced by task-oriented context, and this implies that
outgroup communication is central in executing tasks and maintaining MNOS.

Moreover, as Salk and Brannen (2000) propose that social network concepts enrich
understanding of individual inﬂuence in multicultural context. Thus, the measurement of
communication ties by network analysis is a useful and effective method to observe

ingroup and outgroup communication in the MNOS.

Organizational Context of Intercultural Organizational Communication

In this section, Japanese MNOS in the U.S. are considered from an organizational
perspective. Three beneﬁcial perspectives to organizational communication in the MNOS
are identiﬁed in management and communication literatures: 1) organizational behavior
in the cross-cultural management, 2) Ouchi’s organizational theory, and 3) workplace
diversity. These previous studies directly and indirectly contribute to the investigation of
Japanese MNOS in the U.S.
Organizatiorgrl BQavior in Cross-cultural Management

The study of multinational organizations is different from cross-cultural
management. The latter usually takes comparative approaches or focuses on

management styles of one speciﬁc country. Many of these works bring up differences

19

between American management and Japanese management (Briggs, 1988; Chen &
Chung, 1994; Goldman, 1994; Kim & Paulk, 1994; Klopf, 1991; Lincoln, 1989, 1990;
Lincoln & Kalleberg, 1985; McDaniel & Quasha, 2000; Misumi, 1984; Ouchi, 1981;
Sullivan & Taylor, 1991; Yang, 1984).

American and Japanese organizational behavior and work communication are
very consistent with their interpersonal context illustrated in the intercultural
communication studies. For example, McDaniel and Quasha (2000) characterize
Japanese business style as collectivistic, hierarchical, formal and socially stable (wa),
while American style is described as individualistic, egalitarian, informal, and change.
Briggs (1988) shows ﬁve beliefs of Japanese at work: (1) a job for life, (2)1 love my
company, (3) happy workers, (4) group success, and (5) humble managers. Yang (1984)
and Misumi (1984) point out the Japanese ringi system as an example of the bottom-up
style of decision-making with consensus, in other words, participative decision-making.
Similarly, Lincoln (1985, 1989, 1990) ﬁnds that Japanese are more committed to
decision-making processes and to the company than Americans. Goldman (1994)
summarizes differences between American and Japanese organizational behavior as

follows.

These comparisons help to understand general pictures of American and Japanese
organizational behavior, and they can be applied to the Japanese MNOS in the U.S. Kim

and Paulk (1994), however, directly conduct their research on the MNO where

20

Americans and Japanese work together, and they investigate their coworker relationships.
They had interviews with both groups of employees and asked both parties about
problematic issues in terms of language usage, work styles, and management styles.
When the work styles are focused, American employees complain about detailed work
procedure, strong adherence to the company ways, and unclear job deﬁnitions. In these
points, Americans consider that Japanese ways of working are ‘inefﬁcient’ (p. 127). On
the other hand, Japanese complain about Americans’ unwillingness to devote necessary
time to learn, tendency for rushing a conclusion, and unwillingness to compromise
personal interests.

Considering their statements, Japanese employees pay attention to the process
more than direct conclusions and outcomes. Rehfeld (1990) also introduces Japanese
process-oriented work styles as a kaizen system (a system for continuous improvement)
that exempliﬁes employee participation in the decision process and their strong concerns
about the process of problem-solving. This is related to the strong ingroup norms and
commitment to the organization. Americans often point out many (unnecessary)
meetings. One comment tells that these meetings seem to be wasting time. Japanese
emphasize such processes and meetings as ritualized ways of creating cohesiveness,
enhancing norms of ingroups, and showing their loyalty to corporate authority. American
workers are more market-oriented and more rational in the Western concepts. They
emphasize their personal interests, and they consider that Japanese ways are irrational
and inefﬁcient from their Western point of view. Therefore, some American employees
can not tolerate such Japanese methods so much that they quit their jobs. This Japanese

MNO faces a high turnover rate (Kim & Paulk, 1994).

21

Ouchi’s Organization Theory

Presently Ouchi’s ‘Theory 2’ (Ouchi, 1981) is very popular in the organizational
and management literatures. He contrasts cultural traits of the American type of
organizations (Type A) and Japanese type of organizations (Type J). Type A and Type J
are signiﬁcantly consistent with organizational behavior in Americans and Japanese
respectively. For example, American organizations emphasize “explicit control
mechanism, individual decision-making, individual responsibility, and segment concern”,
whereas Japanese organizations emphasize “implicit control mechanisms, collective
decision-making, collective responsibility, wholistic concerns” (Ouchi, 1981, p. 58). He
argues that these differences in two types of organizations reﬂect cultural differences
between two nations.

His theory posits three basic systems of social coordination: clan, market, and
bureaucracy (Ouchi, 1980). They are control systems in organizations and any system
alone can not completely control an organization. Both Type A and Type I contain three
systems, but they have varying degrees of emphasis or tendencies on these three systems.

Market and bureaucracy are two principle mechanisms for rational means of
control (Hatch, 1997). Market is a source of control as well as a decision—mechanism: a
price is determined by supply-demand through competition. Prices are taken as
indicators of economic performance in competition, and this price-mechanism in
competition always tries to maximize proﬁts. In addition, the organizations based on
market focus on controlling the outputs and the results of organizational behaviors
(Ouchi, 1979). The market control underlies the assumptions that individuals are

independent and they have a freedom of exchange in the market. Such emphasis on

22

independent individuals and exchange relationships are very similar to the characteristics
of individualistic societies. Thus, market control is relatively stronger in Type A than
Type J. The explicit control based on price, competition, and output is typical in Type A
organizations.

However, the effectiveness of a market system is limited in the conditions that the
competition principle makes pricing and output-control meaningful. Without
competition, prices can not indicate internal efﬁciency because there is no comparison
possible (Hatch, 1997). When market fails, organizations generally turn to bureaucracy
(Ouchi, 1980), which relies on a combination of rules, procedure, documentation, and
surveillance to achieve control. Bureaucracy control emphasizes the standardization of
behavior, and it is underpinned by a legitimatized hierarchy of authority (Ouchi, 1980).
Instead of a pricing mechanism, rules govern the process of organizational performance
in bureaucracy and surveillance is critical for behavioral control; therefore,
communication in bureaucracy emphasizes formal aspects rather than informal aspects.
Emphasis on formal authority are more typically identiﬁed in Type I organizations. For
example, Kim and Paulk (1994) show that American employees complain about the
many meetings and detailed procedures of Japanese working styles. Such formality in
bureaucracy seems redundant and irrational for the American ‘market’ point of view.

Rapid changing and complex environments escalate uncertainty and ambiguity,
and these circumstances make bureaucracy ineffective in organizational control. Rational
means of control by market and bureaucracy will not succeed in complex and dynamic
environments (Hatch, 1997). ‘Unobtrusive forms of control’ (Conrad & Poole, 1998, p.

119) is considered more effective in these contexts. Cultural values, norms, and

23

expectations provides the primary mechanisms of control in ‘clan’ (Ouchi, 1980). The
members in clan share the same culture and traditions since socialization processes create
common values and norms. Clan emphasizes implicit understanding of values and
beliefs that guide and control organizational behavior. In this sense, collectivistic (non-
individualistic) cultures are very consistent with the characteristics of clan organizations.
So, clan control can be more clearly identiﬁed in Type I organizations than Type A
organizations. For example, according to Kim and Paulk (1994), American employees
point out unclear job deﬁnitions and strong adherence to the company ways in a Japanese
MNO. They are the examples of normative control and the existence of strong values
within the organization. Lincoln’s ﬁndings (1989, 1990) of Japanese strong commitment
to the company is a case of clan control. Thus, in Type J, ceremonial and symbolic
control is used through the mechanism of commitment and socialization, and
organizational norms and values are central among organizational members.

Ouchi (1980, 1981) provides the fundamentals of cross-cultural norms in
American and Japanese organizations for the present study. In short, teamwork,
workplace harmony, and wholistic concern for people are important principles in
Japanese business practice (Ouchi, 1981). This trait of Japanese business is recognized
as a ‘clan ’ type. On the other hand, competition is a strong value within American
business. Results of competition are directly related to one’s salary and/or promotion.
Outcomes are emphasized in the US while processes are emphasized in Japan. American
business orientations rely on a ‘market’ type. Moreover, Japanese business tends to
emphasize hierarchy rather than American business. As Japanese industrial relations are

famous for life-long employment and the use of seniority systems (Durlabhji & Marks,

24

1993; Kuwahara, 1993), these features underlie the hierarchical system of Japanese
organizations; therefore, Japanese business orientations emphasize formal authority and
bureaucratic control.
Diversity Management Perspectives

Workplace diversity is one of the most important management issues in U.S.
organizations (Chen & Starosta, 1998; Cox, 1993; Larkey, 1996; Limaye, 2000).
Multinational organization is categorized as a part of a culturally diverse workplace, and
a diversity management perspective can be applied to the study of Japanese MNOS in the
U.S. However, the studies of workplace diversity embrace broader issues than cultural
diversity, such as minority, racial, and gender issues. Therefore, in this section, I focus
on multiculturalism and multicultural organizations because J apanese-owned companies
in the U.S. are multicultural. In addition, important conceptualizations of communication
accommodation perspectives are reviewed with regard to workplace diversity.

Multicultural Organizations and Multicultuﬂlism. According to Cox (1991,
1993), a multicultural organization is described in terms of six dimensions: acculturation,
structural integration, informal integration, cultural bias, organizational identiﬁcation,
and intergroup conﬂict. Multicultural organizations must be integrated by a cultural
dimension as well as organizational structure, such as hiring, job status and informal
network among the members of culture groups. In addition to cultural bias, intergroup
conﬂicts should be solved in multicultural organization. In these regards, the six
dimensions are the guidelines to achieve the effectiveness of multicultural organization.

Thus, cultural diversity does not necessarily lead to poor performance. In fact,

cultural diversity might even confer advantages by giving organizational members a

25

broader range of perspectives for managing complex cultural systems (Salk & Brannen,
2000). Effective multicultural workplaces have the capability to overcome serious
conﬂicts and integrate cultural heterogeneity.

Multiculturalism is “a dynamic process that move us toward cultural expansion,
awareness, sensitivity, and competence (Chen & Starosta, 1998, p. 229); therefore,
effective MN Os emphasize and facilitate effective multiculturalism that can transform
negative aspects of cultural heterogeneity to a positive side, and produce beneﬁcial
aspects of cultural diversity. This changing process is started from monoculturalism
(Chen & Starosta, 1998; Wurzel, 1988), in which ethnocentrism is dominant.
Multiculturalism is the ﬁnal stage and goal of cultural diversity.

Communication Accommodation Perspective. Larkey (1996) pays attention
to the interaction types of convergence and divergence in the context of workplace
diversity. These are important factors for effective multicultural organization.
According to her, divergent communicative behavior is evoked, “(a) when conﬂict or
threat to the cultural group boundary occurs, (b) when ethnicity is salient, and (c) when
the solidarity of the ethnic group is strong” (p. 479). Divergence is recognized as
adherence to different patterns of cultural communication. Therefore, divergence
enhances antagonistic intergroup relationships, and it is associated with a negative side of
diversity in the workplace. On the contrary, convergence is recognized as adjustment of
communication style to match the counterpart, and it is effective in mitigating conditions
of threat, conﬂict, and competition.

A similar perspective can be identiﬁed in the theories of relationship development

in intercultural communication. Indeed, Larkey (1996) borrowed the conceptualization

26

of convergence-divergence from communication accommodation theory (CAT)
developed by Gallois, Franklyn-Stikes, Giles, and Coupland (1988), which is used to
explain the relationship development in the process of intercultural encounters. CAT has
three cardinal concepts: convergence, divergence, and maintenance. Convergence refers
to the change of language, vocabulary, speech styles, or tone of voices to become similar
to the intercultural counterpart. Divergence refers to the emphasis on differences in their
communication styles, and maintenance refers to the continuation of their own style of
communication. CAT focuses on a different context from the MNO study, but it provides
interesting implications to the present research. Namely, convergence tends to increase
attraction between two culture groups and divergence tends to inhibit it (Chen & Starosta,
1998).

These conceptualizations in relationship development are applicable to
development of intergroup relationships between Americans and Japanese in MNOS. If
two culture groups emphasize ingroup relationships, there must be few interactions
between them. Strong preference on ingroup communication may lead to antagonistic
relationships, so it is recognized as an interaction pattern of divergence. On the other
hands, if two culture groups have active outgroup communication, there must be a
number of interactions between them. Convergence is a way to develop good
relationships between two culture groups, and therefore emphasis on outgroup
communication is recognized as an interaction pattern of convergence. In this respect,
outgroup communication is critical for effective MNOS, and it must be enhanced for the

sake of convergence and multiculturalism in MNOs.

27

Furthermore, communication frame theory developed by Drake and Donohue
(1996) provide an important insight to the present research. They pay attention to
disputants’ communication frames toward the issues and examine whether their frames
are convergent or divergent: “how frames converge or diverge to form more or less
cooperative interaction contexts” (Drake & Donohue, 1996, p. 306). According to them,
“[c]onvergence refers to adopting others’ communication behavior and values whereas
divergence describes accentuating the differences between self and others” (p. 306).
Therefore, divergence of communication ﬂames between disputants tends to lead to
negative relational deﬁnitions, while frame convergence has positive impacts on their
relationships by evoking social approval and perceptions of attractiveness. Thus,
convergence in communication ﬂames creates a cooperative climate, and it is essential to
reduce antagonism, build trust, and resolve disputes.

This conceptualization is applicable to the context of workplace diversity and the
MNO study. When communication ﬂames between two culture groups are convergent,
they may lead to more cooperative relationships within a diverse workplace. In terms of
the MNO study, either convergence or divergence of communication ﬂames is a useful
indicator for examining the complex context of MNOS.

Thus, the present research recognizes that ‘convergence’ in communication
ﬂameworks and intergroup communication between two culture groups is critical for
multiculturalism in MNOS. Either divergence or convergence in communication
becomes a predictor “for creating a more favorable environment” (Larkey, 1996, p. 484).
I examine whether and how the data show divergence or convergence in terms of the

communication ﬂames and the patterns of ingroup and outgroup communications.

28

Framing Perspectives as Applied to MN Os

The present research attempts to identify how individuals’ ﬂameworks of
communication tend to be convergent or divergent in MNOS. The ﬂaming concept has a
long history in social science, especially in relation to more micro perspectives (Chang,
Johnson, Cox, & Kiyomiya, 1997). Putnam and Holmer (1992) describe the classic
works by Bateson and Gofﬁnan as follows:

Thus Bateson (1972) reminds us that ﬂames are dynamic, reﬂexive, and intertwined
with implicit as well as explicit messages. Gofﬂnan (1974) orients ﬂaming to ways
of deﬁning the social situation. He centers on the multidimensional, multilayered
nature of frames anchored in past experiences. Moreover, Gofﬁnan contributes the
concept of primary ﬂamework, which links ﬂaming to a larger social context
(Putnam & Holrner, 1992, p. 148).

Bateson’s perspective emphasizes that communication is the essence of ﬂaming
through sets of messages, metacommunication, and the premises of interactions. Bateson
(1972) described his terminology of ‘ﬂame’ as follows: “A ﬂame is metacommunicative.
Any message, which either explicitly or implicitly deﬁned, ipso facto gives the receiver
instructions or aids in his attempt to understand the messages included within the frame”
(p. 188). Therefore, a function of a frame is to delimit a set of messages and meaningful
actions and to delimit a logical type.

In contrast, Gofﬂnan’s perspective emphasizes integration between the macro
(social) context and the micro context. “[D]eﬁnitions of situation are built up in

accordance with principles of organization which govern events —- at least social ones --

29

and our subjective involvement in them” (Gofﬂnan, 1974, p. 10). A function of ﬂaming
is to provide an organizing system or principle which governs social events. The
function of social framework is to “provide understanding for events that incorporate the
will, aim, and controlling effort of an intelligence, a live agency, the chief one being the
human being” (Gofﬂnan, 1974, p. 22). His signiﬁcant contribution is that people’s
frames are rooted in everyday life: they are mundane rather than scientiﬁc. Like
Bateson’s metacommunicative nature, ﬂames are viewed as inherently delimiting,
providing individuals with a situated context for action and for interpretations of
particular strips of activity (Goffman, 1974). .

Frames have recently received renewed interest in organizational
communication7, in part because they offer an approach to examining the context of
cooperative relationships (Johnson, 1997b). According to Johnson (1997a), ﬂames are a
basis for coordinated action in collectivities, since cooperation requires a ‘reading’ of
other’s actions and intentions.

There are some works that delineate ﬂames and ﬂaming in organizational context.
For example, Fairhurst and Sarr (1996) suggest that managerial effectiveness rests on the
management of meaning that is largely accomplished through ﬂaming. They focus on
framing as ‘skills’ emphasizing context sensitivity, tools (e.g. metaphor and stories),
avoiding mixed messages, and establishing credibility. Similarly, Bolman and Deal
(1991) view ﬂames as tools for leaders: “the truly effective manager and leader will need
multiple tools, the skills to use each of them, and the wisdom to match to situations” (p.

12).

30

Frameworks for Interaction (F INT) Scaﬁ

Johnson (1997a) created an instrument that assesses communication ﬂameworks
of individual employees in the context of cooperative relationships. He focuses on
frameworks that provide a more encompassing context for interactions within
organizations.

Frameworks for interaction, then, constitute the basic context within which
communication occurs by promoting certain levels of shared understanding of
meanings, orienting interactants to the nature of the event, and establishing the
ultimate purpose of continuing interaction. (Johnson, 1997a, p. 128)

The F INT scale is composed of ﬁve dimensions: (1) exchange, (2) formal, (3)
sentiments, (4) normative, and (5) negotiated order, which are explicated on the basis of
literature and Johnson’s pilot studies.

Exchange. This is the most popular modern ﬂamework as well as the most
fundamental of the ﬂameworks (Johnson, 1997a). Since employees in business
organizations are in a capitalistic society, underlying market conceptions are crucial for
them to communicate with other members to survive in the organizations. The
underlying cultural view of Lockean individualism is dominant in large ﬂame (Johnson,
1997a). Namely, the exchange frame will indicate communication based on their market
conceptions and utilitarianism: individuals are driven to maximize their individual
interests and rewards through interaction with each other. The exchange relationships
underlie the assumptions that individuals are independent and they have a perception of
mutual gain through fair competition, such as maximization of individual interests. This

is very similar to conceptualization of the ‘market’ type that Ouchi (1980) illustrates.

31

Market control is relatively stronger in Type A than Type J. In other words, it can be
hypothesized that American workers tend to more emphasize the exchange frame than
Japanese workers.

M. The formal framework essentially represents the bureaucratic world of
organization, with its speciﬁcation of patterns of super and subordination and the nature
of relationship between parties (Johnson, 1997a, p. 129). Organizational hierarchy and
formal structure provide a principle of action in business, and the formality in the
organizations are shared and legitimated among members. Most often the context of
formal structure can be conceived as embedded in the formal authority structure of
organization, usually associated with bureaucracy. This is similar to Ouchi’s
conceptualization of the “bureaucracy” type. Instead of a market mechanism, formal
authority and rules govern communication process in organizational performance.
Emphasis on formal authority are more typically identiﬁed in Type J organizations as
Kim and Paulk (1994) illustrated Japanese emphasis on bureaucratic procedure and
formality. Stereotypically, it can be hypothesized that Japanese employees tend to more
emphasize the formal frame than Americans.

Normative. “Culture is communication and communication is culture” (Hall,
1959, p. 169). Culture and communication are inﬂuenced with each other and they are
seen as intertwined concepts. Particularly, many organization and communication
scholars have paid attention to organizational cultures in the past two decades, and they
often view culture as shared values, norms, and meanings among the members. The
normative ﬂamework depends on operations of larger collectivities and can encompass a

variety of underlying values for interaction (Johnson, 1997a). Some employees may

32

conform to the ingroup’s norms and values, emphasizing the normative framework.
These communicative behaviors often emerge in the ‘clan’ type of organizations, and
they enhance implicit understanding of values and belief that guide and control
organizational behavior. As Kim and Paulk (1994) exemplify typical normative
behaviors in Japanese personnel, it can be hypothesized that Japanese employees tend to
more emphasize the normative ﬂame than Americans.

Sentiments. Collective sentiments are closely linked with interactions in the
organizations. “Friendship and other more emotional bases for relationship often provide
the underlying context for interaction within organizations” (Johnson, 1997a, p. 130).
Such emotional communication might appear in the informal relationships. In the
domain of the formal relationships, more business-oriented types of ﬂame (e. g. exchange,
formal, and normative frames) must be stronger. In these regards, the sentiments
framework may not appear in the outgroup communication but may appear in the ingroup
communication, since Suzuki’s (1997) study shows that the outgroup communication are
task oriented. Because Japanese distinguish ingroup relationship ﬂom outgroup
relationship and emphasize ingroup norms, the sentiment framework might be strong in
ingroup communication among Japanese employees.

Negotiated Order. Four frameworks are delineated, but it is possible to
communicate with others in the unique mix of the forgoing ﬂameworks. Indeed, absence
of a dominating ﬂamework or the lack of rigid speciﬁcation allows members to create
ﬂexibility within an organization (Johnson, 1997a). This is a very autonomous
framework which rests on the interactants’ mutual agreement. Such negotiation is

designed to “establish a stable ordering of relationship governing interactions within it”

33

(p. 130). This negotiation is either implicitly or explicitly made among interactants. I
could not ﬁnd any literature or empirical studies that relate cross-cultural factors with the
negotiated order ﬂamework. It can be argued that the score of uncertainty avoidance is
one of the rationales to explain cross-cultural differences and the negotiated order
framework. Japan is a high uncertainty avoidance culture, and the U.S. is a low
uncertainty avoidance culture (Hofstede, 1980) and Americans enjoy taking a risk.
Therefore American may have stronger negotiated order ﬂamework because low
uncertainty avoidance is related to ﬂexibility, autonomy, and voluntarism that the
negotiated order framework premises.

Thus, the norms of cross-cultural stereotypes are incorporated into the F INT
subscales as follows: (1) Americans tend to more emphasize the exchange ﬂame than
Japanese; (2) Japanese tend to more emphasize the formal ﬂame than Americans; (3)
Japanese tend to more emphasize the normative frame than Americans; (4) the sentiment
ﬂamework might be stronger in Japanese ingroup communication; and (5) American may
be stronger in the negotiated order ﬂamework than Japanese. They are examined in
terms of how cross-cultural norms differ ﬂom individual’s communication ﬂameworks in

the context of MN Os.

Research Questions and Hypotheses
The present study aims to analyze the complex context of MNOs and the
inﬂuence of national cultures, so it seeks to investigate convergence or divergence of
communication in terms of the F INT scale and ingroup-outgroup communication. This

research contains three important constructs for observation and measurement: 1) cultural

34

difference, 2) communication ﬂame, and 3) ingroup and outgroup relationships. In
addition to these three, there are two more major conceptualizations: 4) the norms of
cross-cultural stereotypes and 5) divergence and convergence. The fourth is the norms or
the expectation of cross-cultural stereotypes ﬂom the past studies and literatures of
intercultural communication and cross-cultural management, which are delineated earlier.
It is used for a benchmark to identify divergence or convergence of communication.

Four phases are logically designed for approaching these constructs. The ﬁrst
phase attempts to investigate the relationship between cultures and communication
ﬂameworks, cultural effects on communication ﬂameworks. In the second phase, the
association between cultural difference and ingroup-outgroup relationships is studied,
cultural effects on ingroup and outgroup communication. The third phase focuses on the
identiﬁcation of either divergence or convergence in terms of communication
frameworks and the communication patterns of ingroup-outgroup relationships. For this
end, the results of ﬁrst and second phases are referred to the norms of cross-cultural
stereotypes. In the fourth phase, I examine whether the total score of F INT scale is
related to outgroup communication in the American and Japanese samples.
Operationalization

It is crucial to operationalize constructs for empirical research (Chaffee, 1991).
The ﬁrst construct of ‘cultural difference’ is operationally deﬁned as different values and
norms identiﬁed by the group members in the organizations. As Suzuki (1997, 1998)
found, two cultural groups exist in Japanese MNOS in the U.S. since Americans and

Japanese compose culture identity groups. American culture and Japanese culture are

35

recognized as two dimensions in a factor of cultural difference. It is hypothesized that a
factor of cultural difference may inﬂuence communication within the MNOS.

Communication ﬂameworks are considered as “the basis for coordinated action in
collectivities” (Johnson, 1997a, p. 128). Frameworks are categorized into ﬁve
dimensions by Johnson (1997a): exchange, formal, normative, sentiments, and negotiated
order. He developed the frameworks for interaction (FINT) scale which is composed of
these ﬁve subscales. Thus, the F INT subscales are the measurements of communication
frameworks.

The studies of ingroup and outgroup relationships underlie the speciﬁc intergroup
relationships, such as gender groups, minority-majority groups, and cultural groups
(Dietz-Uhler & Murrell, 1998; Gao, Schmidt, & Gudykunst, 1994; Kelly, 1990; Oetzel,
1998). The present study of Japanese MNOs in the U.S. assumes two culture identity
groups based on nationality: the American group and the Japanese group. In this regard,
ingroup relationships are recognized as communication within the American group and
communication within the Japanese group. The outgroup relationships are recognized as
American network links toward Japanese and Japanese network links toward Americans.

Emergent communication patterns in ingroup and outgroup relationships can be
observed by social network perspectives (Salk & Brannen, 2000; Suzuki, 1997, 1998).
Ingroup and outgroup communication patterns rely on an individual’s networks toward
other ingroup and outgroup members. This is the approach of a radial network or an ego—
network, and the radial network data is a focal network composed of one individual’s

overall pattern of relationship with others (Chang, et al., 1997). Therefore, it is possible

36

to observe ingroup and outgroup communications in terms of how many communication
ties with Americans and Japanese an individual had in a certain periods.

As Hartman and Johnson (1990) emphasize distinctions between formal and
informal communications in organizational settings, it is assumed that ingroup and
outgroup communications differ in the types of communication. In this research,
therefore, communication ties are categorized and observed in three types based on
communication contents: job-oriented, organization-related, and social. These three
indicates the degree of formality in communication contexts. The job-oriented network is
the most formal communication, the organization-related network is moderate in
formality, and the social network is informal communication. Thus, the construct of
ingroup and outgroup relationship is operationalized as ego-networks toward the ingroup
members and the outgroup members in terms of three communication types.

In addition, the cross-cultural norms are used for a benchmark to identify
divergence or convergence. The norms are cross-cultural stereotypes and strong
expectations about two cultures: Americans are individualistic and Japanese are
collectivistic, Ouchi’s (1980, 1981) Type A and Type J, and signiﬁcant cultural
variations in ﬁve FINT dimensions. These are the norms that stem from past academic
outcomes and business practices.

The conceptualization of divergence and convergence are used to consider the
complex contexts of MNOS. These conceptualizations describe the communicative
relationship between Americans and Japanese. Following Larkey (1996) and Drake and

Donohue (1996), communicative divergence implies increasing differences and distances

37

between two groups while communicative convergence implies decreasing differences
and enhancing integration of two culture groups.
Resgarch Questions and Hypotheses
In Phase One, I examine the relationship between a cross-cultural factor and the

FINT subscales: exchange, formal, normative, sentiments, and negotiated order. The
FINT is the measurement of communication frameworks, and it is the dependent variable,
which is examined in terms of an independent variable of a cross-cultural factor. As
mentioned previously, the cross-cultural norms in FINT subscales are examined in the
American and Japanese samples. A major question is whether the cross-cultural
differences between Americans and Japanese stereotypically appear in the F INT
subscales. The cross-cultural norms in the ﬁve F INT dimensions are analyzed to detect
cultural effects on communication frameworks.

[Research Question 1]

RQl-a: Are there signiﬁcant differences between Americans and Japanese in the

F INT subscales?

RQl-b: Are the scores of the F INT subscales different ﬂom the norm of cross-

cultural stereotypes in the Japanese MNOS in the U.S.?

[Hypotheses 1]

Hl-a: American employees will be signiﬁcantly higher in the exchange ﬂamework

than Japanese.

Hl-b: American employees will be signiﬁcantly lower in the formal framework than

Japanese.

38

Hl-c: American employees will be signiﬁcantly lower in the normative ﬂamework
than Japanese.

Hl-d: American employees will be signiﬁcantly higher in the negotiated order
framework than Japanese.

In Phase Two, the relationships between cultural difference and ingroup-outgroup
relationships are examined. As mentioned early, the ingroup and outgroup relationships
emerge into the individuals’ ego-networks toward Americans and Japanese in the MN Os.
The relative strength between the ingroup and outgroup communication is indicated by
the percentage of the total ego-networks to the ingroup members and the outgroup
members. The ego—networks of ingroup and outgroup communications are the dependent
variables, which are observed in terms of job-oriented, organization-related, and social
communication networks. A factor of cultural difference is the independent variable, and
its direct effects are investigated on ingroup and outgroup communications.

As the norms of cross-cultural stereotypes suggest, it is expected that ingroup and
outgroup communication will differ in individualistic cultures and collectivistic cultures.
I investigate whether the cross-cultural differences between Americans and Japanese
stereotypically appear in the ingroup communication and outgroup communication.
Typically, Japanese as a collectivistic culture prefer ingroup relationships more than
outgroup relationships while Americans do not have speciﬁc emphasis on ingroup and
outgroup relationships.

[Research Question 2]
RQ2-a: Are there signiﬁcant differences between Americans and Japanese in the

communication patterns of ego-networks toward ingroup and outgroup members?

39

RQ2-b: Are the relative strengths of ingroup and outgroup communication different

ﬂom the norm of cross-cultural stereotypes in the Japanese MNOS in the U.S.?

[Hypotheses 2]

H2-a: For Japanese, the ratio of ego-networks to ingroup members will be

Si gniﬁcantly higher than that of ego-networks to outgroup members in informal

communication.

H2-b: The ratio of the ego-network to ingroup members will be signiﬁcantly higher in

the Japanese than in the Americans across all communication types.

H2-c: For the American employees, there will be no signiﬁcant difference between

ingroup relationships and outgroup relationships across all communication types.

The research in Phase Three seeks the identiﬁcation of either divergence or
convergence in terms of the F INT scale and the ego-networks of ingroup-outgroup
relationships. The analysis of the F INT scale indicates either frame convergence or
ﬂame divergence. When F INT scores conform to a direction of cultural stereotypes, it
indicates convergence and implies strong monolithic culture (i.e., American or Japanese
culture is dominant). When the FINT scores signiﬁcantly differ in two groups, it
indicates divergence and implies pluralistic culture. When there is no difference between
Americans and Japanese in the scores of the F INT subscales and they are different ﬂom
cultural stereotypes, it indicates cultural synthesis and multiculturalism.
In addition, the analysis of ingroup-outgroup communication relates to the

interaction types of divergence and convergence. When two culture groups place little
stress on outgroup relationships and only focus on ingroup communication, it is

considered as communicative divergence. When two culture groups mutually stress

40

outgroup communication, it implies convergence of communicative interaction in MNOS.
Outgroup communication is a key to identify convergence. Thus, the measurement
outcomes of the F INT and ego-networks are related to the conceptualizations of
divergence and convergence.

[Research Questions 3]

RQ3-a: Are the patterns of the FINT subscales divergent, convergent, or identical

(synthesis) between Americans and Japanese?

RQ3-b: Do the culture groups stress ego-networks to outgroup members to form

cooperative interactions and to be convergent in MNOS?

In Phase Four, the research focuses on the total score of F INT scale. This scale
was designed to measure individuals’ communication frameworks in the context of
cooperative relationships (Johnson, 1997a). The total score of FINT indicates one’s
cooperative interactions with others. In other words, when the total FINT score
increases, cooperative interactions increase. This positive association between the FINT
and cooperative interactions might be applicable to communication within MNOS.
Outgroup communication is a key to success in MNOS. The more outgroup
communication in two culture groups, the more cooperatively the organizational
members interact each other within a MNO. On the other hand, the more ingroup
communication within a cultural group, the less the organizational members
cooperatively exist within a MNO. Therefore, by relating it with outgroup

communication, the total F INT score will predict the level of cooperation within MNOS.

41

[Hypothesis 3]
H3: The total score of the FINT scale will be positively associated with outgroup

communication and negatively associated with ingroup communication across two

cultural groups.

42

CHAPTER 2

METHODS

Measurement

In order to quantitatively examine the research questions and hypotheses above, it
is necessary to use the appropriate instruments to measure communication frames and
ingroup-outgroup communications. The F INT scale (Johnson, 1997a) and a social
network questionnaire are used in the present research. In addition, the other
methodological issues of research design, sampling, and procedures of data collection are
described in the later section.
Frame for Interaction (FINT) Scale

Johnson (1997a) designed an instrument “to explore the concept of framework for
interactions and to systematically relate it to a variety of established constructs” (p. 131).
For present purposes, it is necessary to compare the communication frames among
different cultures. This instrument meets my research design needs although it is
relatively new and has few replicated studies. There are a number of literatures which
address the importance of ﬂameworks, but there is no speciﬁc instrument to measure
employees’ frameworks and cognitive structures in organizational settings. In these
regards, the FINT scale is the best available instrument and the most effective approach
to accomplish my research objectives.

Johnson (1997a) redesigned the FINT scales since his previous works (Johnson,
1992). According to him, “the results of these early studies were quite encouraging for
this initial stage of scale development” (Johnson, 1997, p. 131). As I illustrated earlier,

the FINT scale is composed of ﬁve subscales. Johnson’s initial works resulted in

43

unidimensionality in terms of content homogeneity, internal consistency, and parallelism.
Also, it has good reliability in terms of Cronbach’s alpha, and its construct validity was
assessed positively. Recent Johnson (1997a) conducted qualitative interviews of key
organizational members and improved the instrument, analyzing the same statistical
criteria. This latest version of the FINT scale has 23 items (four items for exchange and
negotiated order dimensions, ﬁve items for formal and normative dimensions, and three
items for sentiments dimension), measured by an eleven-point Likert-type scale. Some
items did not meet the criteria for acceptable psychometric properties, and they were
dropped ﬂom the ﬁnal FINT model. Eventually, it had relatively strong reliabilities in
Cronbach’s alpha as follows; exchange (4 items) = .85, formal (4 items) = .78, normative
(4 items) = .92, sentiments (3 items) = .76, and negotiated order (4 items) = .88.
Conﬁrrnatory Factor Analysis (CF A) was used to examine unidimensionality,
parallelism, and internal consistency. The 19-item FINT scale met these criteria.

Thus, the latest version of the F INT scale has many advantages for this research.
It has face validity, and it is statistically reliable and valid. It is realistic and feasible: the
number of items is reasonable to ask in business organizations. Most importantly,
selecting an instrument must meet the objective of the research. The F INT scale is useful
to cross-culturally examine communication frames, so the FINT scale is used for the
present research.

When the items were selected for the present research, I attempted to use all of the
original 23 items of the FINT scale since the use of many items can reduce error and
create better statistical reliability. I reviewed all of them and found two items of the

sentiment dimension and one item of the formal dimension that were not appropriate for

44

the present research. The major reasons for this decision were the difﬁculty of translation
and their unclear meaning in Japanese. These three were also dropped in the ﬁnal FINT
model of Johnson’s (1997a) study. Twenty items of the F INT scale were selected ﬂom
Johnson’s 23 items. I pilot-tested these 20 items of the FINT scale on 50 subjects
(Americans and Japanese). Since there was no critique and problem reported on these
items in the pilot-test, the 20-item version of the FINT scale is recognized as satisfactory.

In the next step, the F INT scale for an English version was translated into
Japanese. A bilingual Japanese ﬁrst translated it into Japanese. Then, two bilingual
Japanese who are Ph.D. students back-translated the instrument. They translated it to
English, and these results of back-translation were compared with the original English
questionnaire. Then, the two translators and I discussed each term and decided the most
appropriate expression for every single item. Thus, the 20-item version of the F INT scale
was prepared for both English and Japanese versions. Seven-point Likert-type scales
were used in each version. The participants were asked to answer twenty items of
communication ﬂameworks in the context of cooperative relationship (see the examples
of the questionnaires in Appendix C). Each subscale score was calculated by a statistical
mean of subscale items, and the total FINT score was calculated by summing the ﬁve
dimensions.
SoLial Network Questionnaire

Since the communication patterns of the ingroup and outgroup rely on an
individual’s networks toward other ingroup members and the outgroup members, the
present research uses radial network data, which is a focal network composed of each

individual’s overall pattern of relationship with others (Chang, et al., 1997). In order to

45

investigate ego-network links to ingroup and outgroup members, network analysis was
conducted. I measured how many communication ties with Americans and Japanese an
individual had in a certain period. On the basis of ego-network data, the relative
emphases on the ingroup and the outgroup communication was calculated. It is
expressed as the ratio of the total ego-network to the ingroup members and the outgroup
members. In other words, they are expressed by a percentage of ingroup ties out of total
ego-networks and a percentage of outgroup ties out of total ego-networks. For example,
if the ratio of ego-networks to the ingroup members is 50 %, it indicates equal strength
between ingroup and outgroup relationships. If 70 % appear in the ratio of the ego-
network to ingroup members, then 30 % of ties are to outgroup members, indicating that
ingroup communication is a larger part of a total ego-network and more emphasized than
the outgroup communication.

In order to obtain the ego-network data, a self-reported questionnaire was
developed to ask with whom the respondents had communication contacts, including
face-to-face, telephone, and e-mail communication channels. The respondents were also
asked to report the frequency of communication contacts with these organizational
members. In addition, the types of communication were deﬁned as three kinds of
communication contents. They are articulated with examples on the instructions of the
questionnaire, and the names of network ties and the ﬂequencies of contact are asked in
the following three content areas:

(1) J ob-related: Technical, performance, and other information directly relevant to

the current jobs.

46

(2) Organization-related: Information regarding the organization, NOT directly

related the current jobs.

(3) Social: Informal social activities and conversational topics regardless of the

current jobs.

These three domains of communication are often used to collect social network
data from work organizations (Suzuki, 1997, 1998; Salk & Brannen, 2000). Suzuki
(1997, 1998) calls them the task speciﬁc, the task general, and the non-task
communications. Salk and Brannen (2000) call them the task-related, advice-related, and
private communication. They represent the most formal communication types, the
formal-but-casual communication type, and the informal communication type
respectively. Formal and informal communication structures are quite important in
organizational settings (Hartman & Johnson, 1990; Ibarra, 1997, 1993, 1992). In
particular, Japanese organizational behaviors are considered different in formal and
informal settings. These three distinct types of networks in the present study also
represent the degree of formality in communication contexts. These distinctions are
effective to identify different communication styles between Americans and Japanese in
terms of ingroup and outgroup relationships.

The social network questionnaire was initially developed in an English version,
and I conducted a pilot-test for this initial version on 50 subjects (Americans and
Japanese). There were few comments back to the researcher. They were related to the
difﬁculty of counting the frequency of contacts. Extra explanations were added and
examples were more reﬁned on the instruction sheet. The revised version was also

translated into Japanese by a bilingual Japanese. Then, two bilingual Japanese who are

47

Ph.D. students back-translated the questionnaire same as the procedure in the FINT scale.
In the ﬁnal stage of back-translation, the two translators and I carefully discussed the
instructions and examples of the social network questionnaire and decided the most
appropriate expression for the ﬁnal version. Eventually, as well as the F INT scale, the
social network questionnaire was prepared for both English and Japanese versions (see
the examples in Appendix C).
Demographic Questions

In addition to two major research questionnaires, basic demographic questions
were asked in the end of the questionnaire sheet. The items of demographic questions are
(1) gender, (2) respondent’s formal position in the company, (3) respondent’s years of
services in the research site, (4) languages that the respondent can speak ﬂuently, and (5)
respondent’s ethnic background. The identiﬁcation of either Japanese or U.S. American
is decided on the basis of the question of ethnicity. Participants chose their ethnic groups
ﬂom six categories and reported their nationality. If one’s ethnicity is more complicated

than the six categories, she/he is asked to tell her/his ethnicity in a blank.

Research Design
The present research aims to quantitatively analyze the cognitive and the
behavioral dimensions of communication between Americans and Japanese in the MNOS
and to identify distinctive patterns of organizational communication in terms of the
effectiveness of MN Os. For this end, the measurements of three constructs (cultural

difference, communication ﬂame, and ingroup-outgroup communication) are carefully

48

analyzed with regard to two concepts, the norms of cross-cultural stereotypes and the
concept of convergent-divergent.

The FINT scale is used to measure individual communication ﬂames and to
analyze the cognitive level of communication. There are ﬁve FINT subscales which are
assumed as unidimensional respectively, and this is considered as a ﬁrst order
unidimensionality. Also, the FINT scale is a two-dimensional model, so ﬁve subscales
compose the FINT scale. Namely, the FINT has a second order unidimensionality of a
ﬁve-factor model.

A ﬁrst step is to calculate the subscale scores of individual participants. They are
calculated by a mean of the items in each F INT dimension. Also, the total score of FINT
is calculated by summing the scores of ﬁve subscales. Thus, individual participants have
the score of ﬁve FINT subscales and the total score of FINT. Each item is assessed by
ﬂequencies and basic descriptive statistics. Each subscale is assessed by reliability of
Cronbach’s alpha and factor loading of conﬁrmatory factor analysis. The ﬁrst order
unidimensionality for each subscale and the second order unidimensionality of a ﬁve-
factor model are assessed by internal consistency and parallelism of conﬁrmatory factor
analysis: observing the residuals created by predicted and actual correlations. These
processes are necessary to obtain the effective statistical outcome and these results are
described as the section of psychometric property in the next chapter. Then, the produced
scores of FINT subscales are analyzed with regard to cultural difference, and t-test and
MANOVA are examined to identify the inﬂuence of national culture on communication

ﬂameworks.

49

Ego-networks to Americans and Japanese are measured to analyze ingroup-
outgroup communications as the behavioral level of communication. Participants are
asked to report the names of their three types of communication networks (job-oriented,
organization-related, and social networks), and the reported names are classiﬁed into
either Americans or Japanese. The researcher carefully counted how many ties the
participants had communication with Americans and communication with Japanese. It
was double-checked by the research assistant. In this process, we obtained ego-network
data in terms of the number of communication links with Americans, the number of
communication links with Japanese, and the total number of ego-networks (T) in three
communication types. They were transformed to the number of ego-networks with
ingroup members (E1) and the number of ego-networks with outgroup members (E0) for
both American and Japanese data. Mathematically, a sum of E1 and E0 is equal to T.

The ratio (percent) of ingroup ties is calculated by the number of ego-networks
with ingroup members divided by the total number of ego-networks, individual
percentage of ingroup ties (IGi) is expressed as follows.

IG, = E], / T,
The individual ratio of outgroup ties (0G,) is calculated in the same way, and it expresses
as follows.
OG; = E0, / T,-
A sum of percentages of ingroup ties and outgroup ties are equal to 1.00.
1G; + OG; = 1.00
A statistical mean and a standard deviation are calculated in the number of

ingroup ties and the number of outgroup ties for Americans and Japanese. Moreover, the

50

same statistics are calculated in a percent of ingroup ties and a percent of outgroup ties
for American and Japanese. Then, t-test and MANOVA are examined to identify the
inﬂuence of national culture and the pattern of intergroup communication.

The results of FINT subscales and ego-networks to ingroup and outgroup
members are analyzed with regard to the norms of cross-cultural stereotypes, and they are
assessed in terms of either convergence or divergence. Conformity among ﬁve F INT
subscales is analyzed between the American data and the Japanese data. Also, the
present research design assumes that the emphasis on outgroup communication indicates
convergence in communicative interactions in MNOS, and it is assessed in the patterns of
intergroup communications.

Lastly, correlation between the total score of F INT and the percent of outgroup
communication are inquired in the American data and the Japanese data. The total score
of F INT is designed to indicate cooperative context, and increase of outgroup
communication indicates higher degree of cooperation in MNOS. Therefore, the present
research design seeks to ﬁnd ‘positive’ correlation between the total FINT score and
outgroup communication.

In order to statistically analyze the research design above, SPSS version 10.0 is
mainly used in the present research. Conﬁrrnatory factor analysis is executed by

computer program, CFA (Hamilton & Hunter, 1988).

51

Data Collection
Sampling

Currently, there are about 3,700 J apanese-owned companies in the U.S. Detroit
area has about 200 Japanese companies, members of the Japanese Business Society of
Detroit (JBSD). The researcher decided to collect data ﬂom these Japanese companies in
Detroit through the assistance of the JBSD.

In this sense, the sampling method in the present study is not probability sampling
but purposeful sampling. The purpose of probability sampling is generalization.
Therefore, the logic and power of probability sampling depends on a truly random sample
that will permit conﬁdent generalization ﬂom the sample to a larger population (Patton,
1990). However, “when time and cost considerations severely limits the size of sample,
the judgment sample is often preferable” (Anderson, 1987, p. 167). Purposeful sampling
is characterized by “the use of j udgement and a deliberate effort to obtain the
representative samples by including presumably typical areas or groups in the sample”
(Kerlinger, 1986, p. 120). The logic and power of purposeful sampling lies in selecting
information-rich research sites for study.

Most Japanese establishments in Detroit are engaged in auto-related businesses, a
primary industry in the world. The American Big Three automotive companies and most
Japanese automotive companies have research and development centers in Detroit. Since
late 1980’s, many Japanese transplants have been established and have started their
operations around this area (Cutcher-Gershenfeld, et al., 1998). Thus, J apanese-owned
companies in the Detroit area are recognized as a typical case of MNOS. They were also

convenient research sites.

52

Procedure of Datﬂollection
The JBSD provided a list of 200 J apanese-owned companies. Fifty companies
were selected ﬂom the 200 based on organizational size and the demographic proportion
between American and Japanese workforce. Since many member companies have a few
Japanese employees working with many Americans, these were rejected as unsuitable to
observe subculture conﬁguration in the MNOS. Also, in order to control organization
size, the study did not choose small branch ofﬁces (ﬁve to 10 workers) and huge
companies that have a big assembly plant (more than 150 employees). A majority of
moderate-size organizations were selected, those in which the number of employees is 30
to 100. The researcher communicated and negotiated with each company by mail and
telephone over seven months since November 1998.
Eventually, nine companies were contacted to provide access to their employees.
These nine companies accepted the request for data collection regarding the F INT
scale. Six of these nine companies (Company A, B, C, D, E, and I) cooperated to
collect social network data. The data collection was continued ﬂom November 1998
to May 1999, and each research site had three weeks to administer the questionnaire.
The F IN T and social network questionnaire was distributed to all organizational
members (including managers and executives) in seven companies and also distributed to
all members of the selected departments in the two largest companies (Company H and
F). The English version of the questionnaire was distributed to the American personnel,
and the Japanese version was distributed to Japanese. Individual participants had choices
of where they ﬁll out the questionnaire: either they can take it home and ﬁll it out or they

ﬁll it out in the ofﬁce. There was no time limit to ﬁll out the questionnaire. Since an

53

envelope was provided to the participants, they enclosed a response to a given envelope.
Then, the participants returned their sealed envelopes to a person who administers this
data collection in the research sites. However, the participants had another way to
directly return their answers to the researcher.

Before data collection, the present research design and objectives with actual
questionnaire were evaluated by the University Committee on Research Involving
Human Subject (UCRIHS), and the present research was approved by UCRIHS on
October 29, 1998: IRB# 98651, Category I-C. The procedure of data collection was
evaluated as fair and appropriate. For example, it was promised and articulated that
participants’ responses and privacy kept conﬁdential ﬂom the companies and no personal
information was disclosed to anyone other than the researcher (refer to the sample
questionnaire of Appendix C). Research consent was made by ﬁlling in their name on
the response sheet. Moreover, if some participants requested to see the results of this
research, the summary was supposed to be provided to them. As a result, data collection
was completed in May 1999.

One of most difﬁcult part in the entire research process was data collection ﬂom
business organizations. I spent one year for contacting and negotiating them. Good
research in organizations requires a lot of efforts for data collection. This is not a
convenient sampling, such as data collection ﬂom college students. Data ﬂom real

business organizations provides rich information to us.

54

CHAPTER 3

RESULTS

Description of Sample
The total number of participants who responded to the F INT questionnaire was
152: 89 Americans and 63 Japanese. Sixty-six of these participants (41 Americans and

25 Japanese) reported the social network questionnaire as well.

As Table 2 shows, all the nine companies are engaged in automotive-related
business. Three companies are international trading (Company B, D, and G), and the
others are modest sized research and development (R&D) ﬁrms with small
manufacturing operations. The results of the demographic information are summarized
in Table 3. Since many Japanese ﬁrms in Detroit have localized management to reduce
their Japanese workforce, the proportion of Japanese to Americans is about 40:60. The
workforce in the nine ﬁrms is three fourth male. The proportion between management

and employees is 45:55. The average of years of services is 3.8 years.

The response rates in terms of organizations were varied for this data collection:

ﬂom 30 to 50 percents. These might be problematic in some techniques of network

55

analysis which require the higher response rate (Wasserrnan & Faust, 1994). However,
the present research focuses on individual radial networks, so it is considered that these
response rates do not cause a serious problem. More importantly, as I discuss later, it is
very hard to obtain high response rate for social network data and it is necessary to
improve it for the future research.

In short, the present research was conducted on nine J apanese-owned companies
in Detroit, which are engaged in auto-related business. All the data were appropriately
administered and processed by the researcher. One hundred ﬁfty-two participants
responded to the questionnaire of FINT scale: 89 Americans and 63 Japanese. Sixty-six
of them reported their ego-networks to the ingroup and outgroup members: 41 Americans
and 25 Japanese.

Their nationalities were identiﬁed by their answers of ethnic background, which is
a part of questionnaire. This question prepares six categories of ethnic groups. However,
interestingly, most answers were either Caucasian American or Japanese. There are a
couple of Aﬁ'ican American and Hispanic Americans and one Native American.
Therefore, in the present study, they are integrated in the U.S. Americans, and they are
called the American data while the responses from Japanese participants are called the
Japanese data. The factor of cultural difference is categorized as Americans or Japanese

in the following analysis.
Psychometric Properties of FINT Scale

The analysis of the FINT subscales must proceed in psychometric properties.

First, the FINT subscales are examined in both exploratory factor analysis (EF A) and

56

conﬁrmatory factor analysis (CFA). While the former is beneﬁcial to detect the number
of factor components, the latter is aimed to assess the validity of a factor model. EFA
was initially executed through the extraction method of the principal component analysis
with regard to twenty items, and the rotated component matrix produces ﬁve factor
components, which are exactly identical to the ﬁve dimensions of the FINT model (see

Table 4).

At the same time, CF A supports the ﬁve-factor model of the F INT scale. The
correlation matrix of the twenty F INT items (see Table 5) was used to produce factor
loadings, which express the correlation between each item and its construct. CFA shows
that all the items meet the criteria of face validity: a primary factor loading of .4 or
greater (Johnson, 1997a). CFA also provides the theorems of internal consistency and
parallelism. The residuals are calculated by discrepancies between the predicted and the
observed correlations, and the assessment of internal consistency and parallelism resulted
in the support of unidimensionality in each subscale and the support of a ﬁve—factor

model of the FINT scale.

Moreover, the FINT scale was originally designed as a multidimensional model,

and second order factor analysis (Hunter & Gerbing, 1982) was conducted on the data.

57

The ﬁrst order unidimensionality was identiﬁed in each subscale. Next it was necessary
to assess the second order unidimensionality on the ﬁve subscales and identify a
multidimensional model of the FINT scale. The same procedure as the ﬁrst order factor
analysis was executed on ﬁve subscales, and the results support the second order
unidimensionality for the FINT scale. Thus, the FINT was recognized as validity
measuring communication ﬂameworks, which is a two-dimensional and ﬁve-factor
model.

In addition to the factor analyses and consideration of face validity, the F INT
subscales must be assessed in reliability tests. Reliabilities for the ﬁve subscales were
analyzed on the sample (n=152). Cronbach’s alpha (or) were yielded for reliabilities of
these ﬁve dimensions among twenty items (refer to Table 6), and they are generally
satisfactory since most of them are more than .80, except the exchange ﬂamework. The
exchange ﬂamework had four items and Cronbach’s alpha was .66. Weak items were
deleted and alphas were recalculated on three items of the exchange dimension.
However, none of them could exceed .66 of the initial exchange subscale. The exchange
subscale of the present research is identical to that of Johnson’s, which yielded reliability
alpha of .85. Therefore, these differentials are not caused by the items themselves.
Rather, they may be caused by the procedural errors, such as differences between English

and Japanese versions.

58

In contrast, reliabilities of the other dimensions are quite high. The formal ﬂame
has ﬁve items and its reliability alpha yields .81, which is slightly higher than Johnson’s
(1997a) study (or = .78). Similarly, the normative framework which is composed of four
items produces .89, and it is slightly lower than that of Johnson’s results (.92).
Reliability of the sentiments framework is .92, which is slightly higher than Johnson’s
(.88). That of the negotiated order framework is .82, which is also higher than Johnson’s
(.76). Thus, the F INT subscales have a satisfactory level of reliability and validity, and

these strong psychometric properties enhance statistical outcomes in the present research.

Results of Phase One:
Cultural Difference and FINT Subscales

After the assessment of the F INT scale, the inﬂuence of cultural difference is
analyzed. Table 7 shows the mean comparison between American and Japanese samples
in terms of ﬁve FINT subscales and the total score of F INT. Also, this mean comparison
is visually illustrated on Figure 1. Obviously, Americans are higher in every dimension
and the total of the F INT scale. For example, Americans have 5.72 (The Italic number is
Standard Deviation: .88) while Japanese have 5.32 (.80) in the mean of the exchange
framework. In addition, the sentiment ﬂamework and the negotiated order ﬂamework
have signiﬁcant differences in their means; 4.27 (1.52) in American and 3.52 (1.66) in
Japanese samples in the sentiments ﬂamework; 5.33 (1.18) in American and 4.56 (1.32)
in Japanese samples in the negotiated order. The total of F INT score has signiﬁcant

difference between American and Japanese samples; 24.48 (4.21) in Americans and

59

22.11 (3.42) in Japanese. Americans are slightly higher in the formal and normative

frameworks than Japanese.

The t-test assesses the signiﬁcance of difference between two population means
(p1 — “2) based on data derived ﬂom two samples, of which at least one sample is small
(Smith, 1988). As Table 7 shows, the t-test indicates signiﬁcant difference on the
exchange (t = 2.927), the sentiments (t = 2.811), the negotiated order (t = 3.707) and the

total FINT (t = 3.805) in the signiﬁcant level of .001.

In addition, MAN OVA is effective to diagnose how cultural differences between
Americans and Japanese are statistically signiﬁcant on multiple variables of FINT
subscales (refer to Table 8). The multivariate tests of signiﬁcance (S = 1, M = 1 1/2, N =
71 1/2) show that cultural difference has signiﬁcant effect on F INT subscales with .001
level. F -test Shows statistical signiﬁcance on the exchange ﬂamework; F (1, 149) =
7.934, p < .01; on the sentiments ﬂamework; F(1, 149) = 7.624, p < .01; and on the
negotiated order ﬂamework; F (1, 149) = 14.328, p < .001. Thus, three of ﬁve FINT
subscales reach a statistically signiﬁcant level; the formal and normative frameworks fail

to reach a signiﬁcant level.

60

These results can answer Research Question 1 and Hypotheses 1. The ﬁrst
question is whether there are the cultural differences in the F INT subscales. The data
show signiﬁcant differences between Americans and Japanese on the FINT subscales.
Some of these differences answer to the expectation of the hypotheses. For example, it is
hypothesized that Americans are higher in the exchange and negotiated order framework.
The data support these hypotheses: Hl-a and Hl-d. However, the data do not answer the
other hypotheses; Japanese are higher in the formal and normative frameworks: Hl-b and
Hl-c. Since t-test and F -test do not result in a statistically signiﬁcant level in these
ﬂameworks, the data show neither the opposite evidence nor the supportive evidence to
these hypotheses. It is necessary to further consider Japanese emphasis on formal and
normative ﬂameworks. Thus, the data show cultural difference on some of FINT
subscales and the total of FIN T scale. In average, Americans are generally higher in

every FINT subscale and the data partly support the hypotheses.

Results of Phase Two:
Cultural Difference and Ego-networks to Ingroup and Outgroup
Sixty-six participants reported their ego-networks to the ingroup and outgroup
members. They indicated the names with whom they had communication networks in
terms of job—oriented, organizational, and social communication, so the ego-networks are

obtained in three communication types. A statistical mean and a standard deviation are

61

calculated in the number of ingroup ties and the number of outgroup ties for American
and Japanese samples. Similarly, the same basic statistics are calculated in a percentage
of ingroup ties and a percentage of outgroup ties for both samples. The results are

summarized in Table 9.

Before detailing these results, I would like to look at the outcomes of total ego-
networks in three communication types (see Figure 2, 3, and 4). Three histograms show
the features of three communication types. Figure 2 shows a pattern of ego-networks in
job-oriented communication. The histogram illustrates a bell shape, the ﬂequencies of
many individuals are ranged between 5 to 25. There are few in the range of less than 5
and more than 25. The mean of ego-networks in job-oriented communication is 15.0
(Standard Deviation = 9.43). In contrast, the bell shape of ego-networks in organization-
related communication is more skewed toward the range of 5 and 15 (see Figure 3). It
implies that very few people have more than 20 ties of organization-related networks;
mean is 8.9 (8.62). Interestingly, the mean in the ego-networks of social communication
is larger that of organization-related communication: mean = 11.5 (10.49). However,
there are many people who have no network ties although there are some people who
have more than 20 ties in social communication. It implies that ego-networks are varied
in social communication. Thus, three histograms show different patterns of ego-networks

in three communication types, and it indicates the evidence of their construct validity.

62

Additionally, the overall results of ego-networks show a large amount of standard
deviation, and the present research has to take it into account. Outliers are not detected
from the data. It might be recommended to normalize the original data in order to get rid
of large standard deviations. The normalization was transformed by means and standard

deviations, and the obtained normalized scores are used to supplement the original data.

Now, the ego-networks with ingroup and outgroup are illustrated with regard to
cultural difference in the three communication types. Again, histograms are useful to
look at general tendencies of ego-networks, and therefore four histograms create a set of

a ﬁgure in three communication types in order to compare the ego-network patterns.

Ego-networks in Job-oriented Communication

Histograms in Figure 5 illustrate the number of ingroup and outgroup ties in the
American sample and the Japanese sample in job-oriented communication. Ego-
networks to ingroup members in Americans are varied in the range ﬂom zero to thirty
while ego-networks to ingroup in Japanese narrowly range ﬂom zero to ﬁfteen. The
mean of ego-networks to ingroup is 10.5 (7. 79) in Americans and 6.84 (3.26) in Japanese,
so there are a large difference in a statistical mean. The normalization scores similarly

show obvious differences in ingroup communication of job-oriented network: .51 in

63

Americans and .39 in Japanese. The t-test is used to detect statistical signiﬁcance in
these differences (see Table 9). It ﬁnds statistical signiﬁcance in the .05 level in both the
raw score and normalization score.

The other two histograms in Figure 5 show cultural differences in the number of
ego-networks to outgroup members. The histogram in Americans is very skewed toward
the range ﬂom zero to ten whereas the histogram in Japanese is widely varied ﬂom ﬁve
to twenty. These histogram shows an obvious difference, and a mean in the number of
outgroup ties is 3.63 (2.94) in the American respondents and 10.24 (5.93) in the Japanese
respondents. Similarly, the normalization scores Show differences of outgroup
communication: .33 in Americans and .68 in Japanese. The t-test recognizes these
differences as statistical signiﬁcance in the .001 level in both the raw score and
normalization score.

A mean of ego-network ratio to ingroup ties is .71 (.19) in Americans and .41
(.13) in Japanese. This indicates a comparison between ingroup and outgroup
communication. In average, Americans has 71 percents of ingroup ties whereas Japanese
have 41 percent of ingroup ties. As the t-test shows statistical signiﬁcance in the .001
level, the percentage of ingroup ties are signiﬁcantly different in culture groups. Namely,
Americans emphasize ingroup communication more than Japanese.

Automatically the ratio of Japanese ego-networks to outgroup members is higher
than that of Japanese ego-networks to ingroup member (59: 41). Surprisingly, Americans
have more distinctive patterns in ingroup and outgroup communication. A proportion of
ingroup ties to outgroup ties in the American sample is a ratio of 71 to 21 in average. It

implies that Americans tend to stress ingroup communication in job-oriented networks.

64

MANOVA takes into account of correlation among multiple variables, and it is
used to identify the effect of cultural difference on ego-networks to ingroup and outgroup
members in three communication types. All multivariate tests identify the effect of
cultural difference in signiﬁcance level of .001 (see Table 10). When F-test is
individually examined, it does not Show statistical signiﬁcant in F-test; F (l, 64) = 3.817,
significance = .055 in ingroup communication. However, F—test shows signiﬁcance of
cultural difference in the number of outgroup ties; F (1, 64) = 36.022, p < .001. These

results in the raw score are identical to MANOVA outcomes in normalization scores.

When the ratio of ego-networks to ingroup members and that of ego-networks to
outgroup members are independently examined in three communication types, all
multivariate tests ﬁnd the cultural effect on the percents of ingroup-outgroup
communication in signiﬁcance level of .001. F -test shows statistical signiﬁcance; F (l,
64) = 46.340, p < .001 in both the percent of ingroup communication and the percent of
outgroup communication. Thus Japanese emphasize outgroup communication more than
Americans.

These results answer Research Question 2. In terms of j ob-oriented

communication, there are culturally signiﬁcant differences in ingroup and outgroup

65

communication. However, these cultural differences are completely opposite to the
hypotheses in the previous chapter, and three hypotheses (HZ—a, H2-b, and HZ-c) are not
supported in the job-oriented communication. Japanese do not emphasize ingroup
communication and Americans emphasize ingroup communication. Japanese emphasize
outgroup communication rather than Americans in job-oriented communication.
Ego-networks in Orgamization-thrted Communication

Next, the results of ego-networks in organization-related communication are
focused in terms of the number of ties and the ratios. First, when the histograms of
Figure 6 is compared with those of Figure 5, they are quite similar. This indicates that
ego-network patterns are similar between the job-oriented and the organization-related
networks. American ego-networks to ingroup members are varied ﬂom zero to thirty
while Japanese ego-networks to ingroup members narrowly range ﬂom zero to ten. A
mean of ingroup ties is 5.93 (6.35) in Americans and 4.80 (3.81) in Japanese. A mean of
normalization score is .47 in Americans and .45 in Japanese. Americans and Japanese
are less different in the ego-network patterns of ingroup ties, and the t-test does not show

statistical signiﬁcance in both the raw data and the normalization scores.

Outgroup communication in Figure 6 are also similar to that in Figure 5. There
are very few ego-networks to outgroup members and they mostly range ﬂom zero to ﬁve
in American sample. In contrast, Japanese ego-networks to outgroup members are varied

ﬂom zero to more than twenty. A mean of outgroup ties is 2.07 (2.71) in American

66

sample and 5.56 (6.23) in Japanese sample. A mean of normalization scores is .39 in
Americans and .56 in Japanese. In average, Japanese have higher number of ego-
networks to outgroup members than Americans. The t-test shows signiﬁcant difference
between Americans and Japanese in outgroup communication in organization-related
networks (p < .01 ).

In contrast with the number of ego-networks, the ratio of ego-networks to ingroup
members are signiﬁcantly difference in cultural groups. A ratio of ego-networks to
ingroup members is .69 (.31) in Americans and .52 (.27) in Japanese. A ratio of ingroup
ties to outgroup ties is 69: 26 in Americans and 52:44 in Japanese. Americans more
emphasize ingroup relationships than Japanese. Japanese do not have strong distinction
between ingroup and outgroup relationship.

MAN OVA ﬁnds cultural difference on ego-networks, but when F -test
individually examines the cultural effect on ingroup and outgroup communications in
organization-related network, it does not Show signiﬁcant difference on the number of
ingroup ties. However, the number of ego-networks to outgroup members has signiﬁcant
difference in cultural groups; F (1, 64) = 9.865, p < .01. These MANOVA results of the
raw data are the same as that of the normalization score. In the ratio of ingroup-outgroup
communication, F -test ﬁnds statistical signiﬁcance for the cultural difference; F (l, 64) =
5.046, p < .05. in the percent of ingroup ties, and F(1, 64) = 6.706, p < . 01 in the percent
of outgroup ties.

These results answer the research questions and hypotheses. Similar to the results
ofjob-oriented communication, the data in organization-related communication also

show cultural difference in ingroup and outgroup communications, and these results

67

oppose the hypotheses, H2-b and H2-c. Americans clearly emphasize ingroup
relationship in this data set, and it is more than Japanese. The difference ﬂom a data set
ofjob-oriented communication is that Japanese have more ego-networks to the ingroup
members than ego-networks to the outgroup members; a ratio of ingroup ties to outgroup
ties is 52:44 in average. This result may support one of the hypotheses (HZ-a); Japanese
have a higher ratio in ingroup ties than outgroup ties in informal communication.
Organization-related communication is less formal than job-oriented communication, and
therefore this result might be one example of Japanese preference to ingroup relationships
in informal settings.

Ego-networks in Social Communication

In the histograms of social communication (see Figure 7), American ingroup ties
vary from zero to 25 while Japanese ingroup ties are limited ﬂom zero to 15. A mean of
ego-networks to ingroup members is 8.15 (8.01) in Americans and 6.76 (3.81) in
Japanese. A mean of the normalization score is .49 in Americans and .46 in Japanese.
Americans are slightly higher in ingroup communication, but it is not supported by a
signiﬁcant t-test.

On the contrary, outgroup ties in social communication have signiﬁcant
difference in two culture groups. A mean of ego-networks to outgroup members is 2.15
(2.72) in Americans and 6.64 (8.82) in Japanese. Although both cultural groups have
many zero links in outgroup ties, Japanese have more outgroup communication in
average. The normalization scores also indicate that Americans is lower than Japanese in

outgroup communication: .35 in Americans and .65 in Japanese. These are supported by

68

a signiﬁcant t-test (p < .05). Cultural difference is identiﬁed in outgroup communication,

not in ingroup communication.

In a percent of ingroup ties, a mean is .70 (.35) in Americans and .62 (.32) in
Japanese. An average percent of outgroup ties is .13 (.14) in Americans and .34 (.31 ) in
Japanese. The ratio of ingroup to outgroup communication is 70:13 in Americans and
62:34 in Japanese. The t-test shows statistical signiﬁcance with the .01 level on an
average percent of outgroup communication.

In MANOVA outcomes, F-test does not show statistical signiﬁcance on ingroup
communication, but it shows statistical signiﬁcance of cultural difference on the number
of outgroup ties: F(1, 64) = 9.280, p < .01 and on the percent of outgroup
communication: F(1, 64) = 15.149,p < .01.

Similar to the other communication types, although cultural differences are
identiﬁed in outgroup communication, the data set of social communication does not
support H2-b and H2-c. However, interestingly, this data set support H2-a: Japanese
emphasize more ingroup communication than outgroup communication in informal
settings. In Japanese sample, a ratio of ingroup ties to outgroup ties is 62:34.
Ego-networks in Overall Comunﬂtion

Although the research questions and hypotheses are not directly related, it is
beneﬁcial to look at the patterns of ingroup and outgroup communication in an overall

type of networks, which is obtained by summing three communication types. As Table 9

69

shows, a mean of ingroup ties is 24.12 (17.59) in Americans and 18.40 (9.62) in

. Japanese. Similarly, a mean of the normalization scores is .51 in Americans and .42 in
Japanese in ingroup communication. Americans are slightly higher than Japanese, but
these differences are not supported by signiﬁcant tests.

However, the number of outgroup ties shows signiﬁcant difference in an overall
type of communication. A mean of outgroup ties is 7.85 (7.01) in Americans and 22.40
(15.94) in Japanese. In the normalization scores, a mean of outgroup communication is
.35 in Americans and .65 in Japanese. These differences are supported by t-test of the .01
level. When MANOVA examines the effect of cultural difference on overall ingroup and
outgroup communication, all multivariate tests show statistical signiﬁcant (p < .001). F-
test shows signiﬁcance on outgroup communication: F (1, 64) = 26.083, p < .001. The
same result of MANOVA is obtained in outgroup communication of the normalization
scores: F(1, 64) = 28.066,p < .001.

Thus, in overall communication, Americans emphasize ingroup relationships
more than outgroup relationship in Japanese MNOS, while Japanese emphasize outgroup
relationships more than ingroup relationships. A percent of ingroup to outgroup
communication is 75:25 in Americans and 47:53 in Japanese, and the cultural difference
on a percent of ingroup-outgroup communication is supported by a signiﬁcant t-test (p <
.01).

In sum, the data show a signiﬁcant level of cultural difference in outgroup
communication across three communication types and in ingroup communication in job-
oriented communication. The data oppose the hypotheses (H2-b and H2-c) but support a

hypothesis of Japanese emphasis on ingroup communication in informal settings.

70

Japanese have many outgroup ties and fewer ingroup ties in job-oriented setting but have
larger ingroup ties in organization-related and social communications (informal settings).

Americans have more ingroup ties than outgroup ties in all three communication types.

Results of Phase Three:
Divergence and Convergence in Communication of MNOS

Identiﬁcation of either convergence or divergence of communication in Japanese
MNOS analyzed on the patterns of FINT scores between two cultural groups and their
outgroup communication. I ﬁrst look at the results of F INT dimensions to identify frame
convergence or ﬂame divergence. The highest mean score is the exchange ﬂame (mean
= 5.72), the second highest is the normative ﬂame (5.47), and the third is the negotiated
order ﬂame (5.33) in Americans. The formal and sentiments ﬂameworks are obviously
lower than these three ﬂameworks in Americans. For example, a differential of two
means between the third highest (formal ﬂame) and fourth highest (sentiments ﬂame) is
1.06, and it is recognized as signiﬁcant difference by a paired-sample t-test (p < .001).
Similarly, Japanese have the same results which exchange, normative, and negotiated
order are the ﬁrst, second, and third highest means respectively. A differential of the
third (negotiated order) and the fourth (sentiments) is .34, and it is also recognized as
signiﬁcant difference by a paired-sample t-test; (p < .001).

These results indicate that both Americans and Japanese commonly emphasize the
communication ﬂameworks of the exchange, normative, and negotiated order in their
communicative interactions with others. These emphases of communication ﬂames are

recognized similar or convergent between the communication ﬂames of two culture

71

 

"a

groups. An important point is, as the results of Phase One indicate, that the exchange and
the negotiated order ﬂameworks have signiﬁcant difference between Americans and
Japanese. Therefore, their communication frames are not identical. When a research
question asks whether the patterns of the FINT subscales are divergent, convergent, or
little different (identical) between Americans and Japanese, the data imply that their
emphases of communication frameworks are not identical but convergent in two culture
groups. Two culture groups commonly stress these three ﬂameworks out of ﬁve FINT
dimensions.

The common emphases on exchange, normative, and negotiated order is analyzed
with regard to the expectation of cross-cultural stereotypes. As they are hypothesized,
Americans must be strong in the exchange ﬂamework and the negotiated order
ﬂamework, while Japanese must be strong in the formal ﬂamework and formative
ﬂamework. The results are a mix of these expectations of cross-cultural stereotypes. The
data show that not only Americans but also Japanese have emphases on the exchange and
negotiated order ﬂameworks. This is an example of ﬂame convergent; Japanese
ﬂameworks conform to stereotypical American ﬂameworks.

In contrast, the data shows that both culture groups commonly emphasize the
normative ﬂame, that is supposed be high in Japanese. This is another example of
convergence in communication ﬂamework; American ﬂamework conforms to
stereotypical Japanese ﬂamework. These results indicate that their communication
ﬂameworks are converged upon a mix of two culture groups.

On the other hand, the results of Phase Two directly answer Research Question 3-

b; whether the culture groups stress ego-networks to outgroup members to form

72

cooperative interactions and to be convergent in MN Os? Outgroup communication is a
key indicator for convergence of communicative interactions between two culture groups,
and it is indicated by a percentage of outgroup ties out of total ego-networks. Americans
less stress outgroup relationship and they had more ingroup communication across three
communication types; a percentage of outgroup ties is 29 % in job-oriented, 26 % in
organization-related, 13 % in social communication. In overall communication,
Americans have 75 % of ingroup ties out of total ego-networks. Paired sample t-tests are
conducted between the number of ingroup ties and outgroup ties, and they are statistically
signiﬁcant in all three communication types; I = 6.507, p < .001 in job-oriented; t =
4.760, p < .001 in organization-related, t = 6.337, p < .001 in social communication.
Thus, Americans emphasize not outgroup communication but ingroup communication.

Japanese communication, in contrast, has different patterns, and they emphasize
outgroup communication in Job-oriented communication. A mean of outgroup ties is
22.40 while that of ingroup ties is 18.40: a mean of normalization scores is .65 and .42
respectively, and it is supported by a paired sample t-test on the number of ingroup ties
and outgroup ties; t = 3.536, p < .005. A percentage of outgroup ties is 59 % in job-
oriented context, and therefore Japanese emphasis on outgroup communication is
conﬁrmed in job-oriented context. The data of overall communication support Japanese
emphasis on outgroup communication: a Japanese percent of ingroup ties to outgroup ties
in overall communication is 47:53. A percent of outgroup communication is higher than
that of ingroup communication in Japanese overall communication.

However, in terms of the number of ingroup and outgroup ties, signiﬁcant

differences are not identiﬁed in the other two context: organization-related and social

73

communication. A lower percentage of outgroup ties to ingroup ties appears in
organization-related and social communication. In these regards, Japanese emphasize
outgroup relationships in a formal (job-oriented) communication but less emphasize them
in informal (organization-related and social) communication.

When a research question asks whether the culture groups stress outgroup
communication to form cooperative interactions and to be convergent in MNOS, the data
show that Japanese emphasize outgroup communication in job-oriented context but
Americans do not emphasize it in any context. Thus, convergence of communicative
interactions in intergroup relationships are merely identiﬁed in Japanese MNOS in the
U.S., except Japanese emphasize outgroup communication in the formal setting.
However, communication ﬂame convergence is identiﬁed in the patterns of the FINT

subscales.

Results of Phase Four:
Total FINT Score and Ingroup-Outgroup Communication
The total score of F INT is calculated by summing of ﬁve subscales, and its
correlation with a percentage of outgroup ties is examined in three communication types.
For the Japanese sample, there is no correlation between a total score of FINT and a
percentage of outgroup ties in all three communication types (see Table 14). An
interesting ﬁnding is that the negotiated order ﬂamework has positive correlation with
outgroup ties in social communication; r = .43, p < .05. However, in Japanese data set,

small sample size for ego-network data does not create strong statistical outcomes.

74

On the other hand, for American sample, strong positive correlation can be
identiﬁed in job-oriented and organization-related communications (see Table 13).
Correlation between total F INT score and a percentage of outgroup ties is .57 (p < .01) in
job-oriented, and it is .53 (p < .01) in organization-related communication. Naturally, the
data show negative correlation between total FINT score and ingroup communication; r =
-.57 (p < .01) in job-oriented, and r = -.43 (p < .01) in organization-related
communication.

When ﬁve dimensions of F INT is also examined in correlation with outgroup
communication, four of them (the exchange, formal, normative, and negotiated order
framework) show strong correlation coefﬁcients (ﬂom .30 to .50) in job-oriented and
organization-related communication. The sentiments ﬂamework does not have
correlation in these two types, but it has a signiﬁcant correlation with outgroup ties in
social communication: r = .36, p < .05. Interestingly, only sentiments have correlation
with outgroup communication in social communication and the other ﬂames do not show
strong correlation coefﬁcients in social communication. Also, there is no correlation
between a total score of F INT and a percentage of outgroup ties in social communication.
Thus, for Americans, the total F INT score is positively associated with outgroup
communication in the business settings: i.e. j ob-oriented and organization-related

communications, not in the social setting.

75

CHAPTER 4

DISCUSSIONS

The present study assesses the cognitive aspect and the behavioral aspect of
Japanese MNOS in the U.S. The effect of cultural difference on communication
ﬂameworks is studied as an assessment of cognitive level, and the cultural difference and
ingroup-outgroup communication are discussed as the behavioral level assessment. The
extent of discussion in the present study refer to the unique features and the complex
context of MNOS, which might be different ﬂom the expectation of cross-cultural

stereotypes.

Implications in Cultural Difference and Communication Framework

Critiques of Cultural Stereotypes

The studies in cross-cultural management and intercultural communication have
created the norms of cross-cultural stereotypes between Americans and Japanese.
Typically, Type A and Type J organizations by Ouchi (1981) and individualism-
collectivism by Triandis (1986) and Hofstede (1980) become the typical norms or the
expectation of cross-cultural stereotypes. Their focuses are on the comparisons between
Japanese working at a Japanese organization in Japan and Americans working at an
American organization in the U.S. These perspectives and approaches are useful and
applicable to the study of MNOS, but these cross-cultural stereotypes can not be true in

the context of MNOS.

76

The FIN T scale measures individuals’ ﬂameworks for communicative
interactions with others. The measurement of communication ﬂames in the context of
MNOS opposes the norms of cultural stereotypes developed in many etic studies“. The
exchange ﬂamework is supposed to be high in Americans but low in Japanese. The
results show that the exchange ﬂamework is the highest score in ﬁve dimensions in both
culture groups. It means that the exchange ﬂamework is not only a primary framework
for Americans but also a primary ﬂamework for Japanese in the context of MNOS. Both
Americans and Japanese emphasize the exchange ﬂamework when they communicate
with other workers. In this sense, their major concerns for communication is
maximization of individual interests, and even Japanese stress a communication ﬂame of
Lockean individualism (Johnson, 1997a).

In addition, the second highest score in the FINT subscales is the normative
ﬂamework in both Americans and Japanese. According to the norm of cultural
stereotypes, this ﬂamework is supposed to be high in Japanese and low in Americans.
The normative ﬂamework relies on the culture and value shared in larger collectivities
(Johnson, 1997a), so this must be a primary ﬂamework for Japanese. However, the study
ﬁnds that Americans emphasize organizational norms and values for their communication
ﬂameworks. This also opposes the norm of cultural stereotypes. Similarly, the formal
ﬂamework is supposed to be high in Japanese but low in Americans. However the results
show that both groups do not emphasize this ﬂamework, which is ranked as the lowest in
ﬁve dimensions.

The present research identiﬁes the signiﬁcant difference between Americans and

Japanese in the degree of measurements, but these differences do not support the cross-

77

cultural stereotypes. Both American and Japanese communication ﬂameworks do not
reﬂect the stereotypes studied in the ernic approaches, and they are distinctive patterns in
Japanese MNOS. The context of MNOS is not the same as that of traditional emic
approaches or comparative studies; Americans and Japanese working at MNOS are
contextually different ﬂom Americans in U.S. and Japanese in Japan. Since Americans
and Japanese work together, they have created their distinct ﬂameworks which are
different ﬂom the ﬂameworks deﬁned by their cultural tradition.

Thus, the cross-cultural stereotypes are not true in the context of MN Os, which
may shape the distinctive patterns of communication ﬂameworks in Americans and
Japanese. Intercultural organizational communication (IOC) can not be considered
without context: it is impossible to neutralize context. Therefore, the focus or role of
IOC studies is on the context of MN Os rather than personal attributes. 9
Hybrid Pragmatism in MNOS

The present study ﬁnds the emphases of three communication ﬂameworks
through the measurement of the FIN T subscales. These three are the exchange,
normative, and negotiated order ﬂameworks. In the expectation of cultural stereotypes,
two of them (exchange and negotiated order) are conceived as a major emphasis of
American ﬂameworks, and one of them (normative) is recognized as a core ﬂamework of
Japanese. In these sense, these three ﬂameworks are considered as the hybrid
combination between American and Japanese communication ﬂames, and they are the
primary communication-ﬂameworks working at Japanese MNOS at Detroit.

This hybrid ﬂamework is comparable to Ouchi’s (1980) conceptualizations of

market, bureaucracy, and clan. The exchange ﬂamework underlies independent

78

individuals and their ﬂee competition, and it is relevant to the market control. The
normative ﬂamework underlies the values and norms dominated in organizations, and
this is related to harmonious human relationships in the workplace; it is relevant to the
clan control. Thus, the hybrid communication-ﬂamework emphasizes both
individualistic competition and collectivistic group-harmony and cooperation, and it is
recognized as a combination of the market control and the clan control. Through
emphasizing market control and clan control, MNOS gains more ﬂexibility and more
competitiveness toward global economy"). An important point is that a hybrid
combination of communication ﬂames has economic rationality.

In addition, two culture groups commonly withdraw the sentiments and formal
ﬂameworks, which are related to emotion and bureaucratic control respectively. Under
the highly competitive global market, these two aspects are often considered as
disadvantages for increasing efﬁciency and productivity. Therefore, the primary
communication-ﬂamework reﬂects the organizational context of competition,
cooperation, and ﬂexibility in the MNOS. In other words, this hybrid combination is a
pragmatic outcome that individuals have created under the competitive economic
environment of an automotive industry. In the future study, this line of argument might
be continued on the issues of strong organizational culture (Deal & Kennedy, 1982) and
excellent organizational culture (Peters & Waterman, 1982).

The present study ﬁnds the primary communication-ﬂamework (exchange,
normative, and negotiated order) in Japanese MNOS, which is different ﬂom cultural

stereotypes, and it is recognized as hybrid pragmatismll by Americans and Japanese

79

working together at a competitive auto industry. This results ﬂom their rational decision

and cognitive change toward effective MN Os.

Implications in Cultural Difference and Ingroup-Outgroup Communication

In addition to ﬁndings identiﬁed in the cognitive level, there are interesting
ﬁndings in ingroup-outgroup communication. One of most important ﬁndings is
signiﬁcant cultural difference on ego-networks to ingroup and outgroup members.
Statistical results show a signiﬁcant level of cultural difference in ingroup and outgroup
communications.

Also, there are striking differences ﬂom the expectation of cross-cultural
stereotypes. In the cultural stereotypes, Americans as individualistic culture are supposed
to have no clear distinction between ingroup and outgroup relationships whereas
Japanese as collectivistic culture are supposed to have more ingroup communication than
outgroup communication (Triandis, 1986; Gudykunst, Yoon, & Nishida, 1987).
However, the research results show that Americans have more ingroup communication
than outgroup communication. Japanese have more outgroup ties than ingroup ties in
job—oriented communication. These ﬁndings are different ﬂom the expectation of
cultural stereotypes.

Implication of American Communication Pagterns.

Let’s ﬁrst look at communication patterns of Americans. The research results
indicate that Americans emphasize ingroup relationships more than outgroup
relationships. Americans have more ingroup ties than outgroup ties in all three

communication types, and the percents of ingroup ties and outgroup ties are almost stable

80

across three types. For example, ingroup ties are 71 % in job-oriented, 69 % in
organization-related, and 70 % in social communication. In overall communication, a
ratio of ingroup ties to outgroup ties in Americans is 75:25. There are some
interpretations and implications in American communication patterns.

One plausible argument is language barriers between Americans and Japanese. In
a section of demographic questions in the questionnaire, participants are asked about
languages spoken ﬂuently, and no American can speak Japanese. It means that the
language used in the research sites of Japanese MNOS is English. According to the
research of Kim and Paulk (1994), both culture groups complain about English
proﬁciency and language styles of the counterpart. For example, Americans claim that
Japanese have “misunderstanding of terminology, poor pronunciation, and inadequate
English grammar” (p. 123). Japanese claim that they have to use English, Americans
speak rapidly, and Americans lack intuitive understanding and focus on speaking. These
examples provide the serious language or communication barriers between Americans
and Japanese. Americans are not able to communicate with Japanese smoothly, and
therefore they tend to emphasize ingroup communication. The issue of language barriers
can be applied to Japanese emphases on ingroup communication in informal settings.

Another possible explanation might be organizational politics and power
relationships. “Politics is power in communication” (Conrad & Poole, 1998, p. 250).
According to Conrad and Poole (1998), organizational members can gain power and
control when they 1) are key communicators or gatekeepers in communication networks,
2) occupy formal positions that allow them to distribute legitimate rewards and

punishment, and 3) can obtain access to the symbols of power. There must be such

81

politics and power struggles between Japanese and Americans. In particular, many
Japanese personnel are in managerial positions or relatively high positions in MNOS, so
Americans become more tactical and careful in power relations and organizational
politics. The research of Kim and Paulk (1994) show a number of complaints and
problems in management and human relationships, and American personnel oﬂen quit
their job.

Implication of Japanese Communication Patterns.

Japanese patterns of ingroup and outgroup communication are more complicated
than Americans’. Japanese have more ingroup communication in informal settings
(organization-related and social communication). These results support the hypothesis of
cultural stereotypes; people in collectivistic culture tend emphasize more ingroup than
outgroup relationship. Japanese are expected to have more ingroup communication.

Strong ingroup communication in informal settings can be explained by not only
Japanese collectivism but also Japanese isolation ﬂom their home country. When Suzuki
(1998) ﬁnds a similar result; more ingroup communication than outgroup communication
in task general and non-task communications, she explain that “one possible factor is the
status of Japanese group as a minority in the practicing organizations, particularly in a
foreign country” (p. 176). Salk and Brannen (2000) explain a similar result by saying
Japanese expatriates. Japanese as expatriates think the loss of their relationship with their
ﬂiends and coworkers, and they feel alienated. Thus their Japanese identity is enhanced
in the minority status in the organizations and in the expatriate feeling in foreign life, and

ingroup relationship is emphasized in informal settings.

82

An interesting ﬁnding is that Japanese have more outgroup ties in job-oriented
communication. This phenomenon contradicts the explanation of individualism-
collectivism. However, Gudykunst, Yoon, and Nishida (1987) point out that outgroup
communication is more complicated beyond the hypothesis of individualism-
collectivism. Suzuki also points out the complication of outgroup relationships, and she
suggests that outgroup relationships is more task-related. Japanese stronger outgroup
communication in job-oriented communication is identiﬁed in Suzuki’s (1998) study.

One plausible argument stems ﬂom the extension or critiques of individualism-
collectivism. Some Japanese scholars, ﬂom the emic point of view, criticize that Japan is
labeled as collectivistic culture and they propose contextualism and relativism (N ishida,
1996). Also, Triandis (1986) distinguishes Japanese “contextual collectivism” and
Korean “simple collectivism.” In their views, Japanese are more ﬂexible in contexts.
The context of job-oriented communication makes Japanese communication patterns
different ﬂom the contexts of social settings. Japanese just concentrate on their tasks and
responsibilities, so they understand that they can not communicate within Japanese
networks and they need outgroup communication with Americans when they execute
their jobs and organizational goals.

In addition, it is necessary to consider a fact of the organizational status in
Japanese personnel. The proportion of managers to non-managers is 17 to 6 in the
Japanese sample and 13 to 28 in the American sample. Many Japanese personnel are
upper than the managerial positions, so they have high responsibilities of communication
with American employees. In these regards, more outgroup communication in job

context might be explained by different organizational positions. However, there is no

83

signiﬁcant difference between managers’ communication and non-managers’
communication in both Americans and Japanese when the factor of corporate status is
controlled. American managers as well as American non-managers have less outgroup
communication whereas both Japanese managers and non-managers have more outgroup
ties in job-oriented communication. For example, when a percentage of ingroup ties is
compared, a mean is .70 (Standard Deviation = .13) in American managers and .72 (.22)
in American non-managers. In the same comparison, it is .43 (.11) in Japanese managers
and .42 (.20) in Japanese non-managers. When a percentage of outgroup ties is
compared, a mean is .30 (.13) in American managers, .28 (. 22) in American non-
managers, .57 (.11) in Japanese managers, and .58 (. 20) in Japanese non-managers.
Japanese managers and non-managers commonly emphasize outgroup communication,
and difference between managerial and non-managerial positions does not inﬂuence the
results of ingroup-outgroup communication patterns. Therefore differences of the
corporate status is not a good explanation.

More plausible explanation is the roles and job characteristics of Japanese
personnel, boundary spanning role in particular. This is related to but different ﬂom a
managerial role. Boundary spanners are “individuals who, while members of one social
system, have links to another” and they are usually recognized as individuals “who have
communication ties to people outside their organization as a result of their formal
organizational position” (Johnson, 1993, p. 147). According to Johnson (1993), a
boundary spanning role can be applied to intergroup relations. Namely, boundary

spanning is communicative behaviors that link different groups within the same

84

organization. In this sense, Japanese outgroup communication with Americans is
recognized as the performance of boundary spanning roles in the MN Os.

Suzuki (1998) suggest importance and necessity of managing “boundaries
between two culture groups in such a way as to mitigate between-group distinction and
negative emotional response towards out-group members” (p. 176). Thus, outgroup
communication as boundary spanning is a key to success in effective MNOS because
boundary spanning behaviors reduce between-group distinction and eventually integrate
two culture groups. Japanese personnel are working as managers, trainer, engineer,
technical supporter, so their functional roles and job characteristics require more

boundary spanning in Japanese MNOS.

Implications in Convergence of Japanese MNOS in the U.S.

In the previous two sections, the implications about different patterns ﬂom the
expectation of cultural stereotypes are discussed in the cognitive and behavioral levels.
These ﬁndings are referred to the concepts of convergence and divergence, and the
present research identiﬁes the distinctive features of effective MNOS; namely, the
importance of communicative convergence in MNOS.

Convergence in Cognitive and Behavioral Levels.

The measurement of communication ﬂameworks shows convergence among
Americans and Japanese, and the present research ﬁnds a primary ﬂamework composed
of exchange, normative, and negotiated order. This convergence and a primary
commmrication-ﬂamework are explained as the pragmatic outcomes of the hybrid

combination in American and Japanese ﬂameworks. These ﬁndings imply that the

85

cognitive level of convergence or ﬂame convergence between two culture groups is
important for effective MNOS and multiculturalism”. American and Japanese groups
learn the counterpart’s cultures and also learn the advantages of counterpart’s viewpoints
and communication ﬂameworks.

Ego-networks to ingroup and outgroup members is measured for identiﬁcation of
divergence or convergence in interactions of intergroup relationships. The present
research has two kinds of ﬁndings regarding ingroup-outgroup communication. One is
that both Americans and Japanese stress ingroup communication in informal settings
(organization-related and social networks). The other ﬁnding is that Americans
emphasize ingroup relationships but Japanese emphasize outgroup relationship in a
formal setting (job-oriented network). In these regards, it can be concluded that
Americans and Japanese are divergent in informal settings but they are convergent in a
formal setting. More precisely, Japanese attempt to assimilate their communication to
Americans on the job context. Assimilation is a unilateral process by minority culture
members (Cox, 1991), so it lacks mutual interactions of boundary spanning. Such
unilateral boundary spanning maintains weakness in Japanese MN Os. Divergence in
informal settings is also recognized weakness. In order to facilitate multiculturalism,
both American and Japanese should have more outgroup communication, and they should
be more convergent in any contexts.

In sum, the measurement of communication ﬂameworks indicates ﬂame
convergence between two culture groups, and the measurement of ingroup-outgroup

communication indicates Japanese unilateral boundary spanning to Americans.

86

Contextual Models of MNOS

On the basis of communicative convergence and divergence, I present the
contextual models to systematically observe Japanese MNOS in the U.S. Since a main
goal of the present study is identify complex contexts of MNOS as well as inﬂuence of

the national cultures, it is beneﬁcial to analyze the contextual types of MNOS carefully.

As Table 15 shows, the contextual model is composed of two aspects. One aspect
is the process of acculturation: namely, divergence or convergence of communicative
interactions (Larkey, 1996), which is indicated by ingroup-outgroup communication in
the present study. According to Cox (1991, 1993), acculturation processes are identiﬁed
as assimilation, separatism, and pluralism in the MN Os. Assimilation is considered as a
unilateral communication process by which minority culture members adopt the norms
and values of the dominant culture group (Cox, 1991). When one culture group highly
stresses ingroup relationship and the other group has more outgroup ties (i.e., unilateral
boundary spanning), it is regarded as assimilation. In contrast, convergence is not a
unilateral process but the mutual process toward cultural integration. When two culture
groups have many outgroup ties with each other, this is considered as convergence of two
groups. There might be active boundary spanning between two culture groups. On the
contrary, cultural divergence and cultural separation indicates little adaptation on either

side. When two culture groups strongly emphasize ingroup relationships and little

87

outgroup communication, there are few interactions between two groups and it is
divergence or separation in acculturation process.

The other aspect in Table 15 is the ‘types of organizations’ categorized by Cox
(1991) and Larkey (1996): monolithic, plural, and multicultural organizations. They are
classiﬁed in terms of policies, integration patterns, and diversity climate. This aspect is
also considered as the subculture conﬁguration (Johnson, 1993; Kiyomiya, 2000). A
monolithic organization implies that a mainstream culture dominates minority cultures.
Two possible contexts are conceived. One case is that the home culture dominates the
organization (ethnocentric), and the other case is that the host culture dominates the
organization (polycentric). Next, a plural organization exists when two different national
cultures coexist, and it is a pluralistic arrangement of subcultures in organizations
(Johnson, 1993). Lastly, a multicultural organization is the ideal type of MNOS and the
ultimate goal of acculturation.

These contexts of subculture conﬁguration are indicated by patterns of
communication ﬂameworks between two culture groups. When the organizational
members are in the context of monolithic MN Os, the individual ﬂameworks tend to
conform to either American (host) culture or Japanese (home) culture. When the
members exist in the context of plural organizations, individual’s communication
ﬂameworks must be differentiated in two culture groups. When the members exist in the
context of multiculturalism, individuals’ communication ﬂameworks must become
synthesized, and they may jointly create an alternative communication ﬂames: i.e., strong
or excellent organizational cultures. This is a ﬁnal stage of convergence in

communication ﬂameworks.

88

Thus, communicative convergence and divergence create two aspects of Table 15.
This conceptualization generates such contextual types as (l) ethnocentric divergence, (2)
ethnocentric assimilation, (3) polycentric divergence, (4) polycentric assimilation, (5)
pluralistic separation, (6) geocentric convergence, and (7) multiculturalism.
First, ethnocentric divergence is the context of a culturally divergent situation in the
ethnocentric organization: authority is located at headquarters (Teboul, et al., 1994). The
minority group is isolated ﬂom the mainstream culture. Second, ethnocentric
assimilation is the context of cultural assimilation toward the home culture situated in the
monolithic organization. For example, Japanese culture dominated the organization since
Japanese own the companies and Japanese is the home culture in the Japanese MNOs.
Americans are isolated in this context of ethnocentric divergence. However, Americans
adapt to the home culture in the context of ethnocentric assimilation, and Americans
attempt to assimilate while Japanese do not change and maintain their own styles.

According to Johnson (1993), while the ethnocentric organizations are oriented to
the ‘home’ country (Japan), the polycentric organizations are oriented to the ‘host’
country (U SA.) The polycentric organizations are ones in which the local nationals
hold key positions and dominate business practices (Teboul et al., 1994). In this sense, for
example, American culture is the majority and dominate the organizations, and Japanese
culture is a very small part in the polycentric organizations. Polycentric divergence is the
context of cultural divergence situated in the polycentric culture. This context indicates
that Japanese do not adapt to and do not learn ﬂom the host culture of American
business. On the contrary, polycentric assimilation is the context that Japanese more

actively assimilate their business styles and behaviors.

89

The context of pluralistic separation is a culturally divergent situation in the
plural organizations. It implies the context in which both culture groups make little
efforts for adopting the other side. In this case, Americans and Japanese are culturally
conﬂicting in the MN 0. On the contrary, geocentric convergence is the context of a
convergent process in the plural organizations. It implies that both culture groups have
mutual efforts to adopt norms and values of the other side. Americans and Japanese
mutually learn something ﬂom the other culture, so both culture groups have more active
boundary spanning to each other in order to integrate diverse workplace.

Multiculturalism is the context of full structural and cultural integration over the
cultural identities. “Such organization is extremely rare” (Larkey, 1996, 470), and that is
why multiculturalism is the ideal type of MNOs and the ultimate goal of acculturation. It
implied that there is little gap and no complaint between two culture groups in terms of
management practices and communication. In addition, new culture and new patterns of
communication emerge in both formal and informal networks in the MNOS (Larkey,
1996)

These contextual types provide a systematic view of MNOS and they indicate the
context of the present study. The current research ﬁnds that Japanese emphasize
outgroup communication in the formal (job-oriented) setting and American and Japanese
emphasis on ingroup communication in the informal (organization-related and social)
settings. The former is considered as Japanese assimilation to Americans in the formal
setting, and the latter is considered as divergence in informal settings. Moreover, ﬂame

convergence is identiﬁed in the two culture groups.

90

On the basis of these research outcomes, in the formal setting, the present study
concludes that the individual respondents exist in the context of polycentric assimilation
that Table 15 illustrates. In the informal settings, they exist in the context of polycentric
divergence. Polycentric assimilation and polycentric divergence lack outgroup
communication. Particularly, it is necessary for Americans to increase outgroup
commrmication in the formal and informal settings and necessary for Japanese to
emphasize outgroup relationship in the informal settings. Since convergence is a way to
mitigate antagonism and develop cooperative relationships, ﬂame convergence and
outgroup communication are critical for effective MNOS. Multiculturalism and
effectiveness of MNOS will be facilitated when managers play a key role in ﬂame
convergence and outgroup communication. For example, they rrright be able to exchange
their perspectives and opinions in informal settings. In so doing, they can share their
ideas, information, and ﬂameworks, and simultaneously they have mutual boundary
spanning.

Thus, the contextual models of Table 15 provide practitioners and researchers

more systematic perspective to consider the complex context of MNOS.

Implications in Communication Framework and Outgroup Communication
As mentioned early, outgroup communication within a MNO is considered as
boundary spanning which links different groups within an organization. It implies that
when outgroup communication increase, boundary spanning increase within IVDNIOS.
Unilateral boundary spanning ﬂom Japanese to Americans can be identiﬁed in the current

data, but mutual boundary spanning is necessary to enhance communicative convergence

9]

and lead to multiculturalism eventually. Thus, increase of outgroup communication is
central to develop effectiveness of MN Os.

When the present study analyzes the relationship between the total FINT score
and a percent of outgroup ties, for American data, the study ﬁnds a positive association
between them. As Table 13 shows, the total score of FINT has strong correlation
coefficient with a percentage of outgroup ties in job-oriented communication (r = .57 p <
.01) and in organization-related communication (r = .53 p < .01). However, signiﬁcant
correlation with a percent of outgroup ties can not be found in social communication, but
it is positively associated with the sentiments ﬂamework (r = .36 p < .05).

There are some implications in these results. First, the FINT scale is a good
measurement instrument for assessing the context of cooperative relationships as Johnson
(1997a) designs it. The FIN T score indirectly indicates the degree of cooperative
relationship, showing positive correlation with outgroup communication which is a key
index for effectiveness of MN Os. Increase of ingroup communication strengthens
isolation ﬂom the outgroup and expands divergence between two culture groups, so
outgroup communication, not ingroup communication, is related to cooperative climates
within MNOS. Therefore, the F INT scale correctly shows the positive correlation with
outgroup communication, and it is an indirect index for the cooperative relationships
within a MNO.

Moreover, as strong correlation between FINT and outgroup ties does not appear
in social communication, and the sentiments ﬂamework only have positive correlation
with outgroup commrmication in this context. As Johnson (1997a) predicted, the

sentiments ﬂamework is recognized different ﬂom the other subscales in the FINT, and it

92

is positively related to informal communication structure. In addition, the negotiated
order ﬂamework has relatively strong correlation with outgroup ties in social
communication although it does not reach a level of statistical signiﬁcance (r = .30). It is
also a strongest correlation coefficient in organization-related communication (r = .54, p
< .01). In concept, it is reasonable to think that negotiated order gives a basis for
informal communication. In Table 14, the Japanese data shows evidence that negotiate
order has strong correlation with outgroup ties in social communication (r = .43, p < .05).
Therefore, both sentiments and negotiated order are different ﬂom the other dimensions
of ﬂameworks for interactions, and they are key ﬂameworks in social settings. In
contrast, the other three ﬂameworks of exchange, formal, and normative are strongly
correlated with outgroup communication in business settings: job-oriented and
organization-related communications.

Lastly, the research ﬁndings provide a good implication to practitioners. F INT
dimensions can be considered in terms of improvement of communication skills and
leadership skills for managers, as some works argue these points (Bohnan & Deal, 1991;
Fairhurst & Sarr, 1996). Increase of total F INT score indicates more variety of
communication ﬂames. On the other hands, if the total F INT score is low, it indicates
more narrow or simple communication ﬂames. More different ﬂameworks will lead
sophistication of communication in organizations, and eventually they will shape good
communication skills for leaders. Thus, the FINT scale is not beneﬁcial for academic

researchers but also useful for business practitioners.

93

Limitations and Future Studies

There are several limitations in the present study. One important discussion is
generalizability of this study. As explained in the method chapter, the present research
has purposeful sampling, not probability sampling. Therefore, the ﬁndings and
implications are limited in certain conditions. All participants are working in auto-related
business, one of the most competitive markets in the world economy. These MNOS are
owned by Japanese and they are moderate in size. The research sites are limited in the
Detroit area. Thus, the conclusion of this study is limited in such context as the Japanese-
owned, small and moderate sized organizations, and auto-related business at Detroit.
Other kinds of MNOS are deﬁned and characterized by different conditions, market
environments and cultural background, so there might be different results. However, the
present study can provide good insights and beneﬁcial implications for other MNO
studies.

The present research controls some variables, such as organization size and
differences in corporate positions. However, there are some other variables of workplace
diversity, gender and other ethnic groups. Gender is one of important issues for
workplace diversity (Cox, 1993; Ibarra, 1992, 1993, 1997), and it might become a source
of identity groups. Gender will be intertwined with cross-culture issues in MNOS, and it
may make the study more complicated. Therefore, the present study does not deal with
gender issues. This is an important direction for a future study. In addition, the
American sample in the present study is almost all Caucasian American. This is related

to another issue of generalization. Other ethnic groups, such as Aﬂican, Hispanic, and

94

Asian Americans, may have different communication patterns in MNOS. In this sense,
the American sample in this study is limited in one speciﬁc group of Americans.

The present study has some other limitations. The sample size is relatively small:
a sample size of social network data in particular. As a result, some statistical analyses
did not reach a signiﬁcant level. For example, no signiﬁcant correlation between total
FINT score and outgroup ties is found in the Japanese data because of its small sample
size (refer to Table 14). In order to improve empirical evidence, the future study has to
consider the effect size of data collection.

After ﬁnishing data collection, I realized how difficult the social network data are
collected ﬂom business organizations. Social network questionnaire asks about private
information of their personal networks, and some participants did not respond or skip this
section of questionnaire. However, social network questionnaire is critical for the study
of MNOS. Since social network research and theory have primarily evolved in
uninational setting, Salk and Brannen (2000) recommend that network concept be applied
to enrich understanding of individual inﬂuence in multicultural settings. “More studies
need to be conducted in non-U.S. and multicultural settings to conﬁrm the
generalizability of social network theory” (Salk & Brannen, 2000, p. 201). For this end,
network questionnaire must be improved to gain a high response rate in the future
research.

Moreover, the opportunities to get data ﬂom MNOS are fewer than those to get
data ﬂom American companies. The number of MNOS is growing but they are still
special in the types of organizations. The research in MN Os must be improved in terms

of procedure of data collection. Namely, it requires stronger cooperation with MNOS.

95

Therefore, it might be effective to provide incentives or rewards to individuals or
organizations. The procedure of data collection must be carefully considered for gaining
better cooperation ﬂom MNOS.

Another improvement must be necessary for consistency between English and
Japanese versions of questionnaires. Although reliability is satisfactory in the F INT
scale, Japanese translation must be improved on the items of the exchange ﬂame. Also,
as Johnson (1997a) points out, the sentiments subscale needs more items for enhancing
statistical strength since it has only three items. In addition, this research only uses a self-
reported type of quantitative method. Some qualitative methods may enrich the study,
and multiple approaches will make the future research more ﬂuitful.

In conclusion, the present study has some important ﬁndings and implication.
First, the cultural difference between Americans and Japanese is identiﬁed in
communication ﬂames and patterns of ingroup-outgroup communications in Japanese
MNOS. Second, the measurement of communication ﬂames contradicts the expectations
or the norms of cross-cultural stereotypes. The measured communication ﬂames are
convergent between Americans and Japanese, and they have a primary communication-
ﬂameworks composed of exchange, normative, and negotiated order. This is a pragmatic
outcome of economic decisions and the hybrid combination between Americans and
Japanese. Third, ingroup and outgroup communications in Americans and Japanese are
not consistent with the expectation of cross-cultural stereotypes. Americans emphasize
ingroup communication in both formal (job-oriented) and informal (organization-related
and social) setting. In contrast, Japanese similarly emphasize ingroup communication in

informal settings but emphasize outgroup communication in a formal setting. Fourth, it

96

is concluded that convergence in communication ﬂame and communicative interactions
is a key to success in MNOS. Frame convergence between two culture groups and
Japanese outgroup communication to Americans are recognized as the context of the
polycentric assimilation. F iﬁh, outgroup communication in MNOS is considered as
boundary spanning, and increase of outgroup communication is critical for both
Americans and Japanese to attain the goal of effective MNOS. For Americans, the total
FIN T scale is positively associated with a percent of outgroup ties in business settings
(job-oriented and organization-related). The F INT score becomes a good indicator for
cooperative climates and effectiveness of MN Os as well as excellent leadership skills.

These implications must be reanalyzed in the future studies. The present study
has limitation in the context of Japanese MNOS at an automotive industry, but these
ﬁndings and implications can be applied to other contexts. One future study is to expand
the research to Japanese MNOS in other industries. Another extension of this study is to
conduct the same research in the opposite context of the present research: namely, a
research in American MNOS in Japan. This is an appropriate comparison for the present
research. Furthermore, in order to attain effectiveness of MNOS, the research should be
conducted on many other issues of workplace diversity in MNOS, such as discrimination,
organizational justice and democracy.

This is a beginning study of multinational organizations. The more the world
merges in the societies and economics, the more critical an understanding of MNOS will
be in the future. The study of MNOS will become no more special case of organizational
communication, and intercultural organizational communication will become an effective

approach of interdisciplinary views to culturally diverse workplaces.

97

FOOTNOTE

I Shuter and Wiseman (1994) introduce four research questions that guide
communication scholars to effective studies of intercultural organizational
communication (IOC). The others are; 2) how does national culture inﬂuence the
communication between organizations in different countries? 3) how does national
culture inﬂuence communication within and between organizations in the same country
and world region? 4) how do ethnicity and race inﬂuence the transmission and reception
of information within and between organizations in the same national culture? (p. 8)

2 These are good examples of American and Japanese communication, which examine
their cross-cultural stereotypes. They are not directly related to management and
organizational issues but beneﬁcial for a contextual perspective to American-Japanese
communication in the MNOS.

3 “In the dichotomy A (individualism) and Non-A (collectivism), A’s understanding
could be very accurate but Non-A’s would not be” (N ishida, 1996, p. 108).

4 Allocentrism is correlated positively with social support and negatively with alienation
and anomie in the U.S. Idiocentrism, in contrast, is correlated positively with an
emphasis on achievement and perceived loneliness in the U.S.

5 Independent and interdependent self-construals, which are developed by Markus and
Kitayama (1991), are the most widely used conceptualizations (Gudykunst & Matsumoto,
1996)

6 Recently, Ohashi (2000) developed an instrument to measure high/low context
communication, and she emphasizes the importance of this concept as well as relative
independence ﬂom the constructs of individualism-collectivism.

7 The great attention to “ﬂame” is not only in the ﬁeld of organizational communication
but also in the entire ﬁeld of communication. Putnam and Holmer (1992) and Drake and
Donohue (1996) introduce the studies of ‘ﬂames’ discussed in the ﬁeld of negotiation.
Scheufele (1999) review the framing approaches in media study, and Dearing and Rogers
(1996) illustrate the related study of agenda-setting. Thus, ﬂame approaches are
diversiﬁed in the entire communication ﬁeld, and it can be understood that how the ﬂame
perspective is a crucial concept in communication study. In the present research, it is out
of purposes to expand discussions to these approaches.

8 The etic approach focuses on understanding cultures ﬂom the outside by comparing
cultures using predetermined characteristics. In contrast, the emic approach focuses on
studying cultures ﬂom inside, understanding cultures as the members of the cultures
understand them (Gudykunst & Ting-Toomey, 1996).

9 Some recent studies of intercultural communication bring up the issues of individual
levels of individual-collectivism (Gudykunst & Matsumoto, 1996) and focus on personal
orientation (Triandis, Leung, Villareal, & Clack, 1985) and self-construals (Markus &
Kitayama, 1991). They are not concerned with the macro or structural context which
strongly inﬂuences individual cognition and behavior. As long as there is no context-ﬂee
environment, the context must be taken into a consideration in IOC studies.

10 This might not be an idiosyncratic strategy for MNOS, but it might be a more common
strategy in highly competitive industries.

98

11 A dictionary meaning of “pragmatism” is a method of solving problems and affairs by
practical means (The American Heritage Dictionary, 1983). In the present study, it is
implied that a hybrid communication-ﬂame results ﬂom the rational strategies or
methods of problem-solving in order to improve economic advantages and organizational
strengths.

12 Salk and Brannen (2000) ﬁnd the similar results of convergence on the local norms for
decision-making. Namely, Japanese adjust their behavior and values to German styles,
which are local norms.

99

REFERENCE

Anderson, J .A. (1987). Communication research: Issues and methods. New York:
McGraw-Hill.

Abegglen, J. C. (1958). The Japanese factog: Asmct of its social organization.
London, UK.

Bateson, G. (1972). Steps taan ecology of mind. New York: Ballantine Books.

Briggs, P. (1988). The Japanese at work: Illusions of the ideal. Industrial
Relation_§ Journal, 19(1), 24-29.

Cameron, K. S., & Quinn, R. E. (1999). Diagposing and changing organizational
culture: Based on the competingyalues ﬂamework. Reading, MA: Addison-Wesley.

Chaffee, S. H. (1991). Explication. Newbury Park, CA: Sage Publications.

Chang, H-J., Johnson, J. D., Cox, D., & Kiyonriya, T. (1997). The communication
environment of the CISRC: External communication and boundapy spanning. Paper

presented at the Annual Meeting of the International Communication Association,
Montreal, Canada.

Chen, G-M., & Chung, J. (1994). The "ﬁve Asian dragons": Management
behaviors and organizational communication. Communication Merly, 42, 93-105.

Chen, G-M., & Starosta, W. J. (1998). Foundations of intercultural
commrmication. Needham Height, MA: Allyn & Bacon.

Conrad, C., & Poole, M. S. (1998). Strategic organizational communication: Into
the twenty-ﬁrst century (fourth ed.). Orlando, FL: Harcourt Brace College Publishers.

Cox, T. (1991). The multicultural organization. Academy of Management
Executive, 5(2), 34-47.

Cox, T. (1993). Cultural diversity in organizations: Theog, researca and practice.

San Francisco, CA: Berrett-Koehler Publishers.

Cox, T., & Blake, S. (1991). Managing cultural diversity: Implication for
organizational effectiveness. Academy of Management Executive, 5(3), 45-56.

Cutcher-Gershenfeld, J., Nitta, M., Barret, B., Belhedi, N., Chow, S-C., Inaba, T.,
Ishino, T., Lin, W-J., Mothersell, W., Palthe, J., Ramananad, S., Strolle, M.E.,
WheatonJ., AC. (1998). Knowledge-driven work: Unexpected lesson_s ﬂom Japanese and
United States work practices. New York: Oxford University Press.

100

Dearing, J .W. & Rogers, EM. (1996). Agenda-setting. Thousand Oaks, CA: Sage
Publications.

Deal, T.E. & Kennedy, A.A. (1982). Cor_‘p_orate cultures: The rites and rituals of
Comorate life. Reading, MA: Addison-Wesley.

Dietz-Uhler, B., & Murrell, A. J. (1998). Evaluations of afﬁrmative action
applicants: Perceived fairness, human capital, or social identity? Sex Roles. 38, 933-951.

Drake, L.E. & Donohue, WA. (1996). Communication ﬂaming theory in conﬂict
resolution. Communication Research 23, 297-322.

 

Durlabhji, S., & Marks, N. E. (Eds). (1993). Japanese business: Cultural
persmctives. New York: State University of New York Press.

Fairhurst, G.T. & Sarr, RA. (1996). The art of ﬂaming: Managing the langu_age
of leadership. San Francisco, CA: Jossey-Bass.

Fukunaga, K. (1992). Koshi no "michi," Rousou no "michi" (A means of
Confucianists). President. 3Q 70-77.

Gallois, C., Franklyn-Stikes, A., Giles, H., & Coupland, N. (1988).
Communication accommodation in intercultural encounters. In Y. Y. Kim & W. B.
Gudykunst (Eds), Theories in intercultural communication (pp. 157-185). Newbury Park,
CA: Sage Publication.

Gao, G., Schmidt, K. L., & Gudykunst, W. B. (1994). Strength of ethnic identity
and perceptions of ethnolinguistic vitality among Mexican Americans. Hispanic Journal
of Behavioral Sciences. 16(3), 332-.

Gofﬂnan, E. (1974). Framing analysis: An essay on the organization of
exgrience. New York: Harper & Row.

Goldman, A. (1994). Communication in Japanese multicultural organizations. In

R. L. Wiseman & R. Shuter (Eds), Communicating in multinational organizations (pp.
45-74). Thousand Oaks, CA: Sage Publications.

Gudykunst, W. B. (1997). Cultural variability in communication: An introduction.
Communication Research. 24(4), 327-348.

Gudykunst, W. B., & Matsumoto, Y. (1996). Cross-cultural variability of
communication in personal relationship. In W. B. Gudykunst, S. Ting-Toomey, & T.
Nishida (Eds), Communication in mrsonal relationship across cultures (pp. 19-56).
Thousand Oaks, CA: Sage Publications.

101

Gudykunst, W. B., & Ting-Toomey, S. (1988). Culture and interpersraral
communication. Newbury Park, CA: Sage Publications.

Gudykunst, W. B., & Ting-Toomey, S. (1996). Communication in personal
relationship across cultures: An introduction. In W. B. Gudykunst, S. Ting-Toomey, & T.

Nishida (Eds), Communication in personal relationship across cultures (pp. 3-16).
Thousand Oaks, CA: Sage Publications.

Gudykunst, W. B., Yoon, Y. C., & Nishida, T. (1987). The inﬂuence of
individualism-collectivism on perceptions of communication in ingroup and outgroup
relationship. Communication Monographs, 54, 295-306.

Hall, E. T. (1959). The silent language. New York: Doubleday.
Hall, E. T. (1976). Beyond culture. Garden City, NY: Anchor.

Hamilton & Hunter, LE. (1988). Conﬁrmatory Factor Analysis [Computer
software]

Hartman, R. L., & Johnson, J. D. (1990). Formal and informal communication
group communication structures: An examination of their relationship to role ambiguity.
Sociﬂeﬂvorks, 12, 127-151.

Hasegawa, T., & Gudykunst, W. B. (1998). Silence in Japan and the United
States. Journal of Cross-Cultuaal Psychology. 29(5), 668-684.

 

Hatch, M. J. (1997). Organization theog: Modern, symbolic, and mstmodem
persmctives. New York: Oxford University Press.

Hofstede, G. (1980). Motivation, leadership, and organizations: Do American
theories apply abroad? Organizational Dynamics, 42-63.

Hofstede, G. (1997). Cultures and organizations: Software of the mind. New
York, NY: McGraw-Hill.

Hunter, J. E., & Gerbing, D. W. (1982). Unidirnensional measurement, second
order factor analysis, and causal models. Research in Organizational Behavior. 4. 267-
320.

Ibarra, H. (1997). Paving an alternative route: Gender difference in managerial
networks. Social Psychology Quarterly, 60(1), 91-102.

Johnson, J. D. (1992). Media for interaction and disbandments:

Organizational/institutional communication in the Good Society. Paper presented at the
Annual Meeting of the Speech Communication Association, Chicago.

102

Johnson, J. D. (1993). Organizational communication structure. Norwood, NJ:
Ablex Publishing Corporation.

Johnson, J. D. (1997a). A ﬂamework for interaction (F IN T) scale: Extensions and
reﬁnement in an industrial setting. Communication Studies 48, 127-141.

 

Johnson, J. D. (1997b). Review of the books, "The art of ﬂaming: Managing the
language of leadership" and "Frame reﬂection: Toward the resolution of intractable

policy controversies". Quarterly Journal of Smech, 83, 397-398.

Kerlinger, RN. (1973). F ougation of tmhavioral research (2nd edL New York:
Holt, Hinehart and Winston.

Kelly, C. (1990). Social identity and intergroup perceptions in minority-majority
contexts. Human Relations. 43(6), 583-.

Kim, Y. Y., & Paulk, S. (1994). Intercultural challenges and personal
adjustments: A qualitative analysis of the experiences of American and Japanese co-
workers. In R. L. Wiseman & R. Shuter (Eds), Communicating in Multinational
Organizations (pp. 117-140). Thousand Oaks, CA: Sage Publications.

Kiyomiya, T. (2000). Subculture assessment in Japanese multinational

organizations: Applying commting value model. Paper presented at the F iftieth Annual
Conference of the International Communication Association, Acapulco, Mexico.

Klopf, D. W. (1991). Japanese communication practices: Recent comparative

research. Communication Quarterly, 39(2), 130-143.

Kluckhohn, C., & Strodbeck, F. (1961). Variations in value orientations.
Evanston, IL: Row, Peterson.

Kuwahara, Y. (1993). Industrial relations in Japan. In G. J. Bamber & R. D.
Lansbury(Eds.), International and comparative industrial relations: A study of
industrialised mgkﬁet economies (pp. 220-244). Sydney, Australia: Allen & Unwin.

Larkey, L. K. (1996). Toward a theory of communicative interactions in culturally
diverse workgroups. Academy of Management Review. 21(2), 463-491.

Limaye, M. R. (2000). Managing cultural diversity in the American workplace. In

L. A. Sarnovar & R. E. Porter (Eds), Intercultural communication: A reader (Ninth Ed.)
(pp. 312-324). Belmont, CA: Wadsworth Publishing Company.

Lincoln, J. R. (1989). Employee work attitudes and management practice in the
U.S. and Japan: Evidence ﬂom a large comparative survey. California Management
Review 89-106.

 

103

Lincoln, J. R. (1990). Japanese organization and organization theory. Research in
Organizational Behavior. 12. 255-294.

Lincoln, J. R., & Kalleberg, A. L. (1985). Work organization and workforce
commitment: A study of plants and employees in the U.S. and Japan. American
Sociological Review. 50. 738-760.

Markus, H. R., & Kitayama, S. (1991). Culture and the self: Implication for
cognition, emotion, and motivation. lﬁychological Review. 98. 224-253.

McDaniel, E. R., & Quasha, S. (2000). The communicative aspects of doing
business in Japan. In L. A. Sarnovar & R. E. Porter (Eds), Intercultural Communication:
A reader (Ninth Ed.) (pp. 312-324). Belmont, CA: Wadsworth Publishing Company.

Misumi, J. (1984). Decision-making in Japanese groups and organizations. In B.
Wilpert & A. Sorge (Eds), lntemational Perspective on Organizational Democracy (pp.
525-539). John Wiley & Sons Ltd.

MITI. (1998). White paper on international trade 1998. Tokyo, Japan: Ministry of
lntemational Trade and Industry.

 

Morrow, L. (1992). Japan in the mind of America. Time, 16-21.

MS Research (1999). The number of Japanese companies [Online] Available
http://www.msresearch.com/eng/jpzﬁrmj.htrnl, December 1, 1999.

Neuliep, J. W. (1997). A cross-cultural comparison of teacher immediacy in
American and Japanese college classrooms. Communication Research 24(4), 431-451.

 

Nichibei Hikaku Kenkyuukai. (1989). Japan vs U.S. in ﬁgurj. Tokyo, Japan:
PHP Kenkyukai.

Nishida, T. (1996). Communication in personal relationships in Japan. In W. B.
Gudykunst, S. Ting-Toomey, & T. Nishida (Eds), Communication in Persona
Relationship Across Cultures (pp. 102-121). Thousand Oaks, CA: Sage Publications.

Oetzel, J. G. (1998). The effects of self-construals and ethnicity on self-reported
conﬂict styles. Communication Reports, 11(2), 133-144.

Ohashi, R. (2000). High/low context communication: Concepmﬂmtion and scale
development. Unpublished doctoral dissertation, Michigan State University, Michigan.

Ouchi, W. G. (1979). A conceptual ﬂamework for the design of organizational
control mechanisms. Management Science. 25. 833-848.

104

Ouchi, W. G. (1980). Markets, bureaucracies, and clans Administrative Science

Quarterly, 25, 129-141.

Ouchi, W. G. (1981). Theory Z: How American business can meet the Japanese
challenge. Reading, MA: Addison-Wesley Publishing Company.

Patton, MO. (1990). Qualitative evalu_ation and research methods (2nd ed.).
Newbury Park, CA: Sage Publications.

Peters, T.J. & Waterman, RH. (1982). In search of excellence: lessons ﬂom
America's best-run companies. New York: Harper & Row.

 

Putnam, L. L., & Holmer, M. (1992). Framing, reﬂaming, and issue development.
In L. L. Putnam & M. E. Roloff (Eds), Communication and negotiation (pp. 109-127).
Newbury Park, CA: Sage Publications.

Quinn, R. E. (1988). Beyond rational management: Mastering the paradoxes and
commting demands of high performance. San Francisco, CA: Jossey-Bass Publishers.

Rehfeld, J .E. (1990). What working for a Japanese company taught me. Harvard
Business Review. 68. (6) 167-176.

Salk, J. & Brannen, M.Y. (2000). National culture networks, and individual
inﬂuence in a multinational management team. Academy of Management Journal, 43 (2),
191-202.

Scheufele, DA. (1999). Framing as a theory of media effects. Journal of
Communication 49 (1), 103-122.

 

Shimada, H. (1983). Japanese industrial relations: A new general model? A
survey of the English-language literature. In T. Shirai (Ed.), Contemmﬂy industrial

relations in Japan (pp. 117-143).

Shuter, R., & Wiseman, R. L. (1994). Communication in multinational
organizations: Conceptual, theoretical, and practical issues. In R. L. Wiseman & R.
Shuter (Eds), Communicating in multinational organizations (pp. 3-11). Thousand Oaks,
CA: Sage Publications.

Stephan, C. W., Stephan, W. G., Saito, I., & Barnett, S. M. (1998). Emotional
expression in Japan and the United States: The nonmonlithic nature of individualism and
collectivism. Journal of Cross-Cultural Psychology. 29(6), 728-748.

Sugimoto, N. (1997). A Japanese-U.S. comparison of apology styles.
Communication Research 24(4), 349-369.

 

105

Sullivan, J., & Taylor, S. (1991). A cross-cultural test of compliance-gaining
theory. Management Communigtion Quarterly/JO), 220-239.

Suzuki, S. (1997). Cultural transmission in international organizations: Impact of

interpersonal communication patterns in intergroup contexts. Human Communication
Research. 24(1), 147-180.

Suzuki, S. (1998). In-group and out-group communication patterns in
international organizations: Implication of social identity theory. Commmrication
Research 25(2), 154-182.

Teboul, J. B., Chen, L., & Fritz, L. M. (1994). Intercultural organizational
communication research in multinational organizations. In R. L. Wiseman & R. Shuter

(Eds), Communicating in multinational organizations (pp. 12-30). Thousand Oaks, CA:
Sage Publications.

Triandis, H. C. (1986). Collectivism vs. individualism: A reconccptualization of a
basic concept in cross-cultural psychology. In C. Bagley & G. Verma (Eds), Personality,

cognition. and values: Cross-cultural perspectives of childhood and adolescence. London,
UK: MacMillan.

Triandis, H. C., Leung, K., Villareal, M., & Clack, F. (1985). Allocentric versus
idiocentric tendencies. Journal of Research in Personality. 19. 395-415.

Wasserrnan, S., & Faust, K. (1994). Sociaﬂetwork analysis: Methods and
applications. New York: Cambridge University Press.

Wurzel, J. S. (1988). Toward multiculturalism: A reader in multicultural
education. Yarmouth, ME: Intercultural Press.

Yang, C. Y. (1984). Demystifying Japanese management practices. Harvard
B_u§iness Review. 172-182.

Yum, J. O. (1988). The Impact of Confucianism on Interpersonal Relationships

and Communication Patterns in East Asia. Communication Monoggaphs, 55(December),
374-388.

106

APPENDIX A

TABLES

107

TABLE 1

Japanese and Western Organizational Behavior and Protocol

 

Japanese Protocol

Western Protocol

 

Generalist workers

Advancement by seniority

Private, prescribed channels for
grievances

Publicly conciliatory

Organizational relationships are highest
priority

Long-term organizational agendas
Primarily vertical upward and horizontal
communication

Clear distinction between tatemae and
hone

Accessible informal channels for
manager-employee communication
Decision-making via complete consensus

Amae (interdependencies) crucial in intra-
and interorganizational communication
(e.g., keirestu)

Strong dependencies (giri) and
commitment between organizations and
employees

Organizational security via lifetime
employment in large MNOS

Close workplace proximics

Ritualized, restricted formal codes for
interaction

Interactions more situationally bound
Valuing intuitive, nonverbal
communication (harager)

More reliance on face-to-face
communication

Specialist Workers
By accomplishment
Public areas for grievance and disputes

Publicly more argumentative
Tasks/goals are highest priority

Shorter term organizational agendas
More vertical downward communication

Less distinction between surface
communication and true intentions
More formalized channels for manager-
employee communication
Decision-making via majority vote or
designated leaders

Amae less pronounced and not publicly
sanctioned (e.g., antitrust legislation)

Less binding, more ﬂux in commitments
between organizations and employees

More turnover, less security; layoffs,
ﬁrings

More individualized work spaces
More informal; less restricted codes

More ideologically bound

Values analytical logic over intuitive
communication

Greater use of print communication

 

From “Communication in Japanese Multicultural Organizations” by A. Goldman, 1994, p. 55. In R. L.
Wiseman & R. Shuter (Eds), Communicating in Multinational Organizations (pp. 45-74). Thousand Oaks,
CA: SAGE Publications. Copyright 1994 by Speech Communication Association, lntemational and
Intercultural Communication Division. Reprinted with permission of National Communication Association
(Former Speech Communication Association), lntemational and Intercultural Communication Division.

108

TABLE 2

Description of Research Sites

 

Business and Industry

 

 

 

 

 

 

 

 

 

Company A Production, sales, and maintenance of machines: injection molding,
die casting, extrusion presses

Company B lntemational trading: auto-related materials, equipment, parts, and
investments

Company C Production, sales, and maintenance of assembly-equipment and
material handling systems,

Company D lntemational trading: auto-related materials, equipment, parts, and
plant-building

Company E Research & development of auto-seats and seating systems

Company F Production, research & development, marketing of auto-related /
electronic parts

Company G Marketing and trading of auto-production techniques, quality,
warrantee, and purchase of materials

Company H Research & development and marketing of climate control, exhaust,
and heat exchange systems

Company I Production of surface parts (doors, hoods, etc.), manufacturing

equipment and engineering services

 

109

TABLE 3
Demographic Information on Respondents for Each Company

N ears
service

 

Co: Company

Nationality: Nationality is divided into the two categories, Americans and Japanese. Aﬂican-, Asian-,
Caucasian-, Hispanic-, and Native-Americans were classiﬁed and investigated on the questionnaire,
and they were integrated in Americans since there was no significant variance.

Corporate Status: The hierarchical status in each organization is classiﬁed into four categories, such as Top
Management (CEO, and directors), Managers (senior and junior managers), Upper-level Workers
(senior-workers, supervisors, tearn-leaders), and Workers.

Years service: Statistical mean of years of service.

J: This is not a company, but it is the additional data set of individuals who work at Japanese MNOS at
Detroit: by the researcher’s personal contacts.

110

TABLE 4

Exploratory Factor Analysis for F INT Items: Rotated Component Matrix.

 

 

 

Component

1 2 3 4 5
EXCl .145 .144 -1.012E-02 3.756E-02 .838
EXC2 9.192E-02 .206 .138 5.04 l E-02 .750
EXC3 .242 .284 .l 16 -1.512E-02 .428
EXC4 .496 .221 .132 .335 .250
FMLl 1.186E-02 .167 .788 .155 -6.882E-02
FMLZ -.204 .139 .812 7.711E-02 .103
FML3 .326 8.776E-02 .712 -.102 -5 .732E-02
FML4 -2 .690E-O3 -7.523E-02 .773 -6.3 l4E-02 .142
FMLS .353 -.104 .630 -8.576E-02 .206
NRMl .838 . 109 6.641E-02 -3 .807E-02 .206
NRM2 .817 9.444E-02 5.752E—02 .l 10 8.295E-02
NRM3 .808 .329 -6.549E-03 .153 4.081E-02
NRM4 .797 .270 6.666E-02 7.41 SE-OZ .101
SNTl .225 .161 1.285E-02 .858 1.736E-02
SNT2 .192 .235 3.518E-03 .880 3.73 8E-02
SNT3 -8. l 90E-02 -1 . 157E-02 -2.294E-02 .761 1.428E-02
NEG 1 .197 .788 2.091 E-02 .244 . 138
NEGZ .200 .860 -3 .842E-03 . 1 87 . 180
NEG3 . 174 .863 3.605E-02 5.397E-02 .247
NEG4 .210 .848 .154 7.341E-O3 .105

 

Extraction Method: Principal Component Analysis.
Rotation Method: Varimax with Kaiser Normalization.

Rotation converged in 6 iterations.

111

.30—23-3 _o>o_ 2V6 05 an “gag—aw? m_ GEE—2.50 a...
.28—3-6 .05. mod 05 8 aﬁoumqwmm 2 corn—2.50 e

 

M3.—

3.—

EL $4 .2..— wwé em; cm; no; mm; 24 .Q.m

 

26

mod wmé

me

wcé om.m 0:. wm.m had end 36 mum-m 26 .822

 

an.
:2.-

co.—

0.

3.

co.—

2. 8. :8. 2. .2. :8. :8. :8. :8. 8
2. 8.- 2. :. .2. :8. :8. :8. :8. 2
8. 8.- 2. 8. 8. :8. :8. :8. :8. 2
8. 8. 8. 2. .2. :8. :8. :8. :8. 2
8.- :.- 8.- 8. :. 8. 8.- 8. 8. 2
8. 8.- 8. 8. 2. :3. 2. 2. z. 2
8. 8.- 8.- 8. :. :8. 2. 2. 2. S
:8. 8. :8. 8.- 8. :8. :8. :8. :8. 2
:8. 8.- .2. 8.- 2. :8. :8. .2. :8. 2
2. 8. 8. 8.- 3. 8. 8. 8. 2. :

O .0 O. O. .0

:8. 8. :8. 8.- 8. :8. :8. :8. 8. 2

co; 0v. mm. 3.. mm. 3. 2. 3. mm. a

O. O. .0 O.

8.. :8. :8. :8. 8. 8. .2. 8. w
8.. :8. :8. :8. :8. 8. 8. e

82 :8. 8. 8. :8. 8. e

82 .2. .2. 2. 8.- m

82 :8. :8. 8. e

8.2 :8. :8. m

8.2 :3. 8

82 _

 

8.. :2
8.8 8.8.
82 :8.
82

8 2

t

A:

 

was: HZE «o 2.582 noun—8.80
m mama;

112

TABLE 6
Items of Frame for Interaction (FINT) Scale: Reliability and Factor Loading

 

Exchange Frame: or = .66

(Individuals are seen as driven to maximize rewards through their interaction with others. An

 

 

 

 

 

 

 

 

act of reciprocal giving and receiving.) Factor
Loading

ICOMMUNICA TE WITH OTHERSAT THE ORGANIZA TION BECA USE - - -

1. they have something I need to solve a work related problem. .65

2. they can provide me with technical advice. .61

3. I can help them solve a problem. .53

4. we both get something of value ﬂom the interchange. .49

Formal Frame: or = .81

(Individuals are seen as driven and motivated by the requirement of the positions that they

occupy in the formal structure. Bureaucratic behavior.) Factor
Loading

I COMMUNICATE WITH OTHERS AT THE ORGANIZATION BECA USE - - -

1. formal documents (e.g., job description sheet, etc.) say I should communicate with them. .71

2. my boss told me to. .73

3. it is expected I will communicate with them. .68

4. a work order required me to. .71

5. they are linked to me in the organizational chart. .58

Normgive ﬂame: or = .89

(Individuals are seen as driven by shared norms, share philosophy of management, and some

speciﬁc culture.) Factor
Loading

I COMMUNICATE WITH OTHERS AT THE ORGANIZATION BECA USE - - -

1. our relationship is important to fulﬁlling the company’s mission. .82

2. our relationship helps to maintain the company’s integrity. .81

3. this relationship contributes to the development of teamwork. .81

4. our relationship is important in accomplishing our job. .80

Sentiment: or = .82

(Individuals are seen as driven by ﬂiendship and other more emotional bases.) {3215?

0 mg

I COMMUNICATE WITH OTHERS AT THE ORGANIZATION BECAUSE - - -

l. l have fun with them. ~85

2. I like them personally. 1-00

3. we do things together outside of work. 53

Negotiated Order: or = .92

(Unique mix of the forgoing ﬂameworks to choose what framework (or combination of

ﬂamework) will govern their interactions.) Factor
Loadrng

[COMMUNICATE WITH OTHERSAT THE ORGANIZATION- - -

1. SO THAT we can come to agreement about some action we should take. -78

2 So we can decide what we will be doing in the future. 92

3. So we can come to agreement on an issue. 2;!)

4 .

SO THAT we can decide on how future work on the job should go.

 

113

TABLE 7
Means, Standard Deviations, and T-tests: F INT Subscales in Americans and Japanese

 

 

 

American Japanese Total t-Test
(n = 89) (n = 63) (n = 152)
Mean Std. Mean Std Mean Std t Sig.

Deviation Deviation Deviation (2 tailed)
Exchange 5.72 .88 5.32 .80 5.55 .87 2.297 .004'
Formal 3.69 1.44 3.39 1.06 3.57 [.30 1.495 .137
Normative 5.47 1.32 5.30 1.08 5.40 1.22 .856 .393
Sentiments 4.27 1.52 3.52 1.66 3.96 1.62 2.811 .006
Negotiated 5.33 1.18 4.56 1. 32 5.01 1.29 3.707 .000”
Order
FINT 24.43 4.21 22.1 1 3.42 23.50 4.06 3.805 ."000

 

“' Signiﬁcant at the 0.05 level (2-tailed).
" Signiﬁcant at the 0.01 level (2-tailed).

114

TABLE 8
MANOVA Outcomes: Cultural Difference on Five Dimensions of FINT Scale

Multivariate Tests of Signiﬁcance (S = l, M = 1 1/2, N = 71 1/2)

 

 

Test Name Value Exact F Hypoth. DF Error DF. Sig.
Pillais .13540 4.54145 5.00 145.00 .001"
Hotellings .15660 4.54145 5.00 145.00 .001“
Wilks .86460 4.54145 5.00 145.00 .001”
Roys .13540

 

Univariate F-tests with (1,149) D. F.

 

 

Hypoth. SS Error SS Hypoth. MS Error MS F Sig.
Exchange 5.74561 107.90092 5.74561 .72417 7.93409 .006“
Formal 3.73159 250.68788 3.73159 1.68247 2.21792 .139
Normative .86271 22367951 .86271 1.50120 .57468 .450
Sentiments 19.09719 373.20303 19.09719 2.50472 7.62449 .006“
Negotiated 22.02538 229.04057 22.02533 1.53719 14.32838 .“000

Order

 

“‘ Signiﬁcant at the 0.05 level.
"”" Signiﬁcant at the 0.01 level.

115

82833 2358 08 23.8— 38: 5 23:5: 2:. 882
.28—88-8 85— _o.o 2: 8 “505.8% .3
.30—88-8 _o>o_ moo 05 8 2805,85 ..

 

 

 

 

 

 

 

 

 

mm. mm. 2. S. 2. m4. 8:950:
Fooo. onco- tooo. omvo om. vo. mm. . 3.. mm. 2.. -owm mo 0:3.
R .2 VN .2 a .2 No.0 SN 3- .: x338: 8:80
tooo. 2 m6. moo. ooh; end— mo._N ova o3: was N38 .owm .«o a =Eo>O
Fm. 3.. 3.. mm. 2. mm. x838:
:Noo. ommxn- 2 m. _ _o._ _N. no. «am. S. m _. on. .owm he 0:5—
A N 8 mm .o mm .m a. .8. Nu .m N o .m x838: EEoU
pomo. obvm- mom. too. mod No.8 36 who 2d 2.» -owm we a Eoom
mm. :2 mm. mm. mm. :2 x838:
i.e.—o. m _ 9m- pvmo. mmm.m mm. 3. 3. mm. om. oo. -owm .3 28m 8800
no .» NW .w MN .0 as .m R .N .3 .o 0:950: 8:2
:Qo. God..- :m. So. and. omd end owé nod 36 -owm .3 a 82.8qu
mm. MN. 2. 2. 3. 3. x838:
-ooo. wows- i-ooo. mo:- ov. oo. om. :V. om. 2.. .owm oo 28m .550
mm .m- G 6 mo .8. om .m VQ N on N #538: 333.5
Fooo. mg .m- tvmo. mmm.m Q .o no.” o~.o_ woo mogm moo— -owm he a £2.
.wﬁ a .w_m a 95830 95%.: n:2wSO 9585 9.23:0 9.8m:—
asoLwSO macaw—ho So u 5 Gm n 5 CV u 5
8o 80-..-“ Leo 8a...-“ .58. 82898 mama-88¢.

 

328%.. 28 280522. E manages-ow”; 880-: 98 8:28.239 ohm—ego .8862
o mamas-w

116

TABLE 10
MANOVA Outcomes: Cultural Difference on the Number of Ego-networks

Multivariate Tests of Signiﬁcance (S = l, M = 2 , N = 28 1/2)

 

 

Test Name Value Exact F Hypoth. DF Error DF. Sig.
Pillais .53142 11.15226 6.00 59.00 .000"
Hotellings 1.13413 1 1.15226 6.00 59.00 .000"
Wilks .46858 1 1.15226 6.00 59.00 .000"
Roys .53142

 

Univariate F-tests with (1,64) D. F.

 

 

 

 

 

Hypoth. SS Error SS Hypoth. MS Error MS F Sig.
Job- lngroup 159.90423 2681 .26244 159.90423 41.89473 3.81681 .055
23::(1 Outgroup 669.5181 1 1 189.51220 669.5181 1 18.58613 36.02246 .000"
Org.- Ingroup 19.71951 1960.78049 19.71951 30.63720 .64365 .425
31:32 Outgroup 188.81709 1224.94049 188.81709 19.13970 9.86521 .003"

 

Social Ingroup 29.84835 2913.68195 29.84835 45.52628 .65563 .421
Comm w

 

Outgroup 313.60290 2162.88195 313.60290 33.79503 9.27956 .003

 

* Signiﬁcant at the 0.05 level.
*" Significant at the 0.01 level.

117

TABLE 11

MANOVA Outcomes: Cultural Difference on the Percent of Ingroup Communication

Multivariate Tests of Significance (S = l, M = 1/2, N = 30 )

 

 

 

 

 

 

 

Test Name Value Exact F Hypoth. DF Error DF. Sig.
Pillais .42634 15.35932 3 .00 62.00 .000"
Hotellings .74319 15.35932 3.00 62.00 .000”
Wilks .57366 15.35932 3.00 62.00 .000”
Roys .42634
Univariate F-tests with (1,64) D. F.

Hypoth. SS Error SS Hypoth. MS Error MS F Sig.
Job-oriented 1.38553 1.91356 1.38553 .02990 46.34001 .000"
Ingroup
Org-related .45011 5.70930 .45011 .08921 5.04564 .028'
Ingroup
Social Comm .11367 7.37280 .11367 .11520 .98668 .324
Ingroup

 

"' Signiﬁcant at the 0.05 level.
” Signiﬁcant at the 0.01 level.

118

TABLE 12

MANOVA Outcomes: Cultural Difference on the Percent of Outgroup Communication

Multivariate Tests of Signiﬁcance (S = 1, M = 1/2, N = 30 )

 

 

 

 

 

 

 

Test Name Value Exact F Hypoth. DF Error DF. Sig.
Pillais .4643] 17.91297 3 .00 62.00 .000"
Hotellings .86676 17.91297 3.00 62.00 .000"
Wilks .53569 17.91297 3.00 62.00 .000“
Roys .46431
Univariate F-tests with (1,64) D. F.

Hypoth. SS Error SS Hypoth. MS Error MS F Sig.
Job-oriented 1.38553 1.91356 1.38553 .02990 46.3400] .000"
Outgroup
Org-related .49774 4.75040 .49774 .07423 6.70579 .012"
Outgroup
Social Comm .72648 3.06919 .72648 .04796 15.14881 .000"
Outroup

 

“‘ Signiﬁcant at the 0.05 level.
" Signiﬁcant at the 0.01 level.

119

TABLE 13

Correlation of FINT and Ingroup-Outgroup Communication (American Sample)

 

Exc le Nrm Snt Neg FINT Ing— Out- Ing- Out- Ing- Out-
job job org org soc soc

 

Exc 1.00

PM .32‘ 1.00

Nrm .46" .27 1.00

Snt .07 -.07 .24 1.00

Neg .62“ .40“ .53“ .09 1.00

FINT .68” .63" .77“ .44“ .77” 1.00

Ing-job -.35‘ -.50” -.42" -.21 -39” -.57” 1.00

Out-job .35‘ .50" .42" .21 .39” .57“ -100“ 1.00

lng-org -.27 -.38" -23 -.10 -.38° -.43" .51“ -.51“ 1.00

Out-org .47” .32“ .33” .15 .54“ .53“ -.3s‘ .38‘ -.73 1.00
lng-soc .11 -.22 -.11 .04 -.07 -.10 .20 -.20 .29 -.23 1.00
Out-soc .11 -.01 .14 .36‘ .30 .26 -.08 .08 -.11 .28 .03 1.00

 

"‘ Correlation is signiﬁcant at the 0.05 level (2-tailed).
" Correlation is signiﬁcant at the 0.01 level (2-tailed).

120

TABLE 14

Correlation of F IN T and Ingroup-Outgroup Communication (Japanese Sample)

 

 

Exc le Nrm Snt Neg FINT lng- Out- lng- Out- Ing- Out-
job job org org soc soc
Exc 1.00
mm -.02 1.00
Nrm .27 .02 1.00
Snt .40‘ -.01 -.02 1.00
Neg .42‘ .10 .13 .10 1.00
FINT .68" .36 .41’ .65” .65” 1.00
Ing-job .03 -.12 -.06 .24 -.19 .00 1.00
Out-job -.03 .12 .06 -.24 .19 .00 -1.00 1.00
lng-org .24 -. 13 .04 .21 -.05 . 12 .37 -.37 1.00
Out-org .09 .34 .02 -.27 .16 .05 -.13 .13 -.73” 1.00
Ing-soc -.16 .17 -.12 -.02 -.46‘ -.22 .24 -.24 .41‘ -.21 1.00
Out-soc .08 -.40 .07 .01 .43‘ .10 -24 .24 -.17 -.06 -.80” 1.00

 

"' Correlation is signiﬁcant at the 0.05 level (2-tailed).
"”" Correlation is signiﬁcant at the 0.01 level (2-tai1ed).

121

TABLE 15

Contextual Models of Multinational Organizations

 

Communicative Interactions in Acculturation

 

Divergent

More ingroup communication.

Convergent

More outgroup communication

 

 

 

 

Monolithic Home ETHNOCENTRIC ETHNOCENTRIC
Culture Culture DIVERGENCE ASSIMILATION
Organization - Two culture groups have - Host culture group has
more ingroup comm. more outgroup comm, but
Frameworks tend the other does not.
’0 conform - Home culture is dominant. - Home culture is dominant.
- Frameworks conform to - Frameworks conform to
the home culture. the home culture.
Host POLYCENTRIC POLYCENTRIC
Culture DIVERGENCE ASSIMILATION
- Two culture groups more - Home culture group has
ingroup comm. more outgroup comm, but
the other does not.
- Host culture is dominant. - Home culture is dominant.
- Frameworks conform to - Frameworks conform to
the host culture the host culture
Plural Culture PLURALISTIC GEOCENTRIC
Organizations SEPARATION CONVERGENCE
- Two culture groups more - Two culture groups have
Frameworks diﬂer ingroup comm. more outgroup comm.
1” “”0 3'0“” - Two cultures conﬂict. - Two cultures coexist.
- Frameworks differ in two - Frameworks differ in two
groups. groups, but they learn
from each other.
Multicultural MULTICULTURALISM
Organizations - Two culture groups have

Frameworks are
Synthesized.

balance in ingroup and
outgroup comm.
- Alternative culture exists.
- Frameworks are integrated
for organizational goals.

 

122

APPENDIX B

FIGURES

123

Means of Five FINT Dimensions

T—_._‘ " American} (in = 89) —“ T—Japan’es’eim {65) l

 

 

*4

 

 

 

 

 

 

 

Exchange Formal Normative Sentiments Negotiated Order

FIGURE 1

Means of Five FINT Dimensions for Americans and Japanese

124

Job-oriented Communication

 

 

 

g

0"

a

"1

0

H5

'6

o

O

’E’.

0 St::1.Dev=9A3
Mean= 15
N=66

 

O 5 10 15 2O 25 30 35 40

Number of Ego-networks

FIGURE 2
Histogram: Number of Ego-networks in Job-oriented Communication

125

Organization-related Communication

 

 

 

 

30

g 20‘

3.7

o

"h

5

.2 10

(D
Std.Dev = 8 .62
Mean = 9

0 N = 66
0 5 101520253035
Number of Ego-networks
FIGURE 3

Histogram: Number of Ego-networks in Organization-related Communication

126

Social Communication

 

20

 

g

0"

o

'1

O

H:

"U

o

O

'E-l

n
Std.Dev= 10.49
Mean= 11
N= 66

 

O 5 10 15 20 25 30 35 40

Number of Ego-networks

FIGURE 4

Histogram: Number of Ego-networks in Social Communication

127

Number of Ego-Networks to Ingroup Members
Arum . m. .—, . o1

aldoad J0 JaqumN

 

 

   

Number of Ego-Networks to Outgroup Members

Mam , .. Am 7 .2 _

 

aldoad J0 Jaqum N

 

   

FIGURE 5

Histogram: Ingroup-Outgroup Communication in Job-oriented Communication

128

Number of Ego-Networks to Ingroup Members

 

 

 

 

E
a
S
O
5
1%:
Number of Ego-Networks to OutgroupMembers
, . , ,, ﬁat-m- 7, g
‘ l
1 1
l
l 1
z— " l 1
a l
g l
3 1
.9
1°; 1
| l maj
FIGURE 6

Histogram: Ingroup-Outgroup Communication in Organization-related Communication

129

Number of Ego-Networks to Ingroup Members

 

 

2

C

a

8

5,

3’

:3.

Number of Ego-Networks to Outgroup Members
, , gm 7 ‘ , m
1 l

l
l

C 1

a 1

8

a, 1

8

1°;

0 l

‘1 10 '5 2'0 7 2'5 7
FIGURE 7

Histogram: Ingroup-Outgroup Communication in Social Communication

130

 

APPENDIX C

QUESTIONNAIRE

English Version and Japanese Version

131

A SURVEY ON COMMUNICATION AND VALUES

This survey seeks to explore communication and culture within an international
organization. Your participation will provide very precious information for this study.
This research is conducted solely for the purpose of academic research by a graduate
student at Michigan State University working on his doctoral dissertation. Your
responses will be used strictly for his research only. No personal information will be
disclosed to anyone other than the researcher listed below. Your responses will be kept
conﬁdential from the company, and no part of your responses will be used to evaluate
you in any aspect. A summary of the results will be made available to you at your
request. The participation of this survey is voluntary, and you may reﬁise to answer
certain questions. You indicate your voluntary agreement to participate by completing
and returning this questionnaire with your name on the last page.

It will take approximately 15 to 20 minutes to complete this survey. After ﬁnishing it,
please enclose it in the attached envelope and seal it for conﬁdentiality. Then, Mg
return it to Mr. ABCD. If you want to directly return it to the researcher, you can mail
it to the researcher’s address bellow. I would appreciate it if you complete this survey
and returned it by January 8, 1999. Your participation is greatly appreciated. If you
have any questions or concerns, please contact:

Mr. Tom Kiyomiya
Department of Communication
Michigan State University

East Lansing, MI 48824

Ofﬁce Telephone: (517) 355 — 6615
Home Telephone & Fax: (517) 882 — 0174

E-mail: kiyomiya@pilot.msu.edu

132

Part A: Culture

Individuals and entire groups often think positively of speciﬁc beliefs, customs, and behaviors. When these
positive feeling are strong, such things are said to be valued. An individual may have speciﬁc values, and
so can a group or an entire organization if the people agree on the values. In this survey, we attempt to
identify such shared values among the organizational members as well as your personal values with regard
to some managerial perspectives.

What follow are many things that might be values in your organization (Ube Machinery Inc.) or that you
might value personally. Please use the following scale to describe (1) the extent to which each of following
possible values are operating and emphasized in the organization as a whole and (2) the extent to which
you value each item personally. Please note that just because something might be valued by most of people
who work in your organization does not mean that you will value it yourself. Please circle one number in
each column for each item. On a seven-point scale, with 7 indicating ‘highly valued’ and 1 indicating ‘not
valued at all’, please indicate which of the following items would be valued personally and in your
organization as a whole.

L A
V

Not Valued At All [1 2 3 4 5 6 7] Highly Valued

 

 

(I) In the organization W
asaWhole
1. Predictable outcomes
(being conﬁdent about knowing what
willhappenifcertainactionsaretaken)---1 2 3 4 5 6 7 1 2 3 4 5 6 7
2. Innovationandchange -------------- 1 2 3 4 5 6 7 1 2 3 4 5 6 7

3. Participation and open discussion
(through QC circles, suggestion systems, etc.)-1 2 3 4 5 6 7 l 2 3 4 5 6 7

4. Outcome excellence and quality
(the best outputs or results are more

important than the process) ----------- 1 2 3 4 5 6 7 1 2 3 4 5 6 7
5. Employee concerns and ideas ---------- 1 2 3 4 5 6 7 1 2 3 4 5 6 7
6. Getting the job done ---------------- l 2 3 4 5 6 7 l 2 3 4 5 6 7

7. Controlling the work process
(including schedule and information control)1 2 3 4 5 6 7 1 2 3 4 5 6 7

8. Creative problem solving
(appreciation to various opinions to
overcome some impasse) ------------ 1 2 3 4 5 6 7 1 2 3 4 5 6 7

9. Human relations, teamwork and cohesion--1 2 3 4 5 6 7 1 2 3 4 5 6 7

133

A

Not Valued At All [I 2 3 4 5 6 7] Highly Valued

 

 

 

1 1n the Or anization (2) You Personllly
M016
10. Decentralization (where many people
haveasayindecisionmaking) -------- 1 2 3 4 5 6 7 1 2 3 4 5 6 7
11.Goalachievement ------------------ 1 2 3 4 5 6 7 l 2 3 4 5 6 7
12. Order
(emphasize an organization structure or
rules to make smooth operations) ------- 1 2 3 4 5 6 7 1 2 3 4 5 6 7
13. Morale in the workplace
(respect trust and good relationships
withyourcoworkers) --------------- 1 2 3 4 5 6 7 l 2 3 4 5 6 7
14.Newldeas ---------------------- 1234567 1234567
15. Stability and continuity
(jobs become routinized and smooth
withoutinterruption.) --------------- 1 2 3 4 5 6 7 1 2 3 4 5 6 7
l6.Doingone’sbest ------------------ 1 2 3 4 5 6 7 l 2 3 4 5 6 7

Part B: Communication

In part B, we are interested in determining why you communicate with other people within your
organization (Ube Machinery Inc. ). The next 20 statements describe some of the possible reasons and
meg why you communicate with others. On a seven-point scale, with 7 indicating very likely and 1
indicating very unlikely, please indicate which of the following statements would be likely reasons why
you communicate with others at your company.

 

 

Very Unlikely |l 2 3 4 5 6 7| Very Likely

I COMMUNICATE WITH OTHELS AT THE ORGANIZATION

BECAUSE
1. they have something I need to solve a work related problem ------- 1 2 3 4 5 6
2. they can provide me with technical advice. ------------------- 1 2 3 4 5 6

134

L A
‘ 7

 

Very Unlikely [1 2 3 4 5 6 7] Very Likely
I COMMUNICATE WITH OTHERS AT THE ORGANIZATION

 

BECAUSE
3. I can help them solve a problem. ------------------------ 1 2
4. we both get something of value from the interchange. ---------- 1 2

5. formal documents (e.g., job description sheet, etc.)

say I should communicate with them. ---------------------- l 2

6. my boss told me to. ---------------------------------- 1 2
7. it is expected I will communicate with them. ------------------ 1 2
8. a work order required me to. ---------------------------- l 2
9. they are linked to me in the organizational chart. --------------- 1 2
10. our relationship is important to fulﬁlling the company’s mission. - - - - 1 2
1 1. our relationship helps to maintain the company’s integrity. -------- 1 2
12. this relationship contributes to the development of teamwork. ------ 1 2
13. our relationship is important in accomplishing our job. ----------- 1 2
14. I have fun with them. ---------------------------------- 1 2

15. 1 like them personally. --------------------------------- l 2
16. we do things together outside of work. --------------------- 1 2

17. SO THAT we can come to agreement about some action we should take.-l 2
7

18. So we can decide what we will be doing in the future. ------------ 1 2
19. §9 we can come to agreement on an issue. ------------------- l 2
20. SO THAT we can decide on how future work on the job should go. - - - 1 2

I35

80.22.08 888 8:8 :8 .283 8:88: .
880.8300 :23 8:88. bow 92.8.: :8 .22.... .88.: .:0::. 0. 80 w20w 8 :08 £222.08 888 8.80.8. .
.:.80w :8 ...0.::88 .0200: .8808 .030: 8000.. 83:0: .80.» 8 :08 .882 88.0.. 0.83-87. . 02:88.”:
8:2. 808:0 80.» .8 82:88.. 02:0. 88.880800 :8 02:28.. 888 88.8.8. 4:101:83 A0

.:0 8 :8 68:88 82888:: .0.::0:8 :2800> .0808: .883
80:80.08: 80.80 .539.» 8:280:08 82889.: .8 :08 .882 .0888: :0 2.8.8:: 08: :0.» 8:3 02>:0 .8 88:8 8.80:8. .
A08 .280w 88.88. .0228 :8 .3088 088:..00 £2.88 :.8 882:8: 8:282:88 80:80:80 8020:: .088 888:... ..w.0v
0282890 0:. 2 088 0.88.. 8:8 808:0 .
830:0: 0.83 :8 .0888 .280 0.83 8 :08 .88.. :8 08.8: 88.82830 0. 8.80. :23 8.88.8.8. . 02:88.8.
8:2. 808:0 80.» 0. 88.0.. .».80.:: 502 8282888 0:. 8.8.80. 8.88.8.8. 80.8. 820.2022: 5 Am.

80:80... 80.8.: ..0 88.8: ..:0.» :23 0:28.88: :8 8:28:02: £2802 .
:0“ 808:0 .80.» 0. 8.80. :23 :2200: 0 888 0. :2280 08: :0.» 00:80:08 8 80::8 .:.0: 8.88:5 .
82:8: 88.03.83 8 0>.8 0. 8.8.80.8. :8 003:... .
8:2. 80.» :0 0.00::00: 00888.80: 8 :8 88:8 :2. 80::8 :8 008:... 82:80... . 02:88.”:
.80.: 808:0 80.» 0. 808.0. .»..00.:: 8.88.0.8. 850 :8 00888.80: 808:8... 3:284:10“. C.

808 822: 80:. 8.. 88200 028080.. 20:. 2 .».088:8 208:8 .80.» 80:0: 0. :0.» 00... :.:03 03 “80.3. 80800

:08

8:. 80... 02.3 080.: .2003 0 08: 0:0 8:. 0.88:8 830.. 08: :0.» .2 .2003 8 08: 0:0 :2: 888w 888:8: :0 88:00.. 8880.. 0:. 002m .320:

820.98 0: E3 8:. 808 82:00 00:: .8 0.88. 2 88:8: .8 8:80: 2: :30: 02:3 080.: .2003 ..0: 88.. .8 ..0:8:: 2: 0.. .2003 0 2 0.0880 :8:

:0.» 88: .»:08 30: .».08::x0..::8 082:2 :8 .:8..0: 25 :23 808 80800 2: .8 :80 2 208:8 :282::8800 :0: :0.» 8:0 30: 80:0 0.2:. 080.:
88:88....—

.:282:::.88 .88-”: :.8 08:88. 88 8: :2828888 0080.08,. .».:0 8: 0:202 208:8 80> 6

:88: 88:00 0:. .8 00.8: 808.88: 8 088: :288 0:. 02:3 88 080.: .00— 3.02:8:

0: 3 2:23 .0888 8.82.8888 8 :8: :0.» 80:3 :23 0.:00: 0:. .8 :0 8188423 05 0:..3 080.: 82828888 08: :0.» 80:3 :23 082:2

0. :0.» 0.8 :.:03 03 80:... 80:3 0. 8.828888 2 0:3 305. 0. 00... 2:03 03 8282890 ..:0.» 2 8.838: 888 :0 08:00: 8880.. 2:. 2...
0:08:00 8388:8800

.88: 2:. 0. 2 ::8

:8 .»:8 0 008.: 080.: 8.8.: 0.08 :00: :0.» u. .88: 92328.. 0.: 2 888:8: 0:. :30: 02.3 :8 888: .2. 0. 8.8.: .02.: 80:0 08 08:... 888:8:
20.: 082:2 :8 88:88 80.» 2:23 208:8 8.828888 :8: :0.» 80:3 :23 888: 0:. :0 8:8: 0:. .2. :8 5.8080 82808:. 2232.0: 0:. :00.
080.: .888 88 8:282:88 .:08.0..:0.» 808 80800 00:: .8 0.88. 2 ft... 80:88 :48» 2:23 80.838: 8.828888 80.» 2 8808.2 0.8 03

 

 

0....038z 8.8288800 no .8:

136

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

.208 88.82830 8:8:-:0: 808.880 \ :280m 0807.
F A2003 8::888:~.0:.M 0.0380 028288800
73.... E7. 2.5.200 «.2.. 8:23 mh0<HZO0 20520—752200 §0>_
m .W
W :0... 00.0... M 0008000: .8 .28. 8.00....
m m 20:3...
N m H M N 002028.502 820m. 2... 00805
. m ~0H::00 8.3.030 08:80.30: 0.00.:
.200m m 802.8890 8.0.088: 80:58:00 \ :280m 0807.
A2003 8:. 88:38:? 0.00.80 8.8288800

 

 

.880 88:00:: 0:. 2 ... 8:. m8... :8 .080: 8:08 2.. .080: 2.. 02:3 :0 > 2.8:. 0 8:2. :0 0.::00 0 :2:

:23 82.00:. 28.892 8: 8.80:2 08: :0.» :8 .8280 :00w 0 2 0:. .808 8:282:88. 882 8080 :8.» 8:88 0. :2:.m .2). 0. :028. 08: :0 > o

.808 880m. 2003 0 00:0 :2: :23 0.8m :.0w 0 .»0.: 0.

0w :0.» :8 .::0.:: 0 8 28:28.0: :00w 0 08: :0.» .:22::0 2 .808 88.82880. 2003 0 00:0 :2: :23 .3088 m2 :8 8.82880 08...: 0:. 8:02:
8.2.088 :0.» :8 288880.... 2 008 888.78... 2003 0 023. .»88:220:::0 082:8: 8282008 2:8 0. :8m :2 80:: 8:8: 82.8.88 08: :0 > 0

288882.... 2 .08.0:.:0... .8 8.8 0:. 2 .03... 082:2 :8 0:8: 808:0: :8 8:8: 2: 02:3 :0.» 60:...

.2003 0 8.2. 03: 8:. 0:08 8.: :23 0.0088 03...: :0.» 8:. 0808 .. ..»0:.»8>0 .0088 02.60 .0w 0. .82>:0:8 82:200. 0 8280.22 :2 888:: :0 > o

:50 JAE 0:. >90: :0: mmaméa

I37

 

 

 

 

 

 

 

 

 

 

 

 

 

Econ

 

3.823.530

 

town—05b2,

 

Eustaavn \ steam

oEaz

 

 

99?: .55 “32.2.3.5

 

303.80 =c=3_==EEoU

 

 

:2:—3:8- T: 32 EEOU 1:. 55!» mHD<hZOO ZO~H<U—ZDEEOU MOO:

138

dowﬁomooo So» .8.“ :2:: b? no» xcaﬁ

 

505:2 083%: .v

3:9on 88.3 550 .b 805%.. .c 533:2 0382 .m
535:3 563an .m cactus 574 .N anotoE< 58$? .—

ﬁSéwxoan .8830 59A wear—omen “man 523 338:: 05 29:0 385 3:35.:—

 

 

 

oEZ .N

 

52.2.: 33% :3 .8» «a... own-.954

 

mm?» u>¢mZ—=U<E mm: 5 axiom no 33>

 

SEE—.50 2: E .558.— 25>

 

oEEom ._ 38m "2.32 25>

 

«I. H2<Z >Z<EOU .33., H0:52 hang—.50

.3385: again—can: 2: 0.9.6 was 9:33 2: 5 Eu cane—m

3:530 9529—25

139

ﬁﬁninzﬁ—VaykﬁﬁﬁKﬁT67V7wb

:07V7—b%£mtgébﬁaﬁéﬁwmﬁgwﬁuﬁﬁmtatﬁbrwamt
itéﬂmyw—7wmfwkﬁf6Eﬁmﬁﬁﬁﬁﬁﬁ¢6mémﬁaWeLife
&<Ke@u@%ﬂﬁmﬁﬁibft%®%&E%E¢6twutﬁﬁwﬁgiﬁtw
maaétﬁﬂtknifo$7V7~kﬂﬁi%IK%¥$ﬁxamﬁﬁbvtb<
if$$ﬁ%m%mtwgﬁuﬂ%héﬁﬁwaﬁDifoLtﬁort—A—Aw
ﬁgim:mmﬁawumaﬁbn\ﬁAﬁﬁﬁﬁﬁﬁmﬂuﬁnﬁbné:ku&L
TbniﬁboitﬁA®7V7—bﬁﬁu%&amﬁﬁﬁf&btﬁmwﬁiagu
ﬁbn5:&m~wam:tétﬁﬁﬁ$btﬁif°$7y7—anémmb<i
Tbamﬁﬁfbntﬁiwﬁﬁﬁaugitw:tb‘#A~A®a$ﬁiffo7
V7—F®%%&—9KE£WE%AWtﬁth\ﬁktt7V7—Féﬁwhtﬁ
Wt:&&BOT\$ﬁ§A®iwﬂﬂéﬁéhtlkkéﬁfwtﬁéifoﬁﬁt
%E%%ujﬁbﬂbUiLtB‘£ﬁ®ﬁffﬁ€$btﬁifo

:UDTWI— Nita? < 3 omx<- Hcﬁnmt E’ENCbxrbxéBﬁFaﬁm‘a 15~205N1 8
6T0f&T®N—FEKTLK&‘$7V7—bémﬁwﬁ%uxhﬁmbtﬂﬂg
3b ifrﬁHTén,itmﬁ%uﬁﬁﬁ%émgfé%utUTwEWivﬁi
V) Tau/x, 1 9 9 9%”! 1 5 BiﬂU’wf— bmwé'aikkiiﬁléﬁ‘rwtriét
inkﬁUiﬁz

iétwﬁﬁuﬂﬁffﬁ‘85€ﬂﬁhwﬁéléb<ﬁﬁw$bbﬁifo

7218‘ :EFLS ' ianEIW’S‘bﬁiiﬂ'Fi'C‘fﬁﬁTél/‘o

 

iﬁé‘ ta KIYOMIYA, TORU

i V77 I/‘H‘l 317V?" Department of Communication
13 ‘Z :1 51’7— “/ a Viiﬁi Michigan State University

‘2 ‘I/ﬁV'JH/f “2 }~ '3 I/“/‘/ 7H1” East Lansing, MI 48824
Tivﬁéﬁg USA

Ofﬁce: 517 - 355 —6615
Home: 517 — 882 — 0174
Email: kiyomiya@pilot.msu.edu

I40

"'“i‘ A: mam-am

A4$7W—7$¢m\LdLﬂbbﬁEwﬁﬁmamtﬁmuowrﬁﬁw(ﬂmw)mﬁi
£30:wiitﬁimﬂﬁﬁﬁwkétfwlikEmuﬁbrmﬁéﬁbé/Hﬁbrw
61téiifo:Wﬁﬁﬁuﬁk%ﬁmb@?b&ni1#\ﬁiwﬁﬂwmvbéﬁﬁLt
ﬁﬁﬁéﬁﬁié:tbﬁvifo:0:5kﬂ$%tmﬁﬁ%%ﬁsztmt:mmﬁwi
thBb‘T‘To
HTﬂtbktﬁﬁAmK‘béwmbtt®%ﬁ(XYZ)tbftﬁﬁiﬂbgﬁféﬁﬁ
ﬁﬁbékﬁbnéﬁavfooib&m%h€hwﬁautméﬁtwﬁmﬁéwtbrmﬁ
EﬁwbthniﬁALtQWﬂtﬁﬁAmeméﬁbbmbLniﬁbochutmﬁo
@wkhﬁﬁébfwbﬁmﬁﬁt%th&ttﬁﬁAmK§Ewa5bwt—ﬁbfwb
aumatwmbffo%:T%h€h®ﬁauﬁtmmmﬁkLfﬁﬁbfwéﬁﬁﬁta)
EAWEEEWZmeKwat7%%ﬁﬁ®¢#5ﬁ%fbﬁ$ﬁﬁ&ﬁﬁf?éwe7d
fﬁﬁu§§L1w61\lufiot<ﬁﬁbaw1&§%bifa

 

iot<§ibmw h 2 3 4 5 6 ﬂ ##KEELTwé

l
‘ 7

 

 

(Hﬁﬁiﬁﬁﬁic‘: LT (2)&7£f:ﬂé1ki:&: L'C

L $MW$$§¥
(bbﬁmmgakﬁﬁébtbfm
bamuwﬁﬁﬁézk) --------- 1 23 45 67 12 34 56 7

2 ¢¥®$mt§$ ----------------- 12 34 56 7 1 23 45 67

3 ﬁﬁmmam&ﬁmntﬁﬁ
(QCﬁ—7W$E$ﬂﬁﬁgéﬁbf
EEWK&§%%:%m36:t)U-12 34 56 7 1 23 45 67

4 ﬁhtﬂ&°¥ﬁ&§®§é
(%E®%%M@E&EKELT

inﬁgfbé) ------------------ 12 34 56 7 1 23 45 67
5 ﬁiawiieﬁb$ -------------- 12 34 56 7 1 23 45 67
6 ﬁgéﬁﬁiﬁ%736:t --------- 12 34 56 7 1 23 45 67

7 ﬁ$®ﬁ€€§ﬁfazt
(ﬁﬁ%X79;—»®%EE€U)H-12 34 56 7 1 23 45 67

& mﬁmmgma

(mwuuwgnftama%mma

$iht§§t§ﬁé§§¢6)-n-12 34 56 7 1 23 45 67
a Amw%tf—Av—7tﬁ%h ----- 12 34 56 7 1 23 45 67

uxﬁm®QHM(m%®§<mA4ﬂ
EE&E®%K%§$$E6) -------- 1 23 45 67 12 34 56 7

141

ll.

12.

13.

 

 

iot<§§bﬁw h 2 3 4 5 6 ﬂ ##KHELTwé

L
‘ r

(uﬁﬁéﬁkbf

Bﬁﬁﬁ ------------------------ 1 23 45 67

m%w&¢(ﬂ$émﬁufatw\
mmmﬁemamw—»2§§361n-1 23 45 67

ﬁ%u3ﬁbﬁgﬁ

(ﬁ%’6®ﬁﬁ¥kafaﬁ SEQ: t

m$mxmweﬁ2u§361 ------- 1 23 45 67
.ﬁbw74?47 ------------------- 12 34 56 7
.fiﬁtﬁﬁﬁ

(¢$#§ﬁm/Bﬁm¢5:kuioT\

ﬁﬂhé:t&<ﬂ%t£bgné)n-1z 34 56 7
.éhéﬁ<f:t ------------------- 1 23 45 67

N—F B: mﬁﬂﬁﬁéﬂinz#*92V

(a5at@xmaur

1

234567

CCTH\bﬁtﬁﬁﬁKEWthﬂwkkﬁﬁ(33137—937)€30mtfwﬂ$
ékﬁwLiTOmT06Imtﬁmwm®Atgﬁﬁmb(mM)titﬁmtwm§mfé
#(E%)uowr\bvﬁéﬁaéﬁxrmifo%n%niuomrgwﬁﬁaﬁmﬁrm
1 ii riot< ’é’l'fii

iém&\7&%%ﬁf3§i?éwo7ufﬁﬁtl<%1mi61t

ietw1&§%bi10

 

iot<%ruietw h 2 3 4 5 6 ﬂ ##Ki<§fﬁia

 

L k
‘ r

ﬁft/ab’lx 1 Tﬂﬁﬂw/t é' a 2‘ :z :i—f/a yéﬁow/L

ﬁ$t®ﬂgﬂﬂtwutﬁiﬂﬁgabwéﬁorwémbonu 1
m¥#&WW%Eéﬁif<nbmeo ---------------------- 1
wiwamurwamgwmnuﬂmaﬁo:kﬂvaameo-n1

ﬁﬁéﬁbfﬁﬁbébmékuﬁenameo ...............

ﬁKié<ﬁﬁEﬁ¥&E)K%5Téi§ﬂ$énfwémaou1

142

1

234567

234567

 

 

iot<§rui66w11 2 3 4 5 6 ﬂ ##u2<§ruia

I;
V r

£ﬁKEWTﬂﬁﬁ0A&J$2:#—V3Z§#00#Lm

10.

ll.

12.

l3.

14.

15.

l6.

17.

18.

19.

20.

iﬁ#%5§5m60 .................................... 1
%5T6:&E%ﬁénfw6m5° ........................ 1
ﬁ$iwﬁ%ﬁm6° ------------------------------------- 1
mﬁ®®tﬁm$tﬁwowrmbmeo ..................... 1
ﬁ®&%ééﬁ36tbKXWﬁmbo ....................... 1
KWT®%UHéEﬁﬁ36LT§§ﬁmBO ................. 1
f‘AU—7®%%Kﬁﬁow6° .......................... 1
ﬁﬁéﬁiégﬁmao ---------------------------------- 1
¥Lb‘ﬁ>63 ------------------------------------------- 1
MAmuﬂgﬁmeo .................................... 1
ﬁ$uﬁﬁtﬁééwﬁ26meo .......................... 1
Txéﬁﬂcowrmikﬂmﬁﬁwméﬁfétwo ---------- 1
ﬁkﬁﬁﬁff&élkém$tﬂﬁ&bbtwo .............. 1

ﬁ$$ﬁ$ﬁﬁ$ﬁquTm$kﬂﬁﬁﬁﬁﬁéﬁfétbo""1

witﬂﬁiﬁﬁéwﬂrgilowfEﬁﬁﬂ’\’é7b>9tﬁ>67:bilo ------ 1

I43

234567

swPP32IEAm61162\*.\m\161nmnemmu2#6 4926163<Pexl<2 a
346:6942M. 2661261< mu nanvgﬁtﬁ.
«\62. 2W3622312mtuahxxn 41nuxn 6 23 ,22.
2422£22P211«n\*3592226222 \mx1:42?a;2::.$2

Am#§AHIA2,wm.ﬁm,ﬂwﬁvP;O2hm+ke2#£.
Aﬁm§m&§§mw,ﬁﬁﬁm,ﬁﬁg,ﬁﬁﬁw 22E2P302mwe$<&%weﬁ4.

6xu2. .EMMQPE2$EE< Eiﬁkmt£2§$.

2 “#ﬁWER ﬂ§%§22326\u2. .Emmmkﬁ2 BE .
22K$#eﬁﬂﬂ22%ﬂeﬁ%q22,ﬁﬁnsa2ﬁﬁuiﬁﬁmtﬁ2ﬂw \m\12141ne

  

   

;,;A$

M2322 “2me EH. k\Vklw@t.nwu 31] $2 AHA, DBGQTVUO .
$122.42“ ,ﬂmgtﬁ Qﬁeﬁﬁ &H@%1.EL#,E.NE M24. Rﬁnwpmmtwﬁww.

ELQWHEHNW Niummcﬂrnm AM 2 2. VJ EMVHWEQHE .
ﬁﬁnbwminwﬁﬁ {awnYéLQWﬁ .

 

oswPPJ<ﬁ~2P§222,vﬁPMﬁ22m2PJE2e¢;#%€PEEHE,QumsPme222AmAlb
unmneHaﬁﬁmﬂPmﬁ%osmeWRwﬁew2iwﬁeAmblbnnmn,2mHm;~@m=H&$3,Em222ﬁ;
,2222nswPKﬂHPt$2ﬁEeOmmhgﬁ2Pa,éeuotw2am\nPwﬁﬂﬁem2vwusmemﬁPﬁﬁm
#ew,é202mxm61bnnmnPﬁﬁesmvim2EmH,JR2<G:P:P2o&wAmblbnnmn2:§W%2#£Hﬂ
+29%aﬁﬁePélx$w&Ww,2éauquPowﬁm,ﬂuxmAlbnnmnmsPnu*
A PPﬁwam K:P+R4R 2H%€MENFQVungmmRngmé£¢QMEMEGRGP,2%2;
meWRP4123R ,wE.w€<20¢\m\,,lhlnmnP ﬁts Kﬁwcmmﬁﬂﬁom oﬁPﬁﬂPogwﬂﬁsm $52M»;
2gmﬁﬁWé20$NAAmAlhunmnv£ﬂuﬁﬁﬁ,Q2QZRNQID2aﬁexmblbunmneQmﬁﬁﬂPMEeu
KENEW ”<2okwxm51bnnmn
3mP3$RPJImnaPP#%Q,2¢%
,ggw‘mﬁi {32 oﬁﬁgﬁuniwmneomﬁéﬂmﬁwoiuﬁuglvxeurﬁ a;war/\mmuPﬁﬁ2EGHEwNREQW ,Mﬁmwmnejx
HoimAhmAlenmnvﬁﬁﬁ2ﬁé,

23911.33 .132—$65.3 ,uﬁnbémmmﬁvuwkaEGFH 0.21%;$ERPSOU7J¢WERGPKH~
«Gutwﬁ ﬁﬁﬁﬁbuﬂoh ,Hﬁi 6:23”: oémmﬁncnxm ,quwﬁmamwﬂﬁklb2 AﬁQAmAledmﬁnNﬁonﬂﬁmvﬂP

6162a*.5m\1 b1nm n ”o 21%

144

 

 

 

 

2.212112222222222 22222121121933. 2 22221211252232

IE

 

EQ

 

 

GREG a #SEH C a“

 

<uommxm112uumn

 

 

ﬁxmAIbnnmnﬁﬁg2ﬁﬁmtﬁ2Nta

 

KRN

 

 

 

 

 

 

 

 

 

W mmosommm .8sz £_Ew Emn<
N N .1 N mocchEmE ccom .< 3:52.
.2. .9250 32.30 922802252 So...
2221211....mnﬂwmﬂ 21:212u1mn921iﬁ W 2.212unmneﬂﬁﬂ IE 222$
2#@922£EMH2MR <HoﬁwAmAIbudmn
2 vm~2P§222
E9 22212.11", nenﬁﬁa wwwwmfﬁrwea2 QQPPSEHE 2.221222% mPﬁEE .m2m 92261121.
<12 Pmﬁﬂg omﬁﬁﬁugém 2.22 Snug/R2222 93022222511142 mummum 233 w» ﬁwmwmwnwﬁwmmmnﬁmjx .
2 u 1... n
29121122h12mw11 H2131. v@2m2 ,PwmthéhSMQ/xgﬁﬁPﬁﬁJa2w ovm 323282 2m3n2€uﬁﬁmw2 W
m£§u1Pﬁ;@v@ﬁm P302$meeﬁw P$EE<€$E i QSﬂﬁ ﬂmuﬂ £$m2m22v
MN 2m 12829 2312.11” neHWﬂJ. MMmeﬁmﬁweg P022211. .22 mmwmﬂ m ﬁaﬁﬁmk
3122222223 2.222.219.1111 :3Pt «$.12. Smwa 222122228922 022.312 Kmkbvxxew m2 Pmpﬁﬁﬂﬁ .
2va1ﬁ~ E2 2 8.22.12.11.89 ME ,PJW
ovmuzﬁcoobﬁsoEﬁmem ESE 222 2323M .2 ,mMW4123R228 .Eobweo2212ﬁe2a ﬁsPoﬁw
Ehwmenﬁjmmmm. 9M0 0,3102%:212 332m 32m 2895.. 2w£P1211w12122R1 2.21 92.112 .Eﬁgﬂﬂﬁ.

E§Kmm2

145

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Agnﬁnmngmﬁ m AMALaneﬂﬂam :Alﬁnmneqmﬁ ME EM
“wee u $mﬁ aw <Mo&wsm$lbunmn

 

I46

 

 

 

I T, mAleﬁwnﬁﬁeugﬁmmtmhu§>ﬁ

ﬂaxﬁtn—J &#$9 Ha. aux; ﬁswh m Nun sﬁu.§ ,Rﬁﬂexl lbkk»

 

“ﬂﬁﬁﬂwwwwﬂﬁ¢m,VJNWWE¥H€H#£

A§GRHSw<ﬁ2LIbANG$m%m,PR€$§<%mv

 

 

Apﬁﬁﬂ%mvﬁmwﬂﬂ2ﬁmﬁﬁﬂ£

 

M . m H 3% w ” Hﬂﬁnﬁmep N > x

 

 

”a a ”5...?”

OSwFéﬂwmmwwmwﬁﬂVﬂﬂﬂﬁﬂﬂﬁ,Wﬁﬂﬂwuwwﬁ%ﬁ

EMU 51mm

147

"‘iﬁﬂiﬂﬂjﬂlﬁﬂiﬂ'ﬂivlﬁﬂmm